---

 

CONSTITUTIONAL COURT OF SOUTH AFRICA

 

                        Case CCT 144/23

 

In the matter between:

 

INDEPENDENT CANDIDATE ASSOCIATION SOUTH

AFRICA NPC                                                                         Applicant

                                                                                                                                                      

and

 

PRESIDENT OF THE REPUBLIC OF SOUTH AFRICA        First Respondent

 

SPEAKER OF THE NATIONAL ASSEMBLY                         Second Respondent

 

CHAIRPERSON OF THE NATIONAL COUNCIL

OF PROVINCES                                                                      Third Respondent

 

MINISTER OF HOME AFFAIRS                                              Fourth Respondent

 

INDEPENDENT ELECTORAL COMMISSION                        Fifth Respondent

 

ALL POLITICAL PARTIES REGISTERED FOR

ELECTIONS FOR THE NATIONAL ASSEMBLY                    Sixth Respondent

 

 

Neutral citation:     Independent Candidate Association South Africa NPC v President of the Republic of South Africa and Others [2023] ZACC 41

 

Coram:                      Zondo CJ, Maya DCJ, Kollapen J, Mathopo J, Mhlantla J, Rogers J, Schippers AJ, Theron J and Van Zyl AJ

 

Judgment:                Mhlantla J (unanimous)

 

Heard on:                 29 August 2023

 

Decided on:              4 December 2023

 

Summary:                 Electoral Amendment Act 1 of 2023 — provision limiting seats that independent candidates can contest in the provincial and national elections does not infringe any of the fundamental rights in Chapter 2 of the Constitution

 

Right to stand for public office — Right to vote — Right to free and fair elections

 

 

ORDER

 

 

On application for direct access:

1.                  Direct access is granted.

2.                  The application is dismissed.

3.                  There is no order as to costs.

 

 

JUDGMENT

 

 

MHLANTLA J (Zondo CJ, Maya DCJ, Kollapen J, Mathopo J, Rogers J, Schippers AJ, Theron J and Van Zyl AJ concurring):

 

 

Introduction

[1]              The applicant, the Independent Candidate Association South Africa NPC, is a registered non-profit company that represents and promotes the interests of independent candidates in the electoral system.  The applicant has brought an application for direct access in terms of which it seeks a declarator that item 1 of Schedule 1A (impugned schedule) to the Electoral Act[1] is inconsistent with sections 1(c); 3(2)(a); 9(1); 19(2); 19(3) and 46(1)(d) of the Constitution to the extent that the impugned schedule provides for 200 seats in the National Assembly to be filled by independent candidates and candidates from regional lists of political parties (called regional seats) and 200 seats to be filled by candidates from national lists of political parties (called compensatory seats).  The applicant contends that Parliament acted unconstitutionally by splitting the seats in the National Assembly into 200/200.  This challenge stems from Parliament’s passing of the Electoral Amendment Act[2] which, amongst others, amended item 1 of the impugned schedule to allow for independent candidates to contest for seats in the National Assembly.

 

[2]              The primary remedy sought by the applicant is a combination of a striking down and reading-in, which will result in the word “half” in item 1(a) of the impugned schedule being changed with immediate effect to “350” and, similarly, the word “half” in item 1(b) of the impugned schedule being changed to “50”.  The reading- in will have the effect of amending item 1 of the impugned schedule to apply to the 2024 elections.  In the alternative, the applicant seeks a suspension of the declaration of invalidity of the impugned schedule for a period of 36 months for Parliament to address the unconstitutionality and an interim reading-in in item 1 of the impugned schedule in the manner set out in the preceding paragraph.

 

[3]              The application is opposed by the Speaker of the National Assembly (National Assembly), the Chairperson of the National Council of Provinces (NCOP) and the Minister of Home Affairs (Minister).  The President filed a notice of intention to oppose the application jointly with the Minister but did not file an answering affidavit.  The Independent Electoral Commission (Electoral Commission) abides the decision of the Court and has provided an explanatory affidavit.

 

[4]              In order to understand this case, it is imperative that I commence by setting out the electoral scheme that operated in South Africa since the advent of our democracy in 1994 and the events that culminated in the amendment of the Electoral Act.

 

The electoral system since 1994

[5]              The Constitution itself does not specify the national and provincial election system, instead, it outlines the requirements that the electoral system should comply with.  These requirements include the need for the system to be based on a national common voters’ roll, provide for a minimum voting age of 18 years, and result, in general, in proportional representation.  While there are a number of electoral systems that comply with these requirements, the Constitution leaves the decision on the preferred system to Parliament.  However, this does not grant Parliament unrestricted freedom.  This Court has made it clear that provisions which disenfranchise citizens or preclude them from standing for office are not permitted.[3]  At the heart of this case lies the question of how Parliament chooses to give effect to these requirements.

 

[6]              The number of seats available for contestation in the National Assembly is, in terms of section 46(1)[4] of the Constitution, capped at 400.  Sections 46(1)(d) and 105(1)(d)[5] of the Constitution require the adoption of an electoral system that results, in general, in proportional representation for the election of the National Assembly and Provincial Legislatures.  Sections 46(2)[6] and 105(2)[7] of the Constitution prescribe that national legislation must provide a formula for determining the number of members in the National and Provincial Assembly.  Pursuant to these provisions, Parliament passed the Electoral Act which, amongst others, regulates elections of the National Assembly and the Provincial Legislatures and provides for related matters.

 

[7]              From 1994 to 2019, the election of representatives in the National Assembly was based on a two-tier compensatory system, which filled the 400 seats in the National Assembly as follows: (a) 200 seats were filled from regional lists submitted by political parties; and (b) 200 were compensatory seats based on the national lists of political parties.  The nine regions for which there were regional lists corresponded with the nine provinces.  During this period, only political parties registered with the Electoral Commission could contest elections.[8]  This meant that an adult citizen was not allowed to contest an election as an individual.  Voters had one vote that determined the regional and national seats.  In other words, a voter in any particular region could only cast one vote on a single ballot paper which listed the contesting political parties, but his or her vote for that political party counted once towards the regional seats and once towards the compensatory seats.

 

[8]              The purpose of the regional seats was to ensure that voters in the respective regions (provinces) were represented in proportion to the population size of their region in the National Assembly.  Once votes were tallied for a particular province, they were divided by the number of National Assembly seats allocated to that province plus one, and one was added to the result (ignoring fractions), to determine the quota (the number of votes) needed per regional seat.  The weighted average of the regional quotas in the 2019 national election, in which 17 437 379 people voted, was 83 511 votes per seat.  This method of determining a voting quota – by dividing the total number of votes by the number of seats available plus one, and then adding one to the result – is called a Droop quota.[9]

 

[9]              The purpose of the national (compensatory) seat was to restore overall proportionality as between the represented political parties due to potential distortion created by the regional system.  The total number of votes cast was divided by 400 (being the number of seats in the National Assembly) plus one, and one was added to the result (ignoring fractions), to determine the Droop quota for a compensatory seat.  In the 2019 national election the quota was 43 485 votes per seat.[10]  To determine how many compensatory seats a political party was entitled to, the total number of votes achieved by the party nationally was divided by the compensatory quota per seat to give a provisional number of compensatory seats for the party, and the regional seats won by that party were then deducted from the provisional number of compensatory seats to arrive at that party’s final compensatory seats.

 

New Nation judgment

[10]          Shortly before the 2019 general elections,[11] the New Nation Movement NPC and its associates launched an urgent application for leave to appeal directly to this Court against a judgment of the Western Cape Division of the High Court, Cape Town.[12]  The High Court had dismissed an application to declare certain provisions of the Electoral Act unconstitutional in so far as they did not allow independent candidates to stand for elections.  This Court held that the application was not urgent and heard it in the ordinary course.[13]

 

[11]          On 11 June 2020, this Court, in a judgment referred to as New Nation II,[14] held that the Electoral Act was inconsistent with the Constitution as it made no provision for independent candidates to contest elections for the National Assembly and Provincial Legislatures.  The Court acknowledged that an electoral system that would accommodate independent candidates had to comply with certain requirements, one of which was the section 46(1)(d) requirement that the electoral system must be one that results, in general, in proportional representation.  The choice of an electoral system which must comply with these objectives and requirements was left for Parliament to make.  The operation of the declaration of unconstitutionality was suspended for 24 months to afford Parliament an opportunity to remedy the defect giving rise to the unconstitutionality.[15]

 

[12]          After the New Nation II judgment, Parliament was obliged to design an electoral system that would accommodate both independent candidates and political parties.  That system still had to result in general, in proportional representation as required by the Constitution.  The revised electoral system effectively had to be in place before the 2024 national and provincial elections.

 

Process leading to the Electoral Amendment Act

[13]          Dr Pakishe Aaron Motsoaledi, the Minister of Home Affairs, filed an answering affidavit in which he explained the process that led to the amendment of the Electoral Act.  He explained that after the New Nation II judgment, he commissioned a panel of experts called the Ministerial Advisory Committee (MAC) to produce technical proposals in response to the judgment.  The MAC had to develop policy options on the electoral system that would cure the defects in the Electoral Act.

 

[14]          After investigation and deliberation on different policy options, the MAC proposed two options.  There was a 3:4 split between committee members as to which option was preferable.  The first, and majority, option resembled the current local government electoral system, in that it proposed combining the first-past-the-post method and proportional representation system, making it a mixed-member proportional system.  This option would have entailed making extensive changes to the Electoral Act.  The second option entailed the modification of the existing multi- member electoral system to accommodate independent candidates in the national and provincial elections without many changes in the legislation, including not interfering with the constitutionally required general proportionality.  Under this option, the 400 seats in the National Assembly would continue to be divided into two: 200 regional seats and 200 compensatory seats.  This option was referred to as the minimalist approach.  The Minister and his team preferred the minimalist approach and decided to adopt it.

