One Movement South Africa NPC v President of the Republic of South Africa and Others (CCT 158/23) [2023] ZACC 42 (4 December 2023)

CONSTITUTIONAL COURT OF SOUTH AFRICA


Case CCT 158/23

In the matter between:


ONE MOVEMENT SOUTH AFRICA NPC Applicant

and

PRESIDENT OF THE REPUBLIC OF SOUTH AFRICA First Respondent

MINISTER OF HOME AFFAIRS Second Respondent

INDEPENDENT ELECTORAL COMMISSION Third Respondent

SPEAKER OF THE NATIONAL ASSEMBLY Fourth Respondent

CHAIRPERSON OF THE NATIONAL COUNCIL
OF PROVINCES Fifth Respondent

and

RIVONIA CIRCLE NPC Amicus Curiae



Neutral citation: One Movement South Africa NPC v President of the Republic of
South Africa and Others [2023] ZACC 42

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


Judgments: Zondo CJ (uninamous in respect of direct access, minority in
respect of the signature challenge and majority in respect of the
recalculation challenge): [1] to [212]

Kollapen J (majority in respect of the signature challenge ): [213]
to [355]
Theron J (partial dissent): [356] to [385]

Heard on: 30 August 2023

Decided on: 4 December 2023

Summary: Section 31B(3) of the Electoral Act 73 of 1998 as amended by the
Electoral Amendment Act 1 of 2023 — unconstitutional —
declaration of invalidity suspended for 24 months —
interim reading-in

Unjustifiable limitation — sections 18(1), 19(1) and 19(3) of
the Constitution — regulation versus limitation

Items 7, 12 and 23 of Schedule 1A of the Electoral Act 73 of 1998
as amended by the Electoral Amend ment Act 1 of 2023 —
constitutional — challenge dismissed




ORDER



On application for direct access the following order is made:
1. The applicant is granted direct access to this Court.
2. The recalculation relief sought in prayers 4, 6.2 and 6.3 of the applicant’s
notice of motion is refused.
3. Section 31B(3)(a)(i) and (ii) of the Electoral Act 73 of 1998 ( Electoral Act)
as inserted by the Electoral Amendment Act 1 of 2023 is declared invalid and
inconsistent with the Constitution, to the extent that it unjustifiably limits the
rights to freedom of association, freedom to make political choices and to
stand for public office.
4. The declaration of invalidity referred to in paragraph 3 is suspended for
24 months from the date of this order to afford Parliament an opportunity to
remedy the constitutional defects giving rise to the constitutional invalidity.


5. From the date of the order of this Court and during the period of suspension,
section 31B(3)(a)(i) and (ii) of the Electoral Act will read as follows, the
underlined words being read into the section with the words in strike-out text
being severed:
“(3) The following must be attached to a nomination when it is submitted:
(a) A completed prescribed form confirming that the independent
candidate has submitted, in the prescribed manner, the names,
identity numbers and signatures of voters whose names
appear—
(i) in the case of an election of the
National Assembly in respect of regional
seats, on the national segment of the voters’
roll a nd who support his or her candidature ,
totalling 1 000 signatures for each region in
which the candidate intends to contest an
election;
(aa) totalling 15 percent of the quota for
that region in the preceding election,
if intending to contest only one
region; or
(bb) totalling 15 percent of the highest of
the regional quotas in the preceding
election, if intending to contest more
than one region, provided that where
15 percent of the highest of the
quotas is not achieved, that the
independent candidate may only
contest the region or regions as
determined by the next highest quota;
(ii) in the case of an election of a provincial
legislature, on the segment of the voters’ roll
for the province and who support his or her
candidature, totalling 1 000 signatures least
15 percent of the quota of that province in the

4
preceding election, which the independent
candidate intends to contest,
provided that an independent candidate who was elected to
either the National Assembly or a provincial legislature as an
independent candidate in the preceding election shall be
exempt from this requirement.”
6. In the event that Parliament does not remedy the constitutional deficiency in
section 31B(3)(a)(i) and (ii) within the period provided for in paragraph 4 of
this order, or any extended period grante d by this Court, then
section 31B(3)(a)(i) and (ii) will be deemed to read as set out in paragraph 5
above.
7. The second, fourth and fifth respondents, jointly and severally, are to pay
fifty percent (50%) of the applicants’ cos ts, which costs shall include the
costs of two Counsel. The applicant is not entitled to recover any costs
associated with the report of Mr Atkins.



JUDGMENT




ZONDO CJ (Mathopo J, Schippers AJ and Van Zyl AJ concurring in respect of the
whole judgment. Maya DCJ, Mhlantla J and Kollapen J concurring in respect of only
direct access and the recalculation point):


Introduction
[1] One Movement South Africa NPC (OSA or applicant ) is a registered
not-for-profit company. It has as some of its members individuals who intend to contest
the national and provincial elections that will be held in this country some time in 2024
as independent candidates.

ZONDO CJ
5
[2] Clause 2 of OSA’s constitution provides that OSA’s vision is “a society in which
all men and women, regardless of background and race, live together side by side
peacefully and [will] be able to prosper together”. OSA’s constitution further provides
in part that OSA envisions “a South Africa that is:
2.1.1. a crime free society
2.1.2. a racially cohesive and integrated society
2.1.3. an educated society with high quality jobs
. . .
2.1.8. a society where political leaders are accountable.”

[3] OSA has cited the President of the Republic of South Africa as the
first respondent, the Minister of Home Affairs (Minister) who is the Minister
responsible for the administration of the Electoral Act1 as the second respondent, the
Independent Electoral Commission (IEC or Commission ), which is the body
constitutionally mandated to manage elections in this country, as the third respondent,
the Speaker of the National Assembly (Speaker) as the fourth respondent and the
Chairperson of the National Council of Provinces (NCOP) as the fifth r espondent.
Rivonia Circle NPC was admitted as an amicus curiae. We are indebted to the amicus
for its submissions and its assistance in this matter. The President abides the decision
of this Court. The Minister, the Speaker of the National Assembly and the Chairperson
of the NCOP oppos e the application. The IEC abides the decision but delivered an
explanatory affidavit.

Should direct access be granted?
[4] OSA has brought an application for direct access to this Court as a matter of
urgency for an orde r declaring certain provisions of the Electoral Amendment Act 2
(EAA) inconsistent with the Constitution and, therefore, invalid. This Court grants
direct access when it is in the interests of justice to do so. Direct access means that a

1 73 of 1998.
2 1 of 2023.
ZONDO CJ
6
matter is brought to this Court without first having been taken to another Court. Where
this Court grants direct access, it decides the matter as a court of first and final instance.
That is not something that this Court does lightly because we appreciate the importance
of having the benefit of the views of other courts on issues before they are decided by
this Court. There will have to be exceptional circumstances before it can be said that it
is in the interests of justice for this Court to grant direct access in a matter.

[5] The justification advanced by OSA for its application for direct access and for
this matter to be dealt with on an urgent basis or to be given a certain amount of priority
is this. Parliament passed the EAA to correct the constitutional defect identified in the
Electoral Act by this Court’s judgment in New Nation Movement.3 OSA contends that,
although the EAA will enable independent candidates to stand for election in next year’s
elections and in future elections, it infringes their right to stand for election to public
office, their right to dignity and their right to disassociate.

[6] OSA, therefore, seeks to ensure that by the time of the elections next year the
constitutional defects which it contends are to be found in the EAA have been corrected.
Accordingly, OSA contends that it is of critical importance that this Court decides the
constitutional validity of the relevant provisions of the EAA while there is enough time
for Parliament to correct any constitutional defect of the EAA that this Court may
identify if it upholds OSA’s contentions.