 

[15]          Thereafter, the Minister instructed four independent advocates with expertise in constitutional and electoral law to prepare a draft amendment Bill in line with the chosen “minimalist approach”.  In November 2021, a draft Bill was presented to Cabinet where it was accepted for introduction to Parliament.  The Bill was introduced to Parliament and published in the Government Gazette on 31 December 2021 as the Electoral Amendment Bill, B1-2022 (Electoral Amendment Bill).

 

[16]          An answering affidavit filed on behalf of the Speaker of the National Assembly and the Chairperson of the NCOP sets out the process followed by Parliament in amending the Electoral Act.  The answering affidavit was deposed to by Mr Mosa Steve Chabane, the chairperson of the National Assembly’s Portfolio Committee on Home Affairs (Portfolio Committee).

 

[17]          On 1 and 2 March 2022, the Portfolio Committee heard oral submissions from 20 civil society organisations and individuals.  The Portfolio Committee thereafter split into two groups and conducted public hearings across the country from 9 to 23 March 2022.  The Electoral Commission also made extensive submissions to the Portfolio Committee on the Electoral Amendment Bill.

 

[18]          In terms of this Court’s order in New Nation II, Parliament had 24 months to cure the constitutional defects in the Electoral Act as identified in that judgment.  That period was due to expire on 10 June 2022.  On 26 April 2022, Parliament applied for a six- month extension, which was granted, thereby shifting the deadline to 10 December 2022.

 

[19]          Between 10 June and 20 October 2022, the Portfolio Committee discussed and considered various aspects of the Electoral Amendment Bill in no less than 16 meetings held at least once a week, where the Portfolio Committee also received inputs from various stakeholders and parties.

 

[20]          The Electoral Amendment Bill was passed by the National Assembly on 20 October 2022 and was then referred to the NCOP for its consideration.  Before passing the Electoral Amendment Bill, the National Assembly (either through the Minister or the Portfolio Committee) received at least four memoranda or opinions from Counsel on various questions raised by Parliament or in response to public comments on various aspects of the Electoral Amendment Bill.  Counsel’s opinion specifically addressed questions on the 200/200 split that are the subject of consideration in this matter.

 

[21]          The NCOP’s Select Committee on Justice and Security (Select Committee) called for written submissions from the public upon receiving the Electoral Amendment Bill from the National Assembly.  Between 2 November 2022 and 25 November 2022, the Select Committee deliberated on the Electoral Amendment Bill and received inputs from various stakeholders, including the Minister and the Electoral Commission.

 

[22]          Towards the end of November 2022, the Select Committee proposed various amendments to the Electoral Amendment Bill pursuant to the public engagement process.  As a result, the Electoral Amendment Bill was returned to the National Assembly for further public engagement on the NCOP’s proposed amendments as required in terms of standard parliamentary process.

 

[23]          On 29 November 2022, the NCOP amended and passed the Electoral Amendment Bill.  In December 2022, Parliament sought a further extension on an urgent basis to finalise the Electoral Amendment Bill.  On 9 December 2022, this Court made an interim order suspending the declaration of invalidity to 31 January 2023, pending a final determination of Parliament’s further extension application.  The further extension application was granted by this Court on 20 January 2023, thereby shifting the deadline to 28 February 2023.[16]  Following this further extension, the National Assembly passed the Electoral Amendment Bill on 23 February 2023 and referred it to the President for assent.

 

[24]          On 13 April 2023, the President signed the Electoral Amendment Bill and it was published in the Government Gazette on 17 April 2023, resulting in the Electoral Amendment Act.  The impugned item 1 in Schedule 1A as amended by the Electoral Amendment Act provides as follows:

 

“The seats in the National Assembly are as determined in terms of section 46 of the Constitution and item 1 of Schedule 3 and are allocated as follows:

(a)                Half the seats are filled by independent candidates from lists of candidates of parties contesting the nine regions and these shall be referred to as regional seats; and

(b)                half the seats are filled by candidates from lists of candidates of parties and these shall be referred to as compensatory seats.”

 

Election of the National Assembly under the Electoral Act as amended

[25]          The allocation of seats in the National Assembly in terms of the Electoral Amendment Act is still divided into equal halves: the first 200 seats referred to as regional seats and the other 200 as compensatory seats.  Independent candidates and political parties can both contest for the 200 regional seats, but only political parties can contest for the 200 compensatory seats.  The amended Electoral Act therefore retains the two-tier system while providing an opportunity for independent candidates to contest elections for the National Assembly as required by New Nation II.

 

[26]          Whereas previously one ballot determined the allocation of regional and compensatory seats in the National Assembly, voters now receive two ballots – one to vote for a regional seat and one for a compensatory seat.

 

[27]          The allocation of regional seats is determined in terms of items 4 and 5 of Schedule 1A to the amended Electoral Act, in terms of which, as before, the 200 regional seats in the National Assembly are allocated to nine regions, which coincide with the provincial boundaries.  The Electoral Commission determines the number of seats allocated to each region based on the number of registered voters in each region.  Provinces with higher populations of registered voters will thus be allocated more regional seats compared to provinces with smaller populations of registered voters.

 

[28]          A voter will receive a ballot paper specific to their region, which contains both independent candidates and political parties.  For convenience, I shall refer to this as the regional ballot.  The voter can vote for either a political party or an independent candidate on the regional ballot.  All votes cast in a region will first be used to determine a provisional allocation of that region’s regional seats to both independent candidates and political parties.  If an independent candidate meets the relevant quota for one seat, they will be elected to the National Assembly.

 

[29]          The regional quota at the first stage of the computation is the same for both political parties and independent candidates.  The quota of votes needed to obtain a seat in each region is determined using the Droop Formula.  This provides a provisional allocation of seats for each political party and independent candidate.  Any unallocated seats are then allocated using the largest remainder method.^[17]

 

[30]          Even if an independent candidate gets sufficient votes to be provisionally awarded two or more seats, the candidate is only entitled to one seat.  This is because, unlike a political party which has many candidates and may get many seats, that is not possible for an independent candidate as they are contesting the elections in their individual capacities and are not affiliated to any party.  Once the single seats won by each of the successful independent candidates are allocated, a revised Droop quota is calculated to determine the final allocation of regional seats to the political parties, using the remainder of the regional seats and the number of votes obtained by political parties in the regional ballot.

 

[31]          The amended Electoral Act provides that independent candidates forfeit any seat or votes they get over and above their one allocated seat.  Similarly, if an independent candidate contests for more than one region and wins seats in more than one region, they forfeit the additional seat(s) in the other region(s).  The Minister submits that this is the inevitable consequence of the choice made by a citizen to contest elections as an individual instead of doing so through a political party.

 

[32]          Then there are the 200 compensatory seats, the allocation of which is determined in terms of item 6 of Schedule 1A of the amended Electoral Act.  The voter in a region gets a second ballot paper, listing only the contesting political parties.  For convenience, I shall call this the compensatory ballot.  The compensatory ballot does not elect compensatory representatives directly.  The Droop Formula is also used to allocate compensatory seats.  This quota takes into account all 400 seats in the National Assembly minus the seats allocated to independent candidates and is based on the total number of the votes cast for the political parties in both the regional and compensatory ballots.  As before, this exercise will result in a provisional allocation of compensatory seats to a political party, from which are subtracted the regional seats won by that party in order to arrive at the party’s final allocation of compensatory seats.  The stated purpose of the compensatory seats is to correct the disproportionality in representation in the results of the election.

 

[33]          Therefore, the new position after the passing of the Electoral Amendment Act is that 200 regional seats are filled by independent candidates and candidates from political parties, and the 200 compensatory seats are filled by the candidates from the lists of political parties only.  The independent candidates are excluded from contesting for the compensatory seats.

 

[34]          The affidavit deposed to by Mr Michael Louis, the chairman and director of the applicant, states that the applicant does not object to the principle of a two-tier compensatory system as contained in the impugned schedule.  The applicant also does not object to the principle that independent candidates can only compete for regional seats whereas political parties can compete for both the regional and compensatory seats.  However, the applicant takes issue with the fact that independent candidates can only compete for 200 regional seats, and argues that independent candidates will, as a result, be required to earn more votes in order to secure a single seat in the National Assembly compared to political parties.

 

[35]          The contention that candidates for regional seats will have to earn more votes than candidates for compensatory seats is based on an analysis by Mr Michael Atkins (Atkins Report), whose expertise lies in the fields of mathematics and computer science.[18]  Briefly, Mr Atkins’ thesis is that, based on data from the 2014 and 2019 elections, independent candidates will require more votes than political parties to win a seat in the upcoming national elections.

 

[36]          In the 2019 national election,[19] in which 17 437 379 people voted, the weighted average quota for regional seats was 83 511 while the quota for compensatory seats was 43 485.  Before the Electoral Amendment Act, this difference in quota was non- discriminatory, because there was only a single ballot and only political parties contested elections.  The formulas in the impugned schedule for determining the allocation of regional and compensatory seats were the same for all political parties.

 

[37]          According to Mr Atkins’ analysis, things have changed with the introduction of two ballots in order to accommodate independent candidates in the regional ballot.  Because independent candidates can only contest for regional seats, they would – based on the 2019 data – require 83 511 votes to win a seat.  The new method for calculating the compensatory quota still uses, as it did before, the total number of seats in the National Assembly as the starting point for the denominator in the Droop Formula (400 plus one), save that one must deduct the number of seats won by independent candidates from the denominator.  However, the number used as the numerator is now the total votes cast for political parties in both the regional and compensatory ballots.  So, if all voters voted only for political parties, the numerator – based on the 2019 data – would be 34 874 758 (that is, double the number of single votes cast in the 2019 election).

 

[38]          The compensatory quota will thus be significantly higher than it was in the 2019 election.  Indeed, if no independent candidates obtained votes, the compensatory quota would be very close to the regional quota.  But, says Mr Atkins, the difference is that in relation to the compensatory seats, political parties get the benefit of votes from both the regional and compensatory ballots.  Mr Atkins makes the fundamental assumption that persons who vote for political parties in the regional ballot will vote for the same political parties in the compensatory ballot.  To illustrate what he contends to be a discriminatory effect against independent candidates, he assumes total voter support for independent candidates at two alternative levels, 2% and 8%.