[7] The date for next year’s elections has not yet been fixed. However, it is accepted
by all parties in this matter that constitutionally the date for the elections is required to
be somewhere between early May and sometime in August 2024. OSA argues that, if
it had to first take this matter to the High Court and , thereafter, come to this Court for
confirmation of any declaration of invalidity that the High Court might make, that
would cause such a delay that there would be a serious risk that, by the time the matter
was brought to this Court, there would not be enough time for Parliament to correct any

3 New Nation Movement NPC v President of the Republic of South Africa [2020] ZACC 11;
2020 (6) SA 257 (CC); 2020 BCLR 950 (CC).
ZONDO CJ
7
constitutional defect that this Court could identify if it uph eld OSA’s contention .
Indeed, the IEC might also not have enough time to make all the arrangements that may
need to be made in order to ensure that the elections are free and fair.

[8] None of the respondents opposed OSA’s application for direct access and for this
matter to be accorded a certain amount of priority. I am of the view that OSA has
reasonable prospects of success. I agree that, if OSA was required to first take this
matter to the High Court before bringing it to this Court, by the time this Court handed
down its judgment, there would not be enough time for Parliament to correct whatever
constitutional defect this Court might identify i n the E AA if it upheld OSA’s
contentions. Assuming that OSA’s complaint about the EAA is well founded, that
would cause OSA and independent candidates serious prejudice in next year’s elections.
It is, therefore, in the interests of justice for this Court to grant OSA direct access.
Accordingly, leave is hereby granted to OSA to bring this matter directly to this Court.

Background
[9] Since the advent of our constitutional democracy in the April 1994 elections for
national and provincial legislative bodies in this country have been conducted on the
basis that no individual could stand for election to any such bodies. Voters voted only
for political parties. In turn political parties provid ed lists of persons that they
nominated for membership of those legislative bodies if they won seats in those bodies.
In 2020 this Court handed down its judgment in New Nation Movement and declared
the Electoral Act inconsistent with the Constitution and, therefore, invalid to the extent
that it did not allow an individual adult citizen to stand for election to the legislative
bodies at national and provincial level s without having to be a member of a political
party. This Court suspended that declaration of invalidity for 24 months to afford
Parliament the opportunity to correct the constitutional defect within that period.

[10] In New Nation Movement this Court, through M adlanga J, said that there were
two issues before it. The first was whether – by making acc ess to political office
possible only through membership of a political party – the Electoral Act unjustifiably
ZONDO CJ
8
limited the right to freedom of association guaranteed in section 18 of the Constitution.
The second issue involved a determination of the content of the right enshrined in
section 19(3)(b) of the Constitution and whether the Electoral Act unjustifiably limited
that right.4

[11] Madlanga J later stated that, on its own, the freedom of association challenge
begged the question. He said that that challenge could not be dealt with as a standalone
challenge. He went on to say that this did not mean that freedom of association was
irrelevant to determining the content of the right provided for in section 19(3)(b) of
the Constitution.5

[12] Section 19 of the Constitution deals with entrenched political rights. It reads:

“Political rights
19. (1) Every citizen is free to make political choices, which
includes the right—
(a) to form a political party;
(b) to participate in the activities of, or recruit
members for, a political party; and
(c) to campaign for a political party or cause.
(2) Every citizen has the right to free, fair and regular
elections for any legislative body established in terms
of the Constitution.
(3) Every adult citizen has the right—
(a) to vote in elections for any legislative body
established in terms of the Constitution, and to
do so in secret; and
(b) to stand for public office and, if elected, to
hold office.”


4 New Nation Movement above n 3 para 10.
5 New Nation Movement above n 3 para 4.
ZONDO CJ
9
Early in his judgment in New Nation Movement, Madlanga J said:

“[17] I want to lay emphasis on subsection (1). It affords every
citizen the freedom to make political choices. The fact that what are
itemised in the subsection as being the choices a citizen is free to make
all relate to political parties does not mean those choices concern
political parties only. If that were the case, instead of saying these rights
or freedoms “include”, the subsection would simply have said the rights
“are”. The present formulation means the rights are more than what is
itemised. As the first applicant submits, paragraphs (a) to (c) of
section 19(1) ‘are mere examples of “political choices”; they do not
cover the field of what [section 19(1)] protects ’. A conscious choice
not to form or join a political party is as much a political choice as is
the choice to form or join a political party; and it must equally be
deserving of protection.
[18] Once an adult citizen is forced to exercise the section 19(3)(b)
right through a political party, that divests her or him of the very choice
guaranteed by section 19(1) not to form or join a political party. That
cannot be. We must strive for a reading that does not truncate the full
effect of any of the rights afforded by section 19. The respondents’
reading of section 19(3)(b) resul ts in a diminution of the political
choices afforded by section 19(1). Effectively that gives rise to a
conflict between the two subsections.”

[13] In New Nation Movement this Court referred to UDM II.6 There, the question
was whether members of the National Assembly could vote in accordance with their
conscience when voting in a motion of no confidence in President Jacob Zuma.
Madlanga J then referred to this passage in UDM II:

“As is the case with general elections where a secret ballot is deemed
necessary to enhance the freeness and fairness of the elections, so it is
with the election of the President by the National Assembly. This
allows Members to exercise their vote freely and effectively, in

6 United Democratic Movement v Speaker, National Assembly [2017] ZACC 21 ; 2017 (5) SA 300 (CC);
2017 (8) BCLR 1061 (CC) at paras 73-4.
ZONDO CJ
10
accordance with the conscience of each, without undue influence,
intimidation or fear of disapproval by others.
The frustration or disappointment of the losing presidential hopeful and
his or her supporters could conceivably have a wide range of prejudicial
consequences for Members who are known to have contributed to the
loss. To allow Members of the National Assembly to vote with their
conscience and choose who they truly believe to be the best presidential
material for our country, without any fear of reprisals, a secret ballot
has been identified as the best voting mechanism.”7

[14] This Court went on to say in New Nation Movement that, if all members of the
National Assembly were free to vote as they saw fit regardless of how political ly
sensitive an issue might be and without any risk of reprisals from their political parties,
litigation on this issue would not have been necessary. It went on to say that it was
exactly because the opposite prevailed that the litigation was instituted. It said that this
would not have been an issue at all for someone not subject to encumbrances of party
politics; someone not owing his or her membership of the National Assembly to a
political party.

[15] The concern was that the majority of members of the National Assembly did not
necessarily vote in accordance with their true beliefs or convictions on certain matters
that are very important to the country and its people . This was because they were
instructed by their party leadership on how to vote on certain issues. This was also the
reason why the U nited Democratic Movement (UDM) approached this Court in what
has become known as the secret ballot case.8 In that case the UDM wanted the Speaker
to ensure that a secret ballot – as opposed to a show of hands – was conducted in the
National Assembly. This Court held that the Speaker had a discretion to direct that the
ballot in a vote of no confidence in a President be held in secret. The UDM made this
request for the ballot to be held in secret bec ause it believed that there were many
members of the National Assembly who were members of the majority party who had

7 Id.
8 UDM II above n 6.
ZONDO CJ
11
genuinely lost confidence in President Zuma at the time and who would have liked to
vote in support of the motion of no confidence in President Zuma but would not be able
to do so if the ballot was not secret. This would have been because the leadership of
their party either had already given an instruction, or, it was believed that they would
soon give an instruction, that all A frican National Congress (ANC) members of the
National Assembly should not support the motion of no confidence in President Zuma.