 

[39]          At these assumed levels, and based on the 2019 data, the compensatory quota for a compensatory seat would be 86 532 and 85 846, respectively.  However, on the assumption that voters will vote for the same political parties in both ballots, these figures suggest that a party will need the support of only half this number of voters in order to win a compensatory seat.  After certain minor adjustments, the figure of actual voter support at which Mr Atkins arrives for a political party to win a compensatory seat is 43 703 or 44 712, which is only just over half the voter support which a contestant for a regional seat (including an independent candidate) would need to garner.

 

[40]          After running multiple scenarios on varying numbers of regional and compensatory seats, Mr Atkins concludes that an allocation of 350/50 as between regional and compensatory seats will result in an approximation of the actual voter support required for regional and compensatory seats, respectively, while preserving proportional representation for political parties and posing minimal risk of a so-called overhang.  I shall later explain the concept of overhang.

 

[41]          The applicant therefore seeks an order in terms of which the compensatory seats would be reduced from 200 to 50.  As a result, those competing in the regional tier will contest for 350 seats instead of 200 seats.  This will result in a 350/50 split as opposed to the 200/200 split contemplated in the Electoral Amendment Act.

 

Issues

[42]          The issues to be determined by this Court are as follows:

(a)               Should direct access be granted?

(b)              Is the 200/200 split rationally connected to a legitimate governmental purpose?

(c)               Does the 200/200 split give rise to an unjustifiable limitation of fundamental rights in the Constitution, in particular, the right to vote and to stand for public office?

(d)              If so, is the limitation justifiable?

(e)               If not, what is an appropriate remedy?

 

Direct access

[43]          Section 167(6)(a) of the Constitution, read with rule 18 of the Rules of this Court, provides for direct access to the Constitutional Court.  In terms of these provisions, this Court has discretion whether to grant direct access, but such discretion will only be exercised in favour of the applicant if it is in the interests of justice to grant direct access.  As stated in Zondi,[20] whether it is in the interests of justice to grant leave for direct access depends on various factors, including:

 

“[T]he importance of the constitutional issue raised and the desirability of obtaining an urgent ruling of this Court on that issue, whether any dispute of fact may arise in the case, the possibility of obtaining relief in another court, and time and costs that may be saved by coming directly to this Court.”[21]

 

[44]          For direct access to be granted, an applicant must show that exceptional circumstances exist.  This Court is reluctant to be a court of first and last instance as this deprives parties of the right of appeal and deprives the Court of the benefit of other courts’ insights.[22]

 

[45]          The applicant submits that it would be in the interests of justice for this Court to grant direct access in that: (a) the application raises legal issues of fundamental importance relating to the fairness of the 2024 elections; (b) the Court has to determine the validity and constitutionality of the impugned schedule; (c) the challenge has reasonable prospects of success; and (d) the 2024 elections are imminent, and there is thus insufficient time to approach the High Court for the relief sought and thereafter seek confirmation of the order of invalidity in this Court as required by section 167(5) of the Constitution.

 

[46]          The respondents do not oppose the application for direct access.  They accept that time is of essence and that it is imperative that the issues raised in the application be determined as a matter of urgency to allow the Electoral Commission sufficient time to prepare for the 2024 elections.

 

[47]          I agree with the parties that the circumstances of this case are exceptional and that the issues raised are purely legal in nature.  This application concerns constitutional issues, in particular, whether the impugned schedule violates political rights enshrined in section 19 of the Constitution.  This matter also raises questions of separation of powers, in particular, the extent to which courts can interfere with the electoral process chosen by Parliament.

 

[48]          Further, having regard to the explanatory affidavit filed on behalf of the Electoral Commission, there will be insufficient time for the matter to be considered by the High Court and then brought to this Court for final consideration, either on appeal or confirmation.  Finality is required to enable the Electoral Commission to proceed with the preparations for the 2024 general election, which must be held on a date to be proclaimed by the President of the Republic of South Africa between 22 May 2024 and 14 August 2024.[23]  Therefore, it is in the interests of justice that direct access be granted.

 

[49]          I now proceed to consider the merits of this application.

 

Is the 200/200 split rationally connected to a legitimate government purpose?

[50]          According to the applicant, the 200/200 split is irrational and inconsistent with the rule of law and, therefore, contravenes section 1(c) of the Constitution.^[24]  The irrationality does not stem from the fact that independent candidates may only compete for regional seats but, so the applicant argues, from the quota for regional seats being much lower than the quota for compensatory seats.  As a result, the applicant submits that the equality of the vote and equal opportunity to be elected are violated.  On the applicant’s version, there is no justification or reason for the split other than for the improper purpose of undermining the prospects of independent candidates getting elected.

 

[51]          In determining this question, it is apposite to first remind ourselves of the legal principles relating to the powers of Parliament and the test for rationality.

 

[52]          It is trite that in order to pass constitutional muster, the exercise of public power by the Executive and other public functionaries must, at the very least, comply with the threshold of rationality.  Decisions must be rationally related to the purpose for which they are given.  This is an objective enquiry.  A court is not allowed to substitute its own opinion as to what it thinks is appropriate.

 

[53]          In New National Party,[25] this Court had to consider a constitutional challenge in respect of certain provisions of the Electoral Act that prescribed that only bar-coded identity documents could be used for the purposes of voting and registering for the 1999 national and provincial elections.  Yacoob J, writing for the majority, said the following on the role of Parliament and onus of proof where an infringement of a right has been alleged:

 

“It is to be emphasised that it is for Parliament to determine the means by which voters must identify themselves.  This is not the function of a court.  But this does not mean that Parliament is at large in determining the way in which the electoral scheme is to be structured.  There are important safeguards aimed at ensuring appropriate protection for citizens who desire this foundational right.  The first of the constitutional constraints placed upon Parliament is that there must be a rational relationship between the scheme which it adopts and the achievement of a legitimate governmental purpose.  Parliament cannot act capriciously or arbitrarily.  The absence of a rational connection will result in the measure being unconstitutional.  An objector who challenges the electoral scheme on these grounds bears the onus of establishing the absence of a legitimate government purpose, or the absence of a rational relationship between the measure and that purpose.

A second constraint is that the electoral scheme must not infringe any of the fundamental rights enshrined in chapter 2 of the Constitution.  The onus is once again on the party who alleges an infringement of the right to establish it.”[26]

 

[54]          Chaskalson P reiterated the rationality requirement in Pharmaceutical Manufacturers[27] and held that:

 

“It is a requirement of the rule of law that the exercise of public power by the Executive and other functionaries should not be arbitrary.  Decisions must be rationally related to the purpose for which the power is given, otherwise they are in effect arbitrary and inconsistent with this requirement.

            . . .

The question whether a decision is rationally related to the purpose for which the power was given calls for an objective enquiry.

. . .

What the Constitution requires is that public power vested in the Executive and other functionaries must be exercised in an objectively rational manner.

            . . .

As long as the purpose sought to be achieved by the exercise of public power is within the authority of the functionary, and as long as the functionary’s decision, viewed objectively, is rational, a Court cannot interfere with the decision simply because it disagrees with it, or considers that the power was exercised inappropriately.”[28]

 

[55]          The wide remit given by the Constitution to Parliament to enact an electoral system was restated in New National Party as follows:

 

“The right to vote contemplated by section 19(3) is therefore a right to vote in free and fair elections in terms of an electoral system prescribed by national legislation which complies with the aforementioned requirements laid down by the Constitution.  The details of the system are left to Parliament.” [29]

 

[56]          In AParty,[30] Ngcobo J reiterated the statement in New National Party and said:

 

“Parliament has the constitutional authority and duty to design an electoral scheme to regulate the exercise of the right to vote.  This is apparent from sections 46(1), 105(1), and 157(5) of the Constitution.” [31]

 

[57]          This principle was again restated in New Nation II, where Madlanga J held:

 

“The pros and cons of this or the other system are best left to Parliament which in terms of sections 46(1)(a) and 105(1)(a) of the Constitution has the mandate to prescribe an electoral system.”[32]

 

[58]          In Albutt,[33] Ngcobo CJ said:

 

“The executive has a wide discretion in selecting the means to achieve its constitutionally permissible objectives.  Courts may not interfere with the means selected simply because they do not like them or because there are other more appropriate means that could be selected. . . .  What must be stressed is that the purpose of the enquiry is to determine not whether there are other means that could have been used, but whether the means selected are rationally related to the objective sought to be achieved.”[34]

 

[59]          Therefore, as long as the purpose sought is within the authority of the functionary and as long as the functionary’s decision is rationally connected to that purpose, a court cannot interfere with the decision simply because it disagrees with it.

 

[60]          In the matter before us, the applicant contends that the 200/200 split means that the quota of votes a political party must obtain to be allocated compensatory seats is much lower than the quota of votes the independent candidate must obtain to be allocated regional seats, and this infringes on the equality of vote and equal opportunity.  In this regard, the applicant, as stated earlier, relies on the report produced by Mr Atkins.  On Mr Atkins’ model, the quota for regional seats is much higher than the quota for compensatory seats – almost double.[35]

 

[61]          The applicant argues that the 200/200 split as introduced in the Electoral Amendment Act is arbitrary and therefore substantively and procedurally irrational in that it fails to meaningfully give effect to section 19(3) of the Constitution and is not reasonably capable of attaining the desired outcome.  The applicant adds that although the inadequacies in the 200/200 split were highlighted during the public participation processes, they were not addressed by the Parliamentary Committee.

 

[62]          The respondents submit that the Electoral Amendment Act is rational in that the two different methods used by the Electoral Amendment Act to fill regional and compensatory seats legitimately differentiate between independent candidates and political parties for the express purpose of achieving proportional representation and avoiding the risk of overhang.  Overhang occurs where the election formula requires political parties to be allocated more seats than are actually available in the Legislature.  Put differently, overhang occurs where a party wins more regional seats than the compensatory or national party vote entitles it to.  This, in turn, has implications for proportional representation.  In a Legislature of 400, overhang of one would indicate that, after applying all the relevant formulas in the legislation, 401 seats have to be allocated.  The rationality of the 200/200 split therefore hinges on whether (a) it results in proportional representation, in general and (b) avoids the risk of overhang.