[16] When any member of Parliament makes a decision or allows himself or herself
to be party to a decision that he or she knows deep down in his or her heart is not in the
interests of the people of South Africa, he or she betrays the people of South Africa. In
UDM II this Court said through Mogoeng CJ:

“[37] In anticipation of a President and this constitutionally
envisaged team ’s possible remissness in the execution of their
constitutional mandate, provision was made to minimise or address that
possibility. Those who represent the people in Parliament have thus
been given the constitutional responsibility of ensuring that member s
of the executive honour their obligations to the people. Parliament, that
elects the President and of which the Deputy President, Ministers and
their deputies are members, not only passes legislation but also bears
the added and crucial responsibility o f ‘scrutinising and overseeing
executive action.’”9

[17] As already stated above , this Court concluded in New Nation Movement that
the EAA was inconsistent with the Constitution and, therefore, invalid to the extent that
it did not allow an adult citizen to stand for public office except through membership of
a political party. The legislation that Parliament has passed in response to that judgment
may well allow independent candidates who will serve in the National Assembly to tilt
the scale to get the majori ty of members in the National Assembly to put the interests
of the people first in whatever they do. That would include when they vote in future
motions of no confidence.

9 UDM II above n 6 at para 37.
ZONDO CJ
12

[18] Subsequent to the handing down of the New Nation Movement judgment, the
Minister commissioned a panel of experts, the Ministerial Advisory Committee
(Committee), to provide technical proposals to give effect to the judgment. This
Committee consisted of eight members who subsequently submitted a report to the
Minister. The Minister attached a copy of that Committee’s report to his affidavit. The
members of th e Committee were divided on the advice to give to the Minister. Four
members proposed a combination of a single member constituency and proportional
representation system. This was the majority option. Three members of the Committee
proposed a modification of the existing system so as to permit independent candidates
to contest the elections. This was the minority option. One member of the Committee
chose not to state a preferred option.

[19] In their report the Committee had this to say in part about the two options:

“Option 1 (Minority report): The slightly modified multi -member
constituency (MMC), which stakeholders referred to as the minimalist
option.
This option entails modifying the existing multi -member electoral
system to accommodate independent candidates in the national and
provincial elections without many changes in the legislation. Those in
favour of this option believe that it does not interf ere with the
constitutionally required general proportionality and is the best option
for ensuring inclusiveness, gender representation, simplicity and
fairness for independents.
Option 2 (Majority report): The mixed-member model incorporating
single-member constituencies.
This option entails combining the first -past-the-post and proportional
representation, making it a mixed-member proportional (MMP) system
resembling the current local government electoral system, albeit with
some improvements. It involves electing MPs from 200 single-member
constituencies and the remainder from a single national multi -member
constituency. Thus, voters would vote for a single MP to represent
ZONDO CJ
13
them in single-member constituencies (their first vote) and for a party
to repres ent them in the single national multi -member constituency
based on competing for closed party lists (their second vote). Those in
favour of this option believe that it does not interfere with the
constitutionally required general proportionality and is th e best option
for ensuring inclusiveness, gender representation, simplicity and
fairness for independents.”

[20] The Minister accepted the minority option. It would seem from his affidavit that
he chose the minority option out of pragmatism, given what could be done before the
2024 elections. He left the door open as to what option could be decided upon after the
2024 elections for long term.

[21] In February 2023 Parliament passed the E lectoral Amendment Bill (Bill). The
President assented to the Bill in April 2023 which then became the EAA. It is now
necessary to deal with the relevant features and sections of the EAA. The EAA permits
independent candidates to stand for public office and, if elected, to hold public office.
There are only two respects in whic h OSA is unhappy about the EAA. The first
complaint relates to section 31B(3) of the EAA. This section relates to a
signature requirement about which more is said below . The other one relates to how
the recalculation of seat allocations is done when an independent candidate is elected,
but their surplus votes are either discarded or when an independent candidate vacated
his or her seat. I propose to deal with the signature requirement first and thereafter with
the recalculation challenge.

The signature requirement
[22] For the first time in the history of our constitutional democracy since 1994 adult
citizens will be able to stand for election to public office in the form of the
National Assembly or provincial legislatures. This is because the EAA makes this
possible. As already indicated above, this follows upon the judgment of this Court in
New Nation Movement.

ZONDO CJ
14
[23] Section 4 of the EAA introduces Part 3A in to Chapter 3 of the Electoral Act.
Part 3A deals with independent candidates. The reference to independent candidates is
a reference to candidates who contest elections independently of a political party.
Section 31A governs the nomination of independent candidates. Section 31B governs
the requirements and qualifications for independent candidates to contest elections.

[24] Section 31B(1) and 3(a)(i) and (ii) reads:

“31B. (1) A person may contest an election as an independent
candidate only if that person is nominated on a
prescribed form and that form is submitted to t he
Commission by not later than a date stated in the
timetable for the election and complies with the
requirements of subsection (3).
. . .
(3) The following must be attached to a nomination when
it is submitted:
(a) A completed prescribed form confirming that
the independent candidate has submitted, in
the prescribed manner, the names, identity
numbers and signatures of voters whose
names appear—
(i) in the case of an election of the
National Assembly in respect of
regional seats, on the national
segment o f the voters’ roll and who
support his or her candidature—
(aa) totalling 15 percent of the
quota for that region in the
preceding election, if
intending to contest only one
region; or
(bb) totalling 15 percent of the
highest of the regional quotas
in the preceding election, if
ZONDO CJ
15
intending to contest more
than one region , provided
that where 15 percent of the
highest of the quotas is not
achieved, that the
independent candidate may
only contest the region or
regions as determined by the
next highest quota; or
(ii) in the case of an election of a
provincial legislature, on the segment
of the voters’ roll for the province and
who support his or her candidature,
totalling at least 15 percent of the
quota of that province in the preceding
election, which the indep endent
candidate intends to contest,
provided that an independent candidate who was
elected to either the National Assembly or a provincial
legislature as an independent candidate in the
preceding election shall be exempt from this
requirement;” (Emphasis added.)

The reference to a quota is a reference to the number of votes a political party or
independent candidate must get in an election in a particular region or province in order
to get one seat.

[25] OSA states that its complaint about section 31B(3) is the requirement that an
independent candidate must secure signatures of registered voters amounting to 15% of
the quota in the previous election in the relevant region. The deponent to OSA’s
founding affidavit, Mr Mogoale, pointed out that OSA’s complaint with regard to the
requirement of signatures was that the EAA placed “an impermissible and arbitrary
barrier to entry for independent candidates to register for elections by placing an
unreasonably high signature requirement upon them” . What this means i s that OSA
ZONDO CJ
16
relied upon the contention that the signature requirement was unreasonably high to
submit that the signature requirement was impermissible and arbitrary. In its founding
affidavit OSA said that it would “demonstrate that the 15% signature requi rement is
arbitrary and poses a barrier to entry for independent candidates” (emphasis added).

[26] OSA puts its chief complaint about the requirement of signatures thus:

“[O]ne of the factors informing this choice is the ability to attain the
required number of signatures in order to register with the IEC. The
chief complaint of OSA is that the Amendment Act requires both
political parties and independent candidates to acquire the same number
of signatures in order to register.” (Emphasis added.)

This statement s ays that the applicant’s chief complaint about the requirement of
signatures was that it applied to both independent candidates and political parties. As
one reads this statement, one expects to see some substantiation of this chief complaint
but there is none.

[27] OSA continued and said in its founding affidavit:

“While on the face of it, it is a formally equal requirement, this results
in a substantively unequal outcome, creates an unfair barrier to entry
for independent candidates and goes agai nst the purpose of
New Nation II.”