 

Proportional representation in general

[63]          Section 46(1)(d) of the Constitution requires Parliament to design an electoral system that results, “in general, in proportional representation”.  There are 400 seats available for contestation in the National Assembly which, in terms of the impugned schedule, are divided in half: 200 regional seats and 200 compensatory seats.  The stated purpose of the 200 compensatory seats, according to Parliament, is to correct the disproportionality in representation in the results of the election.

 

[64]          Parliament contends that proportional representation systems are, in general, subject to distortions with the introduction of independent candidates.  This is so because irrespective of how many votes an independent candidate receives, they are restricted to a maximum of one seat.  By contrast, political parties can hold the number of seats proportional to the number of votes received.

 

[65]          According to Parliament and the Minister, the only way to include independent candidates in a two-tier system is to allow them to retain the seats they win and obtain proportionality by allocating the remaining seats to the parties that gain representation.  As a result, independent candidates cannot compete for compensatory seats as they are reserved exclusively for political parties.

 

[66]          The Lynge and Rosen Report,[36] filed on behalf of the Electoral Commission, concludes that the 200/200 split achieves proportionality.  The applicant accepts that the 200/200 split results, in general, in proportional representation.  This is also accepted in the Atkins Report filed in support of the applicant’s case.  The respondents accept that a 350/50 split would also achieve proportionality in general.  The applicant also acknowledges that reserving compensatory seats for political parties is necessary to achieve proportionality because the formulas for regional seats have the potential to distort proportionality, and this potential has increased because independent candidates can now contest for regional seats but can only hold one seat.

 

[67]          Properly construed, the dispute between the parties is, therefore, not whether it is irrational for Parliament to distinguish between regional and compensatory seats in the National Assembly.  Rather, the dispute concerns the number of compensatory seats that ought to be reserved for political parties, with the applicant proposing a 350/50 split.  The applicant argues that the 200/200 split debases the value of votes for regional seats.  In comparison, the 350/50 split will make the quota for regional and compensatory seats similar while still retaining overall proportionality and without giving rise to a significant risk of overhang.  Because both the 200/200 and 350/50 split results in proportional representation, the question is whether avoiding the risk of overhang justifies the disparity in votes needed to obtain a compensatory and regional seat, respectively.  The crucial consideration is whether the 200/200 does, in fact, reduce the value of votes in the regional ballot.  I deal with this point later in determining whether the 200/200 split amounts to an unjustifiable limitation of sections 1(c); 3(2)(a); 9(1); 19(2); 19(3) and 46(1)(d) of the Constitution.

 

[68]          Parliament’s stated objective for the 200/200 split is the achievement of proportional representation as required by the Constitution.  Even if the 350/50 split proposed by the applicant might arguably be fairer and achieve proportionality, an assessment as to whether the 200/200 split achieves proportional representation in general is one that should be conducted by Parliament.  The case law referred to above demonstrates the wide latitude given to Parliament to consider the manner in which to conduct the electoral system.  Sections 46(1) and 105(1) expressly leave the choice of electoral system in Parliament’s hands.  Although the Constitution does set out requirements, there are a number of possible electoral systems which would comply with these requirements.  At least on the requirement of achieving proportionality, the 200/200 split chosen by Parliament passes constitutional muster.

 

Overhang

[69]          The second stated objective of the 200/200 split by Parliament is that it avoids the risk of overhang.  The Electoral Commission explains that overhang is when more seats are required to be allocated to restore proportionality as between represented parties after the allocation of regional (or constituency) votes, than are available in the Legislature.  Overhang occurs when a party wins more regional seats than it is overall entitled to when the compensatory tier outcome is calculated.  This can happen with regionally based parties with strong support but much less support in the rest of the country.  Another possibility is when a small party wins a regional seat based on the largest remainder method but does not meet the threshold for a seat when the compensatory calculations are done.  The consequence of such situations is that seats for parties cannot be allocated in the National Assembly according to their inter-party proportional entitlement.

 

[70]          Section 46(1) of the Constitution limits the number of seats in the National Assembly to 400.  The implication of overhang in our system is that more seats are required to be allocated to ensure proportional representation as between the represented parties after the allocation of regional votes than are available in the Legislature.  Put differently, when overhang occurs, seats in the National Assembly cannot be allocated according to inter-party proportional entitlement and, as a result, the Electoral Commission may not be able to declare the results of an election.

 

[71]          Although the applicant accepts that the 200/200 split avoids the risk of overhang, it contends that the 350/50 split will bring the quota for a regional seat closer to the quota for a compensatory seat without posing a significant risk of overhang or disturbing the proportional representation of political parties.  To verify that the 350/50 split sufficiently minimises the risk of overhang, Mr Atkins ran a Monte Carlo simulation^[37] of elections in terms of the Electoral Amendment Act based on the 2014 and 2019 election results.  The simulation was repeated with 200, 100, 75, 50, 25 and 10 compensatory seats for comparison and with assumed levels of support for independent candidates ranging from 1% to 10% and assessed the risk of overhang with reference to “main parties”,^[38] “minor parties”^[39] and the “largest party”.^[40]  Mr Atkins’ tables show that the risk of overhang almost always reduces as support for independent candidates increases, so the greatest risk is where independent candidates win only 1% of voter support.

 

[72]          According to the Atkins Report, the 350/50 split proposed by the applicant gives rise to a foreseeable risk of overhang of up to 15% per election (based on the 2014 election data), limited to one seat.[41]  The applicant argues that, while foreseeable, this risk under the 350/50 split is remote and limited to one seat.  On the 200/200 split, however, there is virtually no risk of overhang, with the only recorded risk found in the Atkins Report being at 0,03% at 6% support for independent candidates in respect of one seat, based on the 2014 election data.

 

[73]          With each simulation of elections for regional seats, the 2014 and 2019 election results were varied in terms of the number of independent candidates contesting each regional election, as well as the level of support for independent candidates at 1%; 2%; 4%; 6%; 8% and 10%.  The number of votes for independent candidates were added to the total votes for political parties.  As a result, the relative reduction in support for political parties was kept constant across parties.

 

[74]          Once the regional elections were simulated, the votes for political parties were added together and used to simulate the overall proportional representation of political parties for the National Assembly.  The number of seats available to political parties is 400 minus the number of seats won by independent candidates.  The simulation relies on two assumptions.  First, a vote for a political party in the regional ballot will translate to a vote for the same political party on the compensatory ballot.  Second, voters that supported an independent candidate in the regional ballot will not vote for one of the three largest political parties but instead vote for one of the smaller parties.  The basis of these assumptions does not appear from the Atkins Report.  Further, the Atkins Report does not tell us how the risk of overhang would be affected if these assumptions were changed.

 

[75]          Based on the simulation, the 350/50 split gives rise to a foreseeable risk of overhang.  The risk of overhang of one seat is at its greatest where independent candidates obtain 1% support.  The risk of overhang reduces as support for independent candidates’ increases.  On the 2019 election data, and based on 1% support for independent candidates, the risk of overhang of one seat, while being 0% for the largest party, is about 2% for the main parties and about 7% for minor parties.  On the 2014 election data, and while the risk for the largest party remains at 0%, the risk for main parties reduces to just under 1% but increases to about 15% for minor parties.

 

[76]          The Atkins Report further shows that parties with strong regional support have a noticeable risk of a single-seat overhang.  The implication is that political parties with strong regional support may be awarded a seat based on the regional ballot while not being awarded a seat in the National Assembly based on the final seat allocation for political parties.  In light hereof, the applicant argues that while there is virtually no risk of overhang on the 200/200 split,^[42] the risk of overhang under the 350/50 split is remote and limited to one seat.  The applicant also states that, in any event, the Lynge and Rosen Report which the Electoral Commission relies on accepts that a one seat deviation is within the bounds of proportional representation.

 

[77]          In considering the question of overhang, the Electoral Commission filed an explanatory affidavit to respond to the Atkins Report.  Mr Phatudi Simon Mamabolo, the Chief Electoral Officer of the Electoral Commission, explained that there is no mechanism under the current Electoral Act to correct the problem should overhang occur.  Secondly, the applicant has not presented any viable solution to combat the risk of overhang, or remedy it should overhang occur.  Furthermore, overhang poses the risk that the Electoral Commission would not be able to declare the election should the risk materialise and any declaration of the election results where overhang has occurred would likely be challenged in court.

 

[78]          The Electoral Commission states that the fact that a one seat deviation still complies with proportional representation is irrelevant.  The Lynge and Rosen Report did not consider the risk to proportional representation should overhang occur.  Rather, Lynge and Rosen ran a Monte Carlo simulation to determine whether the 200/200 split under the amended electoral formula would introduce additional disproportionality in comparison to the electoral formula prior to the amendment.  They concluded that it would not – the deviation from absolute proportionality never exceeds one seat.

 

[79]          Parliament and the Minister submit that, even if the risk is as remote as proffered by the applicant, the consequences are severe as explained by the Electoral Commission.  To place the risk of overhang in perspective, the Minister refers to the election of the German Federal Parliament, the Bundestag, where overhang occurred and 138 seats had to be awarded to restore proportionality.

 

[80]          I agree with the respondents that, if overhang were to occur in South Africa under the current construction, there would be no option to add additional seats to the National Assembly to restore proportional representation, as was done in Germany.  Germany has a comparable population size to that of South Africa.  The Bundestag is comparable to South Africa’s National Assembly and aims to achieve proportional representation as best as it can.  Similar to our two-tier system, German voters cast two ballots – one for a candidate to represent their single member constituency and one for a political party.  Unlike South Africa’s National Assembly, the German Bundestag is not limited to a specific number of seats.  Thus, when overhang occurs, extra seats are made available to ensure proportional representation.  In the 2021 Bundestag election, which produced overhang, a total of 138 extra seats had to be awarded to restore proportionality.  To put things into perspective, South Africa can only allocate a maximum of 400 seats in the National Assembly.  At 736 members, the Bundestag is now the largest democratically elected parliament.^[43]

 

[81]          We do not have the luxury of pursuing the path taken by the German government — merely increasing the seats when the need arose.[44]  The Constitution of South Africa is clear in this regard and capped the maximum number of seats at 400.  Any increase beyond the 400 seats available for contestation would require a constitutional amendment and this is something the applicant has failed to take into account.

 

[82]          Besides showing that the risk of overhang is remote on the 350/50 split, the applicant does not offer a single solution in its pleaded case to combat the risk or how it should be dealt with should the risk materialise.  The fact that the risk is limited to one seat is immaterial – the issue is that there is a risk that, if it materialises, would have dire consequences.  The Electoral Act does not provide a mechanism that allows the Electoral Commission to correct the problem should overhang occur.  Furthermore, the Electoral Commission has explained that, if overhang occurs, it will not be able to declare that election.

 

[83]          Considering the above, on the 200/200 split there is virtually no risk of overhang.  As a result, Parliament’s second stated objective of the 200/200 spilt – to avoid the risk of overhang – is achieved.  On the other hand, the foreseeability of overhang under the 350/50 split as proposed by the applicant is potentially destructive to the applicant’s case.  So too is the acknowledgement that the 200/200 split serves a legitimate governmental purpose – that is, the split ensures proportionality in results.  Consequently, and as already demonstrated, the 200/200 split passes the rationality test and cannot be said to be arbitrary.

 

Possible solutions to the risk of overhang vis-à-vis the applicant’s pleaded case

[84]          Before concluding on the issue of overhang, I find myself constrained to address the applicant’s submission that the method utilised at the municipal level to prevent overhang offers a potential solution to the risk of overhang in the 350/50 split.  This submission was made for the first time during the hearing.

 

[85]          It is well-established that parties in motion proceedings should make out their case in their founding or answering affidavits and not in their heads of argument or in the course of the hearing.  This is because the right to a fair trial requires that parties know the case they have to meet, and also ensures that the issues that the courts have to determine are properly placed before them.[45]  As such, and in deciding this matter, this Court cannot take into consideration the submission by the applicant on a potential solution to the risk of overhang as it is not properly before this Court.  The exception to this position was articulated in My Vote Counts[46] where Khampepe J said the following:

 

“It is, in any event, imperative that a litigant should make out its case in its founding affidavit, and certainly not belatedly in argument.  The exception, of course, is that a point that has not been raised in the affidavits may only be argued or determined by a court if it is legal in nature, foreshadowed in the pleaded case and does not cause prejudice to the other party.”[47]  (Emphasis added.)

 

[86]          In the present case, this exception does not apply.  The proposed solution to the risk of overhang which the applicant proffered in the hearing is not legal in nature nor was it foreshadowed in the applicant’s pleaded case.  The applicant did not offer any solution to the risk of overhang in its pleaded case.  Had the issue been properly raised by the applicant in its founding affidavit, the respondents may well have put up a case against this proposed solution.  The respondents may have offered reasons why the solution to overhang at municipal level could not work at national level.  If this Court were to consider the proposed solution to the risk of overhang as proffered by the applicant in the hearing, that would certainly be prejudicial the respondents as they would not have had the opportunity to deal with the issue.

 

[87]          While I do not attach weight to the applicant’s proposed solution to the risk of overhang, I wish to make the following remarks: although the 200/200 split is rational in that it avoids the risk of overhang and ensures proportional representation, this should not be interpreted as an indication that overhang will always be an insurmountable challenge.  The risk of overhang in the Atkins Report on the proposed 350/50 split, as well as on the 300/100 or 325/75 split, arises for two reasons.

 

[88]          First, like the Electoral Act, the Electoral Amendment Act prescribes that the full 400 seats of the National Assembly must be made available for contestation.  This is the maximum number of seats that the Constitution prescribes.  On the proposed 350/50 split, where there is a foreseeable risk of overhang, it is impossible to increase the number of seats in the National Assembly by the number of seats required to cure the overhang.

 

[89]          The election of the New Zealand Parliament offers another example of how overhang can be resolved, one which does not increase the size of Parliament to the extent of the German Bundestag.  In New Zealand, the Electoral Act[48] introduced a mixed-member proportional system to elect Members of Parliament, which ordinarily comprises of 120 seats.  Section 192(5) of the New Zealand Electoral Act provides that if overhang seats are won, the number of seats in Parliament is increased by the number of overhang seats until the next election.  For example, there were 121 members after the 2005 and 2011 elections, and 122 following the 2008 election.[49]  Of course, our Constitution does not allow the National Assembly to have more than 400 seats.

 

[90]          However, should the number of seats available for contestation in Parliament be reduced to a number above 350 but below 400,[50] it may be possible to allocate “overhang seats” and increase the number of seats in the National Assembly by the number of overhang seats (up to 400) until the next election.  By way of illustration, if Parliament were to prescribe that 390 seats (of the 400) were available for contestation in the National Assembly, and for the relevant election year it happens that there is overhang of two seats, following the New Zealand model, the Electoral Commission would be able to allocate the two overhang seats until the next election.  This would mean an allocation of 392 seats in the National Assembly until the next election (that is, the 390 available for contestation plus the two overhang seats), which is still below the 400 constitutional cap.

 

[91]          Secondly, unlike the Local Government: Municipal Structures Act,[51] the Electoral Amendment Act does not provide a formula to address overhang should it arise.  In local government elections for a metro council or a local council with wards, each voter has two votes: one for the direct election of a ward candidate (who may be an independent candidate or a political party representative) and one for a political party to ensure that political parties are proportionally represented in the council.[52]  Ward candidates and political parties contest elections for the same number of seats.[53]  Item 16 of Schedule 1 to the Municipal Structures Act provides:

 

“(1)      If, through the election of ward candidates, a party listed on the part of the ballot paper for parties has obtained a number of seats that is equal to, or greater than the total number of seats in the council to which it is entitled under item 13, that party must not be allocated any seats from its list of party candidates.

(2)        The seats of ward candidates are not affected.

(3)        A new quota for a seat must be determined in accordance with the following formula (fractions to be disregarded)

            . . .

(4)       (a)        The total number of valid votes cast for each party, both on the party vote and for the ward candidates representing the party, excluding the party that has excessive number of seats, must be divided by the quota of votes for a seat.  The result is the total number of seats to which each party is entitled.

            (b)        If the calculation in paragraph (a) yields a surplus not absorbed by the seats awarded to a party, that surplus must compete with similar surpluses accruing to any other party or parties and any undistributed seat or seats must be awarded to the party or parties concerned in sequence of the highest surplus.

            (c)        If the surplus for two or more parties is equal, the seat must be awarded to the party that received the highest number of valid votes”.[54]

 

[92]          I recognise that the absence of a formula to address overhang in the Electoral Act as amended may well be a result of the fact that there is no risk of overhang on the 200/200 split.  I refer to the Municipal Structures Act to illustrate that the risk of overhang is not a novel concept in South Africa and that the Legislature has, in the context of local government elections, provided a mechanism to deal with it should it arise.

 

[93]          On the facts and submissions made by the parties in this case, I conclude that the 200/200 split passes the rationality test as it achieves proportional representation and avoids the risk of overhang, as intended.  The next question is whether the electoral scheme has infringed any fundamental rights in the Constitution.

 

Does the 200/200 split give rise to an unjustifiable limitation of fundamental rights in the Constitution?

[94]          The applicant contends that the 200/200 split as contained in the Electoral Amendment Act is unconstitutional as it violates the sections 1(c), 3(2)(a), 9(1), 19(2), 19(3) and 46(1)(d) of the Constitution.

 

[95]          In New National Party, where an infringement of voters’ rights was alleged, this Court held that the onus is on the party who alleges an infringement of the right to establish it.[55]  The applicant therefore bears the onus to establish the rights violations so alleged.  I now proceed to consider the issues raised by the applicant.

 

Sections 1(c) and 46(1)(d) of the Constitution

[96]          The applicant submits that the section 1(c) provision is infringed upon in that the split is irrational and inconsistent with the rule of law.  The applicant also alleges in its founding affidavit that the split violates section 46(1)(d)[56] of the Constitution in that the election of the National Assembly in terms of the Electoral Amendment Act would not result, in general, in proportional representation.  This is because, so argues the applicant, there are bound to be wasted votes and excess votes for independent candidates which will be allocated to political parties, thereby skewing the outcome of the elections.  The applicant did not persist with this argument in the hearing.

 

[97]          As already shown, the applicant acknowledges that there is a legitimate governmental purpose served by the 200/200 split – the split ensures proportionality in results.  The split also avoids the risk of overhang.  Consequently, and as already demonstrated, the 200/200 split passes the rationality test.  The section 1(c) challenge is, therefore, without merit.

 

[98]          Similarly, the applicant’s section 46(1)(d) challenge to the 200/200 split as alleged in its founding affidavit should fail.  During the hearing, counsel for the applicant accepted that the 200/200 split achieves proportional representation in general as contemplated in section 46(1)(d).  This was also accepted in the Atkins Report.

 

Sections 3(2)(a) and 9(1) of the Constitution

[99]          The applicant argues that the 200/200 split violates section 3(2)(a)[57] of the Constitution in that a vote for an independent candidate carries less weight than a vote for a political party, and that this is inconsistent with citizens voting equally for independent candidates and independents running for public office.  This is because, argues the applicant, the compensatory quota requires only half the number of voters to support a political party than is required by an independent candidate.  In reaching this conclusion, the Atkins Report rests on the critical assumption that voters who vote for a political party in the regional ballot will vote for the same political party in the compensatory ballot.  Consequently, a vote for a political party will count twice when compared to a vote for an independent candidate.

 

[100]     The applicant argues further that the split violates section 9(1)[58] of the Constitution in that it arbitrarily differentiates between independent candidates and political parties.