After the above excerpt, OSA quoted paragraphs 52 and 53 of this Court’s judgment in
New Nation Movement.10 However, there is nothing in those two paragraphs which

10 New Nation Movement above n 3 at paras 52-3 provides that:
“[52] Being coerced to form or join a political party is an issue that may fundamentally touch
one’s inner core; a matter that goes to one’s conscience. And freedom of conscience is protected
by section 15(1) of the Constitution. It is so that individual members of a legislature remain
‘free to follow the dictates of personal conscience ’. But they do so at their own peril because
‘if [they] wish to be re-elected they need to bear in mind party discipline ’. A classic Hobson’s
choice for somebody who does not want to be shackled by party politics and constraints.
ZONDO CJ
17
supports OSA’s statement at the end of the above excerpt that the requirement that an
independent candidate and a political party that is not yet represented in the National
Assembly or provincial legislature must obtain signatures that are equal to 15% of the
quota goes against the purpose of the New Nation Movement judgment. I shall refer to
such a political party in this judgment simply as a new political party.

[28] OSA states that, in real numbers, for the upcoming 2024 national and provincial
elections independent candidates w ill be required to attain the following number of
signatures (equating to 15% of the relevant quotas) to contest an election:

National Legislature Provincial Legislature
Eastern Cape 11 657 4 672
Free State 11 340 4 286
Gauteng 13 890 8 757
KwaZulu-Natal 13 045 6 664
Limpopo 11 329 4 357
Mpumalanga 11 925 5 886
North West 11 652 4 213
Northern Cape 10 271 4 920
Western Cape 13 201 7 176


Ms Revell, the second applicant, is such a person. We cannot make light of her choice. In the
interpretative exercise, her personal situation is merely illustrative.
[53] We cannot dismissively say if you stand for political office through a political party, it
makes no difference; you still do stand for political office. It may make all the d ifference to
some. And it does to Ms Revell. She explains that, as a representative and leader of the Korana
nation, a section of the Khoi and San people, she is averse to forming or joining a political party.
Hers is not a for -the-sake-of-it objection. I understand her point perfectly. I read it to mean
that, as a leader of a nation, she does not want to be constrained by that kind of partisanship that
comes with being a member of a political party. That partisanship makes you ultimately
answerable to the party. Being free of those shackles will make Ms Revell directly answerable
to her nation, not to a political party. That is the choice she is making. In my book, it is a valid
choice. Surely, her example is not isolated. There must be many and varied other examples.
Subject to the Lavigne threshold, we cannot make light of them.”
ZONDO CJ
18
[29] OSA then deals with the purpose of the signature requirement as it understood
it. OSA states that the stated purpose of imposing a signature requirement on
independent candidates “is that such a requirement will ensure that candidates have a
serious intention of contesting elections and limit the number of frivolous candidates on
the ballot”. OSA state s that this supposedly ensured that the IEC “[was] able to run
elections effectively ”. It also state s that “the signature requirement supposedly
prevented voting from be coming too complicated for voters and prevent ed vote
counting from being overly complicated and taking a much lengthier period”.

[30] OSA also states that, prior to the EAA, political parties were required to submit
1 000 signatures to register themselves as n ational parties with the IEC. OSA goes on
to state that “the only reasonable inference to draw is that 1 000 signatures satisfied the
same purpose so that political parties are serious about entering the election race”. OSA
points out that, instead of ma intaining this 1 000 signature requirement, the EAA
changed this requirement to direct that political parties also attain the 15% entry
requirement. In other words, says OSA, instead of requiring independent candidates to
also obtain 1 000 signatures in the same way that political parties were obliged to obtain
that number of signatures under the Electoral Act, now both political parties and
independent candidates are required to secure 15% of a quota of the relevant region in
respect of a previous election.

Minister Motsoaledi
[31] According to the Minister, the purpose of the signature requirement is “to ensure
prospective independent candidates or prospective political parties seeking to contest
the national elections in fact have some plausible chance of ga ining sufficient public
support to be elected”. The Minister goes on to say in the same paragraph:

“With respect, this eligibility requirement is eminently rational,
sensible and constitutionally permissible. If an independent candidate
(or political party) seeking to contest the elections is unable to gather
enough support to equate to 15% of the votes likely to be required to
ZONDO CJ
19
obtain a seat, they would be extremely unlikely to obtain enough votes
to meet the quota for a seat in the election.”

[32] If the quota of the previous election was 44 000, an independent candidate would
need to obtain 15 % of 44 000 which is 6 600. If, therefore, an independent candidate
cannot obtain 6 600 signatures of registered voters to contest elections, how can he or
she hope to get 44 000 votes during the actual election? It would be highly unlikely.
15% is under one sixth of the votes he or she would require to win a seat.

[33] The Minister says that the purpose of the 15% signature requirement is also to
act as a sifting mechanism to ensure the integrity of the general elections. He says that
this requirement seeks to discourage independent candidates and political parties who
have no plausible hope of obtaining a seat from contesting the elections. The
requirement also seeks to ensure that the ballots are kept to a manageable length and
that voters are not overwhelmed by unnecessarily long and unwieldly ballots.

[34] The Minister contends that OSA’s attack on section 31B (3) of the EAA is
misplaced because, if it were to be upheld, the result would be to lower the threshold
for independent candidates (unrepresented in the legislatures ) to contest an election
while making it more difficult for political parties (unrepresented in the legislatures) to
contest elections. He points out that this would be so because , while OSA challenges
the constitutional validity of section 31B (3), it did not direct any challenge to
section 27(2)(cB) of the EAA which makes the signature requirement applicable to
political parties that are not yet represented in the legislative bodies. He submits that
there is no basis for this differentiation and Parliament had correctly made the signature
requirement applicable to both independent candidates and new political parties.

[35] The Minister dispute s OSA’s contention that section 31B(3) limits the right to
stand for public office in section 19(3)(b) of the Constitution and other political rights.
He contends that the Constitution recognises that it is necessary to regulate the exercise
of the right to stand for public office. He points out that, without reasonable regulation
ZONDO CJ
20
of such right, it would be impossible to give substantive content to the right. He states
that the Constitution anticipates this by requiring that elections be held in terms of an
electoral system prescribed by Parliament. He submits that the requirements that must
be met by a contestant to register for and to contest an election are a constitutiona l
imperative and not a limitation of the right to stand for public office. He also states that
there is a limited number of seats which may be the subject of electoral contestation .
The Minister states that, unless the exercise of that right is regulated properly, too many
candidates would contest elections irrespective of their prospects of success with the
result that the ballot paper could be too long and unwieldly. He submits that, if that
were to happen, namely if anybody and everybody could stand for elections irrespective
of their prospects of success, it would result in an election that is not free and fair. OSA
did not file any further affidavits to dispute this.

[36] In concluding on the OSA challenge to the signature requirement, the Minister
refers to the fact that in the past this Court “similarly evaluated the constitutionality of
voter registration requirements (including the need to have a bar-coded ID book)”. He
points out that this Court concluded that these requirements did not limit rights.

Parliament
[37] Mr Mosa Steve Chabane is the Chairperson of the Portfolio Committee on
Home Affairs in the National Assembly. He deposed to an answering affidavit on
behalf of the Presiding Officers of both the National Assembly and the N COP.
Mr Chabane points out that the purpose of the signature requirement is to minimise the
prospect of frivolous entries into the election race. He also states that “the
signature requirement avoids a situation where by the ballot contains thousands of
names of candidates who have no prospect whatsoever of achieving a sufficient number
of seats”. He says that it avoids a situation of an unwieldly ballot, full of candidates
who have no hope of being elected. He says that this would be confusing to voters and
ultimately could undermine the freeness and fairness of the election.

ZONDO CJ
21
[38] Some of the points made by Mr Chabane are the following:
(a) a balance was struck between ensuring that persons contesting the elections
are serious about the election and have some prospects of being elected and
not placing barriers which unnecessarily and unjustifiably dissuade
participation in contesting elections.
(b) The threshold of 15% is imminently reasonable when regard is had to the
fact that, if an independent candidate or political party is not able to meet
the 15% threshold, this means that they are not able to demonstrate the
support of less than one fifth of the total number of votes required for a seat
in the election.

[39] Mr Chabane also points out that OSA ’s contention that , if the requirement of
15% signatures were to be changed to a requirement of 1 000 signatures, that would
equally achieve the same objective as the objective of 15% is mistaken. He explains
that the requirement of 1 000 signatures applies to political parties when they register
with the IEC and not when they seek to contest elections. He states that the signature
requirement relating to section 31B(3) and the signature requirement that relates to the
registration of political parties with the IEC serve different purposes.

[40] Mr Chabane points out that the signature requirement relating to the registration
of political parties conveys the seriousness of the political party to operate as a serious
and well-organised party. He said that this requirement is not directly concerned with
the ballot or the manageability of elections as a whole. Mr Chabane points out that
there were 331 registered national political parties and a ballot that tried to
accommodate all these political parties would be unworkable.

IEC
[41] Mr S Mamabolo, the Chief Executive Officer of the IEC, states that
section 31B(3) requires an independent candidate to submit supporting signatures from
registered voters in the region or province in wh ich the candidate intends to compete
ZONDO CJ
22
totalling 15% of the quota for a region or province in the previous election (or, if the
candidate is contesting in more than one region in the national elections, totalling 15%
of the highest of those regional quotas). It then says that, based on the 2019 election
quotas, 15% thereof equates to between approximately 10 000 and 14 000 signatures to
contest in the national elections (depending on the region) and between 4 000 and 9 000
signatures to contest in provincial elections (depending on the province).

[42] Mr Mamabolo points out that the requirement of proof of support to contest the
election must be distinguished from the requirement of supporting signatures for the
registration of political parties. The requirements for registration of a political party are
set out in section 15 of the Electoral Commission Act 11 and the Regulations for the
Registration of Political Parties, 200412 (2004 Regulations).

[43] Mr Mamabolo responded to OSA’s statement in its founding affidavit that
political parties were previously required to submit 1 000 signatures of registered voters
to contest elections. OSA said in effect that there was no reason why independent
candidates and new po litical parties were not required to also submit 1 000 signatures
instead of 15% of the quota based on the previous elections. Mr Mamabolo explains
that the requirement of 1 000 signatures of registered voters is still a requirement for
political parties, not to contest elections but to register with the IEC. Mr Mamabolo
explains that, upon registration with the IEC , a political party obtains a registration
certificate and its registration particulars , including its name, abbreviated name and
symbol are p ublished in the Government Gazette which are then afforded legal
recognition and protection.

[44] The effect of the registration of a political party with the IEC is to be gathered
from regulation 8 of the 2004 Regulations. Regulation 8 reads:


11 51 of 1996.
12 Regulations for the Registration of Political Parties, 2004, GN R217 GG 26058, 7 January 2004.

ZONDO CJ
23
“A party registered under these Regulations shall—
(a) be entitled to be represented on a party liaison
committee as contemplated in the Regulations on
Party Liaison Committees;
(b) have free access to any voters’ roll compiled and
maintained by the Commission; and
(c) be entitled to protection by the Commission of its
name, abbreviation of its name and distinguishing
mark or symbol.”13

[45] Mr Mamabolo goes on to point out that the fact that a political party has been
registered with the IEC does not entitle the party to contest an election. This is
important to note because it means that the requirement of 1 000 signatures with which
political parties must comply is not a requirement for contesting elections. It is a
requirement for the registration of a political party with the IEC. Once a political party
has been registered with the IEC and then wants to contest elections, there are
requirements for contesting elections with which the political party must comply. These
are provided for in section 27(1), (2), (3) and (4) of the Electoral Act. These are that:
(a) it must submit a list of its candidates and a declaration of
its candidates’ qualificat ion to stand for election and
acceptance of the nomination;
(b) a declaration of commitment by a duly authorised
representative and each of its candidates, to comply with
the Electoral Code of Conduct;
(c) it must pay the prescribed deposit for participati on in the
election; and
(d) political parties that are not yet represented in either a
provincial legislature or in Parliament are required,
additionally, to submit the signatures of registered voters

13 Regulation 8 of 2004 Regulations.
ZONDO CJ
24
equating to 15% of the relevant quota in the previous
election.

[46] Mr Mamabolo also points out that to register, every political party whether they
are already represented in one or other legislative body or not is required, among others,
to submit 1 000 supporting signatures of registered voters. He points out that, contrary
to OSA’s assertion that this is a repealed requirement, this requirement is currently
applicable. Mr Mamabolo also points out that OSA’s statement that this requirement
was prescribed in terms of section 27 of the Electoral Act is not correct. He states that
this requirement is prescribed in terms of section 15(3)(a) of the Electoral Commission
Act and regulation 3 of the 2004 Regulations.

[47] Mr Mamabolo also states that, in order to contest an election, political parties
which are not represented in one or other of the legislative bodies must prove their
capacity to participate successfully in the election by submitting signatures amounting
to 15% of the quota of the relevant region from the previous election and this
requirement also applies to independent candidates. In respect of political parties, the
requirement is provided for in section 27(2)(cB) whereas in respect of independent
candidates it is provided for in section 31B(3).

[48] Mr Mamabolo also states that the signature requirement for the registration of a
political party serves a different purpose to the signature requirement for contestation.
He says that OSA mistakenly conflates the two. He goes on to say:

“The signatures requirement on registration aims to ensure that the
associations seeking to be recognised and legally protected as political
parties convey their intention, and demonstrate the capacity, to operate
as a serious and well -organised party. While it serves a different
purpose to the requirement of a signature for contestation of an election,
both signature requirements are clearly complementary. Both signature
requirements ultimately prevent frivolous participation in elections.”

ZONDO CJ
25
[49] Mr Mamabolo also states that the trend indicates that the signature requirement
for the registration of a party – even when increased to 1 000 signatures – did not serve
as an indicator of voter support to justify a candidate as a serious contender in an
election. He sa ys that in the most recent election nearly three -quarters of participants
(political parties) were unsuccessful and forfeited their deposits. He points out that the
1 000 signature requirement also did not protect against the risk of an election with an
unmanageable number of contestants. He sa ys that this risk has been substantially
heightened with the inclusion of independent candidates as participants. This simply
has to be true. Whereas in the past it could, for example, be 24 or 48 political parties
which did not get even a single seat, with independent candidates it could even be
thousands of candidates who will contest the election if all they need is a number of
signatures which is not related in any way to the number of votes they must get in the
election in order to get one seat. One thousand signatures is such a number.

[50] Mr Mamabolo also points out that a secondary but not unimportant factor is the
cost of multi-page ballot papers. He says that the cost of the long ballot papers used in
the 2019 election came to about R35 million. He states that with the sharp rises in the
cost of paper since 2019, the inflation adjusted 2024 cost – which includes the
introduced third ballot – was projected and budgeted at R66 million. He says that that
takes into account the single-column ballot paper similar to the 2019 ballot, permitting
48 contestants. One only has to imagine what would happen if there were thousands of
independent candidates in the ballot who will contest elections because the threshold is
too low and bears no relationship to the number of votes they will have to get in the
elections to get one seat.