 

[101]     In response, Parliament submits that to the extent there is a rational basis for the 200/200 split, there can be no infringement of sections 1(c), 3(2)(a) or 9(1) of the Constitution.  For this proposition, Parliament submits that this Court has made it clear that there can be no infringement of the right to equality even where there is a differentiation between different categories of persons, provided that there is a rational connection between the measures taken and the legitimate government purpose of facilitating the effective exercise of the right to vote.

 

[102]     Parliament argues that, on the evidence, the 200/200 split does not infringe the right to equality as protected in the Constitution as it is rationally connected to a legitimate government purpose in that the split (a) ensures a system that results in proportional representation in general, and (b) avoids the possible risk of overhang which threatens constitutional imperatives.

 

[103]     In response to the applicant’s sections 3(2)(a) and 9(1) challenge, the Minister argues that the applicant’s contention that a vote for an independent candidate carries less weight for no reason is fundamentally flawed.  The Minister submits that independent candidates compete for the same quota as political parties in respect of regional seats, and they do so before the same voters and on the basis of the same ballot paper.

 

[104]     With regard to the compensatory seats, the Minister argues that the differentiation between political parties and independent candidates is not a result of a devaluation of the votes for independent candidates as suggested by the applicant.  Rather, it is because an independent candidate can only hold a single seat regardless of the percentage of the vote they get and this distorts proportionality.  By contrast, political parties can fill seats in proportion to the percentage of votes they get, and the compensatory seats can thus be used to restore proportionality.  The Minister adds that just as like cases should be treated alike, unlike cases should be treated differently.  This is a general axiom of rational behaviour.

 

[105]     In its explanatory affidavit, the Electoral Commission submits that the assertion made in the Atkins Report that the number of votes an independent candidate or political party must obtain in order to gain a seat in the regional tier should be equivalent to the quota for a compensatory seat is logically flawed.

 

[106]     This submission in the Atkins Report, explains the Electoral Commission, ignores the fact that South Africa has a two-tier electoral system for the national election, and as such, different quotas apply to the regional tier and the compensatory tier.  The two tiers are inherently different and serve different purposes: they are not the same in size of the constituency; the voter base or voters’ roll; and the seats in contention.  Consequently, the basic premise underpinning the applicant’s criticism that the number of votes per seat in the regional tier must be equal or approximately equivalent to the number of votes required per seat in the compensatory tier is flawed.  At the regional tier, where independent candidates compete with political parties, the quota per seat, as already stated, is exactly the same.

 

[107]     The Electoral Commission further explains that, while the quota for a regional seat is based on one ballot, the compensatory quota is based on the aggregate of two ballots.  The use of two ballots increases the quota for the allocation of compensatory seats to be broadly similar to those that apply in the regional tier.  On projecting the 2019 election results onto the new electoral system, the Electoral Commission found that the compensatory quota tier would likely be higher than it would be in six of the nine regions.

 

[108]     The Electoral Commission therefore concludes that the argument that independent candidates require approximately double the number of votes to gain a seat in the National Assembly, when compared to political parties, is misguided.  The Electoral Commission submits that this distorts the true calculation required for political parties to gain a seat in the National Assembly.

 

[109]     The Atkins Report and the applicant accept that nominally the quota for a compensatory seat is higher than the quota for a regional seat.  However, and on the basis of the assumption that a vote for a political party counts twice, the Atkins Report concludes that political parties require half the votes needed by an independent candidate to gain a seat in the National Assembly.  Based on the 2019 election results, the Atkins Report finds that “the likely threshold for [political] parties to gain a seat in the National Assembly is support from about 44,000 voters” whereas independent candidates would need support from about 85,000 voters.  In its explanatory affidavit, the Electoral Commission aptly points out that, in concluding that independent candidates need double the votes required by political parties to gain a seat in the National Assembly, the Atkins Report assumes, without any basis, that voters will vote for the same political party on both ballots.

 

[110]     In New National Party, the applicant in that case argued that the requirement of a bar-coded identity document excluded eligible voters who did not have bar-coded identity documents and that this breached section 9(1) and 9(2) of the Constitution.  In considering this argument, Yacoob J said the following:

 

“Before this Court, the appellant advanced an argument based on what was alleged to be a breach of sections 9(1) and 9(2) of the Constitution.  However, it is clear from what has been said in this judgment that although the documentary requirements in issue may be said to differentiate between different categories of people, there is a rational connection between the measure and the legitimate governmental purpose of facilitating the effective exercise of the important right to vote.  No discrimination or unfairness has been established.”[59]  (Emphasis added.)

 

[111]     In Democratic Party,[60] this Court heard a similar application to that in New National Party on appeal from the then Transvaal High Court.  The relief sought was an order that the bar-coded identity document requirement for registering and voting in the 1999 elections was unconstitutional.  The applicant argued that although facially neutral, the documentary requirement for voting constituted indirect discrimination against discrete vulnerable groups.[61]  To support this contention, the applicant had relied on surveys conducted by the Human Sciences Research Council (HSRC) and from Opinion 1999: Voter Participation in the 1999 Elections.[62]  In assessing the evidence provided by the applicant in support of its contention that the documentary requirement infringed on the right to equality, Goldstone J for the majority said the following:

 

“On the assumption that the opinions expressed in the HSRC and Opinion 99 reports are correct, there is no evidence as to which category of persons referred to therein might be among the millions of South Africans who, after the promulgation of the Electoral Act, applied for and were issued with the necessary documents, and as a result were able to register on the national common voters’ roll.  In the absence of evidence showing that the impugned provisions have had the effect suggested by the DP, it cannot be found that the provisions, on that account, were unconstitutional.”[63]  (Emphasis added.)

 

[112]     This Court, therefore, rejected the section 9 equality challenge in Democratic Party and in New National Party on the basis that the applicant failed to provide evidence to support the alleged inequity.  This Court in Democratic Party continued and said the following:

 

“[I]n any event, it must be accepted that there are very few laws of general application that will not, directly or indirectly, have the potential to affect different categories of people in different ways, whether, for example, by reason of where they live, their standard of literacy or political beliefs.  There is no evidence to show what the impact of the Electoral Act has in fact been on the various categories of persons referred to by the [Democratic Party].  Whatever the different impact, if any, might be, it is not possible to determine whether such impact constitutes unfair discrimination within the principles endorsed by this Court, unless it is established that such different impact is caused by the impugned legislation, and is not the result of some other cause.”[64]

 

[113]     From this Court’s judgments referred to in the preceding paragraphs, it is clear that the fact that a law affects different categories of people differently does not prove a violation of the right to equality as provided in section 9 of the Constitution.  The person alleging the violation should provide evidence to that effect.

 

[114]     This principle was reiterated by this Court in United Democratic Movement,[65] where the legislation being challenged made it possible for a member of the National or Provincial Legislature or a municipal council to leave the party to which she or he belonged without at the same time losing her or his seat (so-called ‘floor crossing’).  The legislation provided that, except for the initial period, at least 10% of a party had to leave that party for the legislation to apply.  In assessing the challenge, this Court held:

 

“The fact that a particular system operates to the disadvantage of particular parties does not mean that it is unconstitutional.  For instance, the introduction of a constituency- based system of elections may operate to the prejudice of smaller parties, yet it could hardly be suggested that such a system is inconsistent with democracy.  If defection is permissible, the details of the legislation must be left to Parliament, subject always to the provisions not being inconsistent with the Constitution.  The mere fact that Parliament decides that a threshold of 10% is necessary for defections from a party, is not in our view inconsistent with the Constitution.”[66]

 

[115]     In my view, the applicant has failed to discharge the onus of proving that the model articulated by Parliament infringes on the equal protection provisions provided for in sections 3(2)(a) and 9(1) of the Constitution as alleged.

 

[116]     Firstly, and as already demonstrated, the applicant’s argument that the 200/200 split arbitrarily differentiates between independent candidates and political parties is without merit.  Parliament has articulated and proven that there is a rational basis for the split, which is to facilitate proportional representation and to avoid the risk of overhang.  In keeping with what this Court said in New National Party, as there is a proven rational connection between the split and the articulated legitimate governmental purposes, it cannot be maintained that the split violates section 9(1) of the Constitution.

 

[117]     Secondly, the proposition that a vote for an independent candidate carries less weight when compared to a vote for a political party is also without merit.  As argued by the respondents, independent candidates and political parties compete for the same quota in regional elections and the votes carry exactly the same weight.  There is no differentiation in respect of regional seats.

 

[118]     The applicant accepts the reservation of compensatory seats for political parties but rejects the reservation of 200 seats on the basis that independent candidates need double the votes that political parties need to win a seat since a vote for a political party counts twice.  It is on this basis that the applicant alleges that a vote for an independent candidate carries less weight when compared to a vote for a political party in violation of section 3(2)(a).

 

[119]     This argument must also be rejected.  As pointed out by the Electoral Commission, the applicant fails to appreciate that, in a two-tier system, there are two different quotas that apply in respect of compensatory and regional seats.  Within the different regions, there will also be different quotas.  Consequently, the proposition that the number of votes an independent candidate or political party must obtain in order to gain a seat in the regional tier should be equivalent to the quota for a compensatory seat is logically flawed.  Evidently, some difference must be had, the issue lies in the extent of the difference and whether it serves to exclude independent candidates disproportionately.  This, though, the applicant has failed to establish.

 

[120]     Finally, the applicant’s argument that the Electoral Act as amended devalues a vote for an independent candidate is based on the assumption that a voter that votes for a political party will vote for the same party on both the regional and compensatory ballot, which in turn means that independent candidates need double the votes needed by political parties to gain a seat in the National Assembly.  Put differently, the applicant and the Atkins Report, without providing any basis to this Court, assume that voters who vote for a political party will not split their votes across different parties in the two ballots.  By parity of reasoning, vote splitting beyond a de minimis (trivial) level by voters who vote for political parties would be fatal to the applicant’s right’s violation challenge.

 

[121]     While the 2024 national elections will be the first elections South Africans will get to vote using two ballots for the National Assembly, it is not inconceivable that voters may choose to split their regional and compensatory votes.  There is in fact evidence of vote splitting by South African voters in the last provincial and national elections.