[51] Mr Mamabolo also makes the point that, if an independent candidate or political
party cannot secure 15% of the quota of a previous election, it means that the
independent candidate or political party is unable to show voter support of less than
one-fifth of the support he, she or it will need to win just one seat in the elections . He
then says that it is difficult to think how a person or political party that cannot get less
ZONDO CJ
26
than one-fifth of the votes required in order to obtain one seat before the elections can
manage to get five times that number of support in the elections.

[52] Mr Mamabolo points out that there is no international standard for parties and
individual candidates to prove their support before participating in an election. He says
that there is a wide disparity in the signature requirements adopted by different
countries, based on their own circumstances, needs and electoral systems many of
which are not based on proportional representation. He sa ys that the Commission
considers that Denmark provides a useful comparator as it has an electoral system akin
to South Africa’s, including the participation of independent candidates. Denmark
requires one signature for every 1 75 votes cast in the previous election. He points out
that in the 2022 elections in Denmark, the formula resulted in a requirement of 20 194
signatures, based on 3 535 952 votes cast in the previous election.

[53] Mr Mamabolo state s that, by comparison, in the 2019 provincial election in
KwaZulu-Natal a broadly similar number of 3 652 577 votes were cast. He then sa ys
that, if the Danish formula of one sign ature for every 175 votes were used, the
requirement would be 20 872 signatures. That, of course, is much more than is required
in South Africa for the election in 2024. Mr Mamabolo goes on to say that in Gauteng,
where a total of 4 537 402 votes were cast in the provincial election in 2019, the
requirement would be 25 928 signatures. That is, if we used the Danish formula of one
signature for every 175 votes.

[54] Mr Mamabolo states that South Africa’s signature requirement differ s from the
Danish model in that instead of taking into account all the votes cast, it sets the signature
requirement as a percentage of the level of support required to win a seat. That is a
percentage of a quota. Mr Mamabolo states that this approach results in a more
inclusive outcome with a much lower threshold of support required for participation.

[55] Mr Mamabolo points out that the number of signatures required under
section 31B(3) falls in the range of 10 000 to 14 000 for the National Assembl y and
ZONDO CJ
27
provincial legislatures depending on the region. He says that this means that an
independent candidate or a political party not yet represented in any of the legislative
bodies is only required to show that “it has the support of 15% – less than one-fifth – of
the number of votes that were required for a seat in the previous election ”.
Mr Mamabolo says 15% is less than one-fifth – it is actually less than one-sixth – of the
number of votes that were required for a seat in the previous election.

[56] Mr Mamabolo sets out the election outcomes from the 2004 election to the last
election in 2019. This helps to show how the numbers of political parties that were
allowed to contest elections but did not win any seats has been going up:
(a) In the 2004 election for the National Assembly 21 political parties
participated in the election but only 12 gained representation
while 9 did not gain representation . This means that just under
50% of the political parties that had been allowed to contest the
elections failed to get enough votes for even one seat in the
National Assembly.
(b) In the 2009 election for the National Assembly 26 political parties
participated in the election and 13 parties gained representation
and 13 did not. T his means that in 2009, 50% of the political
parties which were allowed to contest the elections could not get
enough votes to get simply one seat.
(c) In the 2014 election for the National Assembly 29 parties took
part in the elections but only 13% gained representations and 16
failed to gain even one seat. This meant that 55% of the parties
which took part in the elections did not get even one seat.
(d) In the 2019 election for the National Assembly , out of 48 parties
that participated in the elections, only 14 gained representation
and 34 did not gain representation. This meant that 70.8% of the
political parties allowed to contest elections failed to get even one
seat.
ZONDO CJ
28
(e) In the 2019 National Assembly and provincial legislature
elections, a combined 78 parties pa rticipated and 15 gained
representation in one or both legislatures and 63 did not gain
representation in either. This means that in the 2019 elections –
the most recent – there was an astronomical rise in the number of
political parties that were allowed to contest elections but that did
not get even one seat. More than 80% of the political parties
failed.

[57] Lastly, with regard to the Commission ’s evidence on the signature requirement
of 15% of the quota of the previous election, Mr Mamabolo says:

“39 The rationality and fairness of the requirement of 15% of the
quota of the previous election can also meaningfully be
assessed with reference to two data sets.
40 First, it can be evaluated by considering how many support
signatures a candidate would have to obtain per voting district
in each of the regions to meet (or exceed) the required total
number of supporting signatures. These are set out in the
following tables for the National Assembly elections (regional
tier) and the provincial elections respec tively. Column 1
reflects the number of supporting signatures required to contest
for a seat in the legislature per region; columns 2 to 4 show the
calculation of the signatures the candidate would need to obtain
per voting district in that region to meet (or exceed) that figure.
40.1 For the National Assembly (regional tier) election;
ZONDO CJ
29


40.2 For provincial legislatures:
Signatures
Required
Number of Voting
Districts
Signatures per
Voting District
Total
Signatures
Eastern Cape 4 672 4 869 1 4 869
Free State 4 826 1 582 4 6 328
Gauteng 8 757 2 799 4 11 196
KwaZulu-Natal 6 664 4 972 2 9 944
Limpopo 4 357 3 223 2 6 446
Mpumalanga 5 886 1 813 4 7 252
North West 4 213 1 737 3 5 211
Northern Cape 4 920 732 7 5 124
Western Cape 7 176 1 572 5 7 860

41 The average number of registered voters per voting district
currently stands at 1 120 prior to the registration weekends
before the 2024 election. Higher averages will apply in the
Signatures
Required
Number of
Voting Districts
Signatures per
Voting District
Total
Signatures
Eastern Cape 11 657 4 869 3 14 607
Free State 11 340 1 582 8 12 656
Gauteng 13 890 2 799 5 13 995
KwaZulu-Natal 13 045 4 972 3 14 916
Limpopo 11 329 3 223 4 12 892
Mpumalanga 11 925 1 813 7 12 691
North West 11 652 1 737 7 12 159
Northern Cape 10 271 732 15 10 980
Western Cape 13 201 1 572 9 14 148
ZONDO CJ
30
metropolitan and urban voting districts, while they will be
lower in rural areas.
42 Second, it can be evaluated by considering the number of
signatures req uired as a proportion of the pool of registered
voters from which support for participation can be obtained.
This is reflected in the following tables.
42.1 For the National Assembly (regional tier) election:


42.2 For provincial legislatures:
Signatures
Required
Registered Voters Signatures as a Portion
of Registered Voters
Eastern Cape 4 672 3 226 252 0,14%
Free State 4 826 1 394 838 0,35%
Gauteng 8 757 6 146 680 0,14%
KwaZulu-Natal 6 664 5 434 281 0,12%
Limpopo 4 357 2 612 185 0,17%
Signatures
Required
Registered Voters Signatures as a Portion
of Registered Voters
Eastern Cape 11 657 3 226 252 0,36%
Free State 11 340 1 394 838 0,81%
Gauteng 13 890 6 146 680 0,23%
KwaZulu-Natal 13 045 5 434 281 0,24%
Limpopo 11 329 2 612 185 0,43%
Mpumalanga 11 925 1 900 122 0,63%
North West 11 652 1 676 687 0,69%
Northern Cape 10 271 613 577 1,67%
Western Cape 13 201 3 092 488 0,43%
ZONDO CJ
31
Mpumalanga 5 886 1 900 122 0,31%
North West 4 213 1 676 687 0,25%
Northern Cape 4 920 613 577 1,8%
Western Cape 7 176 3 092 488 0,23%

43 The above figures indicate that it is reasonably possible for any
serious contender to collect the required number of signatures
to support their participation in an election.
43.1 When broken down into the number of signatures
required per voting district in the region, it becomes an
achievable requirement that will not obstruct serious
contenders from participating in an election.
43.2 Support for participation may, in some instances, not
be widespread in a region. Still, it is open to parties
and candidates to concentrate on areas with stronger
support to compensate for the lack of general support
or organisational infrastructure elsewhere in a region.
Moreover, if an independent candidate or
unrepresented party intends to contest in more than one
region, it only needs to focus on the region with the
highest number of signatures. If that threshold is not
met, the independent candidate or unrepresented party
may contest in the region or regions where the number
of signatures co llected meets the next highest quota
threshold.”

[58] Mr Mamabolo concludes by saying that , for the reasons given above , the
signature requirement of 15% of the quota of the relevant region for the previous
election is, in the Commission’s submission, a rational and justifiable requirement and
one that serves a legitimate government purpose.

ZONDO CJ
32
Analysis
OSA’s real complaint
[59] OSA is challenging the constitutional validity of the statutory requirement that
an independent candidate should obtain and produce supporting votes totalling 15% of
the quota of the relevant region from the previous election but a close examination of
OSA’s complaint reveals, in my view, that its real problem is not the 15%. It is the size
of the regions and, therefore, the size of the quotas. There is not one quota that applies
throughout.

[60] There are different regions or provinces with different quotas. 15% of the quotas
of some regions is low, for example, KwaZulu -Natal, and 15% of the quotas in some
provinces, for example, Gauteng, is high. Where the number of registered voters in a
region or province i s low, the quota will also be low and, therefore, the 15% will also
be a low figure. Where a region or province is big and the number of registered voters
in the region or province is high, the quota will also be high and that will mean that the
15% is also high. How high or low the number representing 15% of the quota will
depend on the size of the region or of the quota.

[61] The reason why OSA’s complaint cannot be with the 15% is that, if the 15% was
a 15% of a quota that is low, for example, 20 000, the 15% would translate to just about
3 000 about which I do not think OSA would complain. Another example is that, if the
quota was 100 in which case 15% thereof would be 15 supporting signatures of
registered voters, OSA could not complain. OSA could not complain about a
requirement that independent candidates and new political parties obtain 15 supporting
signatures of registered voters. However, when the 15% of a quota translates to 9 000
or 13 000 votes, OSA complains that the 15% requirement is a problem. However, their
problem is the s ize of the quota which in turn is based on the size of the region or
province. If the 15% were to be calculated on the basis of much smaller regions or
constituencies, OSA would have no problem with the 15% requirement. This proves
that the source of OSA’s true complaint is the size of the regions.
ZONDO CJ
33

[62] Mr Mogoale made this statement in his founding affidavit:

“The stated objective must however be understood in the context of the
legislative framework. For all intents and purposes provinces are
constituencies. Further, the signature requirements for independent
candidates are based on the quotas from those very regions from the
previous elections.”

This is found in paragraph 144 of OSA’s founding affidavit. In the first sentence of the
next paragraph, namely, paragraph 145, Mr Mogoale then says:

“I respectively submit – that the problem originates in and is caused by this
fact.”

The reference to “this fact” in this sentence is a reference to what he said in the previous
paragraph, namely paragraph 144.

[63] After stating in effect in the first sentence of paragraph 145 re ad with
paragraph 144 that “the problem originates in and is caused by” the fact that Parliament
decided that the provinces would be constituencies and that the signature requirement
for independent candidates (and new political parties) “a re based on the quotas from
those very regions from the previous election”, Mr Mogoale then continued in
paragraph 145 and said:

“Parliament elected or chose to have provinces as large constituencies. It
then made the signature requirement relative to the large vote thresholds of
these provincial regions or constituencies. The inescapable inference is
that the deliberate decision to use provinces as a unit of measurement for
the number of signatures required is arbitrary.”

[64] In the next paragraph, paragraph 146, Mr Mogoale states:

ZONDO CJ
34
“Not only does this approach lead to exorbitant numbers of signatures
being required, but the principle is out of kilter with the rest of our electoral
system. The previous 1 000 signature requirement had no link to the vote
threshold nor was it linked to the number of voters per province. It
represented an ordinary, feasible and reasonable measure.”

[65] OSA says that Parliament’s choice to have provinces as large constituencies and
to make the signature requirement relative to the “large vote thresholds of those
provincial regions or constituencies ” leads to “exorbitant numbers of signatures being
required”.

[66] That OS A’s real complaint is about the size of the constituencies that were
adopted by Parliament, namely the provinces , is also supported by Mr Mogoale’s
statements in paragraph 220 to 222 of OSA’s founding affidavit. In those paragraphs
Mr Mogoale says:

“220. While OSA does not ask this Court to step in the shoes of the legislature,
it does bring to this Court’s attention that an appropriate system should
be highest remainder system, because parties will be rewarded based
on their highest remaining votes keeps in step proportionality.
221. The allocation of the vacancy should ideally be dealt with by the normal
cause of a by -election. This proposal was declined by Parliament
during the Parliamentary process on the basis that it was not feasible
because regions which are provinces, are too large to conduct by -
elections.
222. However, I must point out that this conundrum is a direct consequence
of Parliament refusing to implement a constituency system that was
recommended by the majority of the Ministerial Advisory Committee.”
(Emphasis added.)

[67] The Minister’s response to paragraph 222 of OSA’s founding affidavit is in
paragraph 121 of his answering affidavit. It is this:

ZONDO CJ
35
“Indeed, the OSA reveals its true complaint at paragraph 222 of the
founding affidavit where it complains that the ultimate c ause of these
difficulties is Parliament’s failure to adopt a constituency-based system.
While the OSA's preference for such a system is noted, this is not a
basis to invalidate the Electoral Act.” (Emphasis added.)

[68] OSA is not asking this Court to set aside or invalidate Parliament’s decision to
use provinces as regions or constituencies . It is also not asking this Court to direct
Parliament to consider or decide to use any other places as constituencies on the basis
of which quotas will be calculated so that the 15% requirement will translate to lower
numbers than the numbers that one gets when 15% is used in relation to provinces. In
my view, OSA has acted wisely in not asking us to make those decisions.

[69] The consequence of the fact that OSA is not asking this Court to set aside or in
any way invalidate Parliament’s choice of provinces as constituencies or regions and
Parliament’s decision to base the 15% requirement on the quota of a region in the
previous election is that, as long as the level of support of the independent candidates
or new political parties needs to be tested, it can only be tested by way of a percentage
of the quota of the relevant region in the previous election. In other words, if there were
going to be one quota for all regions, one could fix a number as opposed to a percentage.

[70] If, for example, the quota for all regions or provinces were 30 000, one could fix,
for argument’s sake, 3 000 signatures which would be 10% of 30 000 as the number of
signatures that an independent candidate or a new political party would have to obtain
in order to contest elections on the basis that it would be an indication of adequate voter
support. However, if the quotas for different regions or provinces differ according to
the size of a region or province, whether or not an independent candidate or a new
political party has credible support to contest elections can only be determined on the
basis of a percentage of voter support in relation to the quota of a particular region or
province. You cannot fix one number such as 1 000 for all regions when the regions
have vastly different sizes. A one-size-fits-all approach cannot work.

ZONDO CJ
36
[71] Fixing one number such as 1 000 signatures when the regions have vastly
different sizes of voter populati ons simply cannot serve the purpose of the
15% requirement. This is so because 15% is 15% of the quota of each region but 1 000
may be enough to show adequate support for a region whose quota is 10 000 but, once
a region has a quota of 44 000, a requirement of 1 000 signatures cannot conceivably
be said to be an indication that such an independent candidate or new political party has
credible prospects of gaining a seat in elections. Therefore, in my v iew requiring an
independent candidate or new political party to obtain 1 000 signatures on the basis that
1 000 signatures would be an indication that he , she or it has credible prospects of
gaining one seat in elections is irrational. It would not serve any government purpose.