 

[122]     By way of illustration, the 2019 report of the Mapungubwe Institute for Strategic Reflection (MISTRA)[67] looked at, amongst others, vote splitting in the 2019 provincial and national elections in South Africa.[68]  The report found that “there were no cases where the party-specific vote totals for the national and provincial votes corresponded exactly”.[69]  In this regard, the report said the following:

 

“The ANC in both Election 2014 and 2019 was the only party that scored more votes nationally in all nine provinces than provincially.  In addition, in all provinces but the Eastern Cape, this ANC margin (or ‘vote edge’) grew from 2014 to 2019 (Table 9).

For all of the rest of the top-four parties (DA, EFF and FF+ in 2019) there were diverse trends.  Amongst the 2019 opposition parties that fared better at national than provincial level, were the FF+ in six of the provinces, the EFF in four, and the DA in three (KwaZulu-Natal, North West and Limpopo).  In the Eastern, Northern and Western Cape, and in KwaZulu-Natal, the EFF did better nationally than provincially.  In 2014, in the EFF’s first provincial elections, it fared better nationally than provincially in six of the nine provinces.  The DA in contrast fared better at the provincial than the associated national level in 2019 in six of the provinces – compared with doing better nationally than provincially in seven provinces in 2014.  This indicates that many of its supporters endorsed another party. . .at the national level rather than voting DA at both the provincial and national levels in 2019.  The EFF’s 2019 national-level performance was affected similarly at several sites.  In five of the provinces, the EFF recorded fewer votes at the national than the provincial level (the inverse therefore of the situation outlined regarding 2014, when its national performance was better): Gauteng, North West, Limpopo, Mpumalanga and the Free State.  Hence, judged by these vote-splits, both DA and EFF supporters in 2019 bolstered the ANC’s ‘Ramaphosa project.’”[70]

 

[123]     Considering the above, it is clear that one cannot discount the possibility of vote splitting with the introduction of the second ballot.  When one considers the 2021 local government elections, different political party support across the different provinces is even more stark when compared to the 2019 provincial and national elections.  The Electoral Commission makes this exact point in its explanatory affidavit.  The Electoral Commission explains that “in municipal election where independent candidates stand for political election as ward candidates, most voters that support independent candidates also choose to support a political party on the second ballot”.  Admittedly, local government elections are conducted differently to provincial and national elections, not to mention that voters may well have different considerations when voting in local government elections compared to national elections.  However, the different levels of support that political parties received in the provinces across the different elections is nonetheless instructive with regard to vote splitting.

 

[124]     It must be acknowledged that the opportunity for vote splitting is not the purpose of the recent amendments to the Electoral Act.  The possibility of vote splitting is an unavoidable consequence of Parliament having introduced two ballot papers in order to accommodate independent candidates to comply with this Court’s ruling in New Nation II.  Nevertheless, because two ballots have been introduced, voters will now have a tactical opportunity to split their vote between regional and compensatory ballot (or vote for the same political party on both ballots if they so choose) and one cannot say with certainty that voters will not use this opportunity to split their vote.  This proposition is even more compelling when regard is had to the fact that some voters appear already to be voting for different political parties in the national and provincial elections as evidenced by the 2019 election results.

 

[125]     Notwithstanding the above, I do not think an electoral law could ever be properly formulated on the basis of an assumption as to how people will vote as the applicant’s challenge suggests.  As already explained, the applicant’s challenge to the 200/200 split is premised on the assumption of identical voting across the two ballots for political parties.  An electoral law must always assume a complete freedom for the electorate to vote as they please, including splitting their vote where an opportunity to do so exists.  Section 19 of the Constitution guarantees citizens a right to make free political choice.  If, when formulating an electoral law, the Legislature had to take into account how people have voted in the past, it might have to change the law every election cycle as voters are free to vote as they choose and often vote differently in different elections.

 

[126]     In addition to the above, it must be stressed that the applicant has failed to provide compelling evidence to support the alleged inequity between independent candidates and political parties as required by the jurisprudence of this Court.  The applicant has not been able to prove that (a) the split is arbitrary and (b) a vote for an independent candidate carries less weight.  Bearing in mind that the mere difference in effect of the split is not, in and of itself, evidence of unequal treatment in the manner protected against in the Constitution.

 

[127]     For these reasons, the section 3(2)(a) challenge also fails.

 

Section 19(2) and 19(3) of the Constitution

[128]     The applicant argues that the 200/200 split violates section 19(2)[71] of the Constitution in that it undermines the fairness of the outcome of the elections.  The applicant argues further that the split violates the “right to vote” and to “stand for public office” as provided in section 19(3)[72] of the Constitution.  The applicant submits that section 19(3) entails a right to a vote that has equal weight and the right to stand an equal chance of being elected.  In support of this submission, the applicant relies on international law^[73] and jurisprudence from Canada,^[74] the United States of America^[75] and Australia,^[76] which recognise the principle of one person, one vote.  The applicant also refers to the United Nations Human Rights Committee’s General Comment 25 on Political Participation in which this principle is reflected as:

 

“[T]he vote of one elector should be equal to the vote of another.  The drawing of electoral boundaries and the method of allocating votes should not distort the distribution of voters or discriminate against any group and should not exclude or restrict unreasonably the right of citizens to choose their representatives freely.”^[77]

 

[129]     Therefore, the applicant submits that the 200/200 split infringes on the right to equal benefit of the law and the right to stand for public office as votes for independent candidates weigh less than votes for political parties and makes it more difficult for an independent candidate to be elected.

 

[130]     Parliament submits that the split does not infringe on the right to free and fair election, nor does it infringe on the “right to vote” and to “stand for public office”.  Parliament notes that there is no internationally accepted definition of the term “free and fair election”, and that this is ultimately a value judgement.  Parliament submits that this Court has distilled some elements in Kham[78] as being fundamental to the conduct of free and fair elections, which include, amongst others, that: (a) a person entitled to vote should be registered to do so, and those not entitled to vote should not be permitted to do so; and (b) the Constitution protects not only the act of voting and the outcome of elections, but also the right to participate in elections as a candidate and to seek public office.

 

[131]     Parliament argues that the Electoral Amendment Act gives effect to section 19(2) and 19(3) in that the Act allows for citizens to stand for political office and to hold office if elected.  The Electoral Amendment Act also allows for citizens to vote in an election, and they can choose to either vote for an independent candidate (at the regional tier) or a political party.  The requirements and quota for a single regional seat are the same for both independent candidates and political parties.  The compensatory seats are created and reserved for political parties for reasons already discussed.

 

[132]     The difference in outcome, argues Parliament, emanates from the inherent difference between political parties and independent candidates.  The principle of one person one vote is not violated by the Electoral Amendment Act.  According to Parliament, the international law and foreign jurisprudence cited by the applicant do not advance the applicant’s case because those countries do not have a constitutional injunction for an electoral system that results, in general, in proportional representation.  Further, the Australian and Canadian cases that the applicant relies on do not require strict equality of voting power.  Instead, in Reference re Prov. Electoral Boundaries,[79] the Canadian case that the applicant relies on, it was recognised that some deviation from the concept of equal representation can be acceptable if it is justifiable under the Canadian Charter.

 

[133]     Parliament further argues that, pursuant to this Court’s judgment in New National Party,[80] for the section 19 challenge to be successful, the applicant must not merely show that Parliament has performed its function in a manner which has resulted in a denial of the vote to a substantial number of citizens.  Rather, they must show that the machinery, mechanism or process provided by the Electoral Amendment Act is not reasonably capable of ensuring the vote for those who want to vote and take reasonable steps to exercise their right to vote.

 

[134]     The Minister, in response to the section 19 challenge, submits that the applicant has failed to demonstrate how the freeness and fairness of the election would be compromised by the 200/200 split as contended.  This challenge, argues the Minister, is inconsistent with the concession made by the applicant and Mr Atkins that the 200/200 split will achieve proportionality.  The Minister also states that the argument that votes for independent candidates count for less is without merit since independent candidates compete for the same quota as political parties in respect of regional seats and the difference in compensatory seats has been justified.

 

[135]     Like Parliament, the Minister also states that the jurisprudence of Australia, Canada and the United States of America does not assist the applicant.  According to the Minister, while the South African Constitution prescribes an electoral system that must result, in general, in proportional representation, the electoral systems of Australia, Canada and the United States of America are constituency-based.  In constituency- based systems, the political party or candidate who wins the highest number of constituencies wins the election regardless of whether they have the highest level of proportional support.  Thus, overall party support is inconsequential.  The Minister submits that this is the exact opposite of what our Constitution envisages and what it guards against.  Further, the case law referred to concerned issues surrounding whether voters in constituencies were evenly distributed from a geographical standpoint.  According to the Minister, this is irrelevant to the issues in this matter.

 

[136]     As already demonstrated, the Electoral Commission denies that the 200/200 split causes a vote for an independent candidate to count for less towards the outcome.  It argues that this argument is premised on fundamental misconceptions by the applicant and Mr Atkins as discussed in this judgment.[81]

 

[137]     For reasons I have stated in the preceding paragraphs, the applicant has failed to establish that a vote for an independent candidate carries less weight than a vote for a political party.[82]  As such, the argument that the 200/200 split infringes the section 19 rights on the basis that a vote for an independent candidate does not have equal weight and that independent candidates do not stand an equal chance of being elected, is without merit.

 

[138]     The applicant fails to get out of the starting blocks on this challenge.  Notwithstanding this shortcoming in the applicant’s case, it is important to consider how this Court has dealt with section 19(2) and 19(3) challenges in other cases.

 

[139]     I think it apt to remind ourselves of the test for determining a rights violation set out by this Court in Ex Parte Minister of Safety and Security: In Re S v Walters,[83] where, in assessing whether the impugned provisions limited the right to life, to human dignity and to bodily integrity as protected by the Constitution, Kriegler J said the following:

 

“As observed at the outset, the Bill of Rights spells out the fundamental rights to which everyone is entitled and which the State is obliged to respect, protect, promote and fulfil.  An enactment (like section 49) may limit these rights only if – and to the extent that – the limitation can be justified under section 36 of the Constitution.  Otherwise it has to be declared invalid under section 172(1).  This is essentially a two-stage exercise.  First, there is the threshold enquiry aimed at determining whether or not the enactment in question constitutes a limitation on one or other guaranteed right.  This entails examining (a) the content and scope of the relevant protected right(s) and (b) the meaning and effect of the impugned enactment to see whether there is any limitation of (a) by (b).  Subsections (1) and (2) of section 39 of the Constitution give guidance as to the interpretation of both the rights and the enactment, essentially requiring them to be interpreted so as to promote the value system of an open and democratic society based on human dignity, equality and freedom.  If upon such analysis no limitation is found, that is the end of the matter.  The constitutional challenge is dismissed there and then.”[84]  (Emphasis added.)

 

[140]     The first leg of the Walters test examines “the content and scope of the relevant protected right(s)”.  In the present case, this refers to the political rights protected under section 19(2) and 19(3) of the Constitution.  In New National Party, this Court recognised the importance of the right to political participation, where it stated that “[t]he importance of the right to vote is self-evident and can never be overstated. . . .  [T]he right is fundamental to a democracy for without it there can be no democracy.”[85]

 

[141]     In August,[86] Sachs J, commenting on the importance of the right to vote, stated that “[t]he universality of the franchise is important not only for nationhood and democracy.  The vote of each and every citizen is a badge of dignity and of personhood.  Quite literally, it says that everybody counts.”[87]

 

[142]     The Constitution in various sections provides for the regulation of the exercise of the right to vote.  As stated in New National Party, the effect of these measures on the effective exercise of the right to vote is the following:

 

“(a)      National, provincial and municipal elections must be held in terms of an electoral system which must be prescribed by national legislation.

(b)        The electoral system must, in general, result in proportional representation.

(c)        Elections for the national assembly must be based on the national common voters roll.

(d)        Elections for provincial legislatures and municipal councils must be based on the province’s segment and the municipality’s segment of the national common voters roll respectively.”[88]

 

[143]     Commenting on the inter-relatedness of the rights conferred in section 19(2) and 19(3), this Court in New Nation II said the following:

 

“The condition [the singular condition that section 19(3) imposes for the exercise of the right to vote] reveals the inter-relatedness between the right to vote and the right to free, fair and regular elections which is guaranteed by section 19(2).  If the elections are not free and fair, there can be no proper exercise of the right to vote and consequently the content of the right to vote itself would be emasculated.  And that would place at risk the entire democratic project.  This illustrates that the right to vote is vital to our democratic order, not only with regard to who gets the honour of exercising political power, but also in respect of the policies to be adopted in governing the country.”[89]

 

[144]     Therefore, the right to vote and the right to free and fair elections are closely linked.  While the right to vote gives content to the right to free and fair elections, the right to free and fair elections has implications for how the right to vote must be exercised.

 

[145]     In Kham, this Court set aside the results of municipal by-elections held in various wards in the Tlokwe Local Municipality between August and December 2013 on the ground that they were not free and fair due to irregularities in voter registration.  In a unanimous judgment, Wallis AJ held that there is no internationally accepted definition of what constitutes a free and fair election, and that ultimately, whether an election is free and fair is a value judgement which must always be assessed in context.[90]  A Court required to make this value judgement must weigh all the evidence and determine whether the constitutional requirement that an election be free and fair was satisfied.[91]

 

[146]     The Court then went on to distil the following elements as fundamental in the conduct of free and fair elections:

 

“First, every person who is entitled to vote should, if possible, be registered to do so.  Second, no one who is not entitled to vote should be permitted to do so.  Third, insofar as elections have a territorial component, as is the case with municipal elections where candidates are in the first instance elected to represent particular wards, the registration of voters must be undertaken in such a way as to ensure that only voters in that particular area (ward) are registered and permitted to vote.  Fourth, the Constitution protects not only the act of voting and the outcome of elections, but also the right to participate in elections as a candidate and to seek public office.”[92]

 

[147]     The cases cited in the preceding paragraphs give meaning to the “content and scope” of the section 19 political rights.

 

[148]     The second leg of the Walters test requires of us to examine the “meaning and effect” of the 200/200 split as provided in the amended Electoral Act to assess whether there is any limitation to the section 19 political rights as alleged by the applicant.  I have already explained the meaning of the amended Electoral Act, and in particular, the impugned 200/200 split.[93]  It is the “effect” of the impugned provisions on the section 19 rights that we are concerned with at this stage.

 

[149]     Before determining the question of the effect of the impugned provisions, it is worth noting that Walters did not concern section 19 political rights, but rather, at issue was the constitutional validity of the provisions of section 49(1)(b) and 49(2) of the Criminal Procedure Act[94] which essentially permitted force to be used when carrying out an arrest.  The Walters two-prong test nonetheless applies with authoritative weight to this case as it is a general approach for determining a rights violation – regardless of the content of the right(s) alleged to have been violated.

 

[150]     Now returning to the effect of the 200/200 split on the section 19 political rights.  As shown above, this Court in Kham distilled elements fundamental to free and fair elections as protected under section 19(2) of the Constitution; these elements include that (a) persons entitled to vote should be registered to do so, and (b) the protection of the right to participate in elections as a candidate and to seek public office.  In my view, the elements distilled by this Court in Kham are met under the 200/200 split.  As Parliament points out in its submissions, the Electoral Amendment Act entitles every adult citizen to a vote, and they can, in keeping with this Court’s ruling in New Nation II, vote for either a political party or an independent candidate.  The Electoral Amendment Act also allows citizens to stand for political office and to hold office if elected.

 

[151]     The contention that a vote for a political party counts double to that of an independent candidate, which as the applicant argues leads to a limitation of the section 19 rights, is, as I have already demonstrated, without merit.  This argument is based on the assumption that voters will not split their vote, an assumption which I have already shown to be unsustainable.

 

[152]     Further, and as stated in New National Party, the effective exercise of the right to vote as protected under section 19(3) includes, among others, the following: (a) that elections must be held in terms of an electoral system which must be prescribed by national legislation; (b) an electoral system that must, in general, result in proportional representation; (c) that elections for the National Assembly must be based on the national common voters’ roll.  The Electoral Amendment Act gives effect to these principles in that (a) it is national legislation that prescribes an electoral system, (b) the 200/200 split prescribed therein results in proportional representation, and (c) it provides for a national common voters’ roll for the election of the National Assembly.  Therefore, it cannot be said that the Electoral Amendment Act limits section 19(3).

 

[153]     It is worth reiterating what this Court said in United Democratic Movement,^[95] that is, laws of general application will invariably, directly or indirectly, have the potential to affect different categories of people in different ways, and the fact that a law affects different categories of people differently is not evidence of a violation of the Constitution.

 

[154]     While it is quite clear that provisions which disenfranchise citizens or preclude them from standing for office are unconstitutional, this must be contrasted with instances where the design of the electoral system (which entails legitimate competing options available to Parliament) is challenged – that is, the manner in which Parliament chooses to give substantive content to the right to political participation.[96]

 

[155]     The applicant was required to show that the measures adopted by Parliament constitute a limitation of the political rights alleged.  For the reasons stated, the applicant has failed to discharge this onus and as such the 200/200 split passes the second leg of the Walters test.  It follows that the section 19 challenge must also fail.

 

Foreign jurisprudence and international instruments

[156]     Before I conclude, I wish to provide a brief response to the submissions on the foreign jurisprudence relied on by the applicant.  I agree with the respondents that it does not assist the applicant’s case.  The South African Constitution prescribes an electoral system that must result, in general, in proportional representation.  The electoral systems of Australia, Canada and the United States of America, which the applicant relies on, are constituency-based, where the overall party support is inconsequential.  The Minister correctly submits that this is the exact opposite of what our Constitution envisages and what it guards against.

 

[157]     We must never forget that the Constitution represents a decisive break from the past.  The Constitution is intentional in affording all South African citizens the right to political participation on equal terms.  While this means that all South African citizens now have the right to vote and stand for office, I believe that, in the context of this case, the more important consideration is the break from geographic voting constituencies which allowed the National Party to come to power in 1948 with 37% of the total vote while the United Party had 49%.  Throughout its rule, the National Party designed a system which entrenched their power and excluded most South Africans from participating in public affairs.  The respondents are, therefore, correct that our electoral systems are incomparable.

 

Section 36 limitation analysis

[158]     In the light of my conclusion that the impugned schedule does not violate any of the fundamental rights alleged by the applicant, it is not necessary to conduct the justification analysis under section 36 of the Constitution.

 

Conclusion

[159]     The applicant has not established that the impugned legislation is irrational, nor that it infringes a provision in the Bill of Rights.  Therefore, it has not made out a case to justify the declaration of constitutional invalidity.  Accordingly, the application must be dismissed.

 

Costs

[160]     As established in Biowatch,[97] in order to combat the unduly chilling effect of adverse costs on constitutional litigation, this Court held that the general rule for an award of costs in constitutional litigation between a private party and the state is that if the private party is successful, then it should have its costs paid by the state; however, if the private party is unsuccessful, each party should then pay its own costs.[98]  In this matter, the applicant has not been successful, therefore, each party must pay their own costs.

 

Order

[161]     The following order is made:

1.                  Direct access is granted.

2.                  The application is dismissed.

3.                  There is no order as to costs.

For the Applicant:

 

T G Madonsela SC, H J De Waal SC, T Palmer, Y S Ntloko and C Louis instructed by Strauss Daly Incorporated.

 

For the Second and Third Respondents:

 

K Pillay SC and L J Zikalala instructed by the State Attorney.

 

For the Fourth Respondent:

 

S Budlender SC, A Nacerodien and M De Beer instructed by the State Attorney.

 

For the Fifth Respondent:

 

A Bham SC and J Bleazard instructed by Moeti Kanyane Incorporated.

 

Counsel for the Amicus Curiae:

 

M du Plessis SC, P Maharaj-Pillay and C Kruyer instructed by Webber Wentzel.

---