[72] OSA has not anywhere in its founding affidavit said why 15% of the quotas
based on the previous elections constitutes a barrier. One may have understood if the
position was that, for example, there were two months left before the elections and OSA
was saying there was too little time left to collect signatures from so many people. OSA
does not anywhere say that there is little time left to collect so many signatures. When
OSA launched its application – which was in May 2023 – there was still about at least
a year before the 2024 election s. The question is: why would even six months before
the elections not be enough to collect 4 000 signatures or 9 000 signatures or
13 000 signatures? An independent candidate will have campaign workers or
supporters. If he or she has 10 campaign staff, there is no reason why each one of them
cannot secure 2 000 signatures per month which would be 500 per week which would
mean 20 000 signatures per month which would be 40 000 in two months.

[73] Another point is this: OSA does not seek the reduction of the size of the quotas
or the quota of the regions or provinces on which the 15% is based. It seeks a change
from the use of a percentage of a quota to a fixed number, namely, 1 000 in order for an
independent candidate to be eligible to contest an election but it has not shown why
Parliament’s decision to use a percentage of a quota is bad or is irrational. There is,
therefore, no justification to change from the use of a percentage of a quota to a fixed
number. This was Parliament’s judgment call and there is no basis to interfere with it .
ZONDO CJ
37
OSA has not shown any basis on which it would be justified to effectively set aside
Parliament’s decision to use a percentage and not a number.

[74] Lastly, I wish to make this point. The 15% is 15% of the number of the registered
voters that the independent candidate will need to achieve in the elections in order to
secure one seat. If OSA says 15% of the quota of the region in the previous election is
too high because it will be 9 000 or 14 000 votes and says the requirement should be
1 000 signatures, we must remember that the quota will remain the same. It will not be
reduced if the requirement is altered to 1 000. If the requirement remained at 15% of
the quota and an independent candidate secured 15% before contesting the election,
then later – during the election campaign – that independent candidate would not need
to go back and canvass the people whose signatures he or she already secured. In that
event, the candidate would only need to win over the remaining 85% of the registered
voters in the relevant region to reach the quota.

[75] In other words , an independent candidate who complied with the 15%
requirement at the beginning would have an easier job thereafter than an
independent candidate who obtained only 1 000 signatures at the be ginning and has to
obtain, maybe, 90% of the quota later. That is if 1 000 signatures constitute 10%. If
1 000 signatures constitute 2% of the quota, that would mean the independent candidate
who secures only 1 000 signatures to contest the elections, ha s to work hard to secure
98% of the quota during the election campaign. The question that arises is: what logical
basis can there be for thinking that the same candidate has credible prospects of getting
in the election with 98% of the votes that he or sh e needs in order to get one seat? If
there was a reward offered for a student who would get 100% in a particular subject in
the final matric examination, nobody would ever think that a student who could not get
15/100 in the trial examinations in August or September would be able to suddenly get
100% in the final examination in November and get the promised reward. So, it seems
to me that anyone who cannot get 15% of the qu ota over a year before the election is
highly unlikely to be able to get 100% of the required votes in the election for one seat
in a few months’ time.
ZONDO CJ
38

[76] I think that we can take judicial notice of South Africa’s world famous race,
namely, the Comrades Marathon which is r un between Durban and Pietermaritzburg.
The distance is about 90 km. It is a very popular race and there are always too many
people who want to run the Comrades Marathon than numbers that the organisers can
accommodate. 15% of the distance of 90 km is about 13,5 km. If you wanted to enter
the Comrades Marathon and the organisers asked you to run a distance of 13,5 km in
order to be eligible and you said that that was too much, would they allow you to enter
the Comrades Marathon? I don't think so.

[77] The point would be that the distance you will have to run to fin ish the race is
90 km and, if you cannot run a distance of 13,5 km, you have no chance of successfully
running 90 km on the day of the race. Here, too, if over a year before elections, an
independent candidate cannot obtain the support of registered voters amounting to 15%
of the votes he or she will need in the election in order to get one seat, they have no
chance of getting 100% of the votes they will need. I go back to saying that the source
of OSA’s problem is not the 15% but the size of the reg ions and the size of the quotas.
I now turn to deal with the complaint that OSA says in its founding affidavit is its chief
complaint.

OSA’s declared chief complaint
[78] As stated earlier , OSA says in its founding affidavit that its chief complaint is
that the EAA requires both the independent candidates and the political parties which
are not yet represented in the National Assembly and provincial legislatures to obtain
the same number of signatures in order to be allowed to contest elections. OSA goes
on t o say that, although the fact that the signature requirement referred to in
section 31B(3) applies to both the independent candidates and new political parties
meant that it was a “formally equal requirement”, “this results in a substantively unequal
outcome, creates an unfair barrier to entry for independent candidates and goes against
the purpose of New Nation II ”. The reference to New Nation II is a reference to
this Court’s judgment in New Nation Movement. The passages of this Court’s judgment
ZONDO CJ
39
in New Nation Movement which Mr Mogoale quotes in OSA’s founding affidavit do
not contain anything that supports the statement that the fact that the signature
requirement applies to both ind ependent candidates and political parties is contrary to
the purpose of that judgment.

[79] OSA did not anywhere in its founding affidavit say why it was complaining
about the fact that the signature requirement applied to both the independent candidates
and political parties. All it said was that the fact that the signature requirement applied
to both independent candidates and political parties “creates an unfair barrier to entry
for independent candidates and goes against the purpose of New Nation II”. There was
no elaboration provided as to how that fact created a barrier and, if it created a barrier,
why OSA contended that it was unfair. At the hearing I said to OSA’s Counsel that I
could not see any substantiation in OSA’s founding affidavit of its bald assertions about
its complaints and I asked her to indicate to us where we could find substantiation of
the general assertions in the founding affidavit. She could not point to any.

[80] OSA drew a table which it said reflected the real numbers which would constitute
15% of the quota from the previous election in the different regions. I n the different
provinces the figures were somewhere between 10 271 to 13 890 signatures in respect
of the National Assembly, depending on the size of the province, and 4 213 to 7 176
signatures in respect of the provinces depending on the size of the province. One would
have thought that OSA ’s purpose in indicating what 15% would translate to , in real
numbers, of people was to then elaborate by saying why independent candidates would
not be able to achieve those numbers but OSA did not do this in its founding affidavit.
So, OSA did not say that there would be any logistical or practical difficulties that an
independent candidate could not reasonably overcome. The Court should not assume
practical or logistical difficulties that independent candidates could possibly have which
OSA has chosen not to articulate. It is difficult to think how anybody could contend
that it would be difficult for any independent candidate worth his or her salt to obtain
around 4 000 or 4 500 or so signatures to obtain a seat in a provincial legislature.

ZONDO CJ
40
[81] OSA also contended that Parliament should have required
independent candidates to obtain 1 000 signatures because in the past, as OSA
understood the position , political parties had been required to obtain 1 000 signatures
as one of the requirement s for contesting elections. OSA contended that , if the
requirement of 1 000 signatures served its purpose in respect of political parties, there
is no reason why it would not serve its purpose in respect of independent candidates.
OSA contended that Parliament should have made the requirement of 1 000 signatures
applicable to independent candidates.

[82] I did not understand this contention by OSA to mean that independent candidates
and political parties would have to be required to obtain different numbers of signatures.
I understood OSA to suggest that both independent candidates and political parties
should obtain the same number of signatures. However, the difficulty with this is the
fact that OSA did not challenge the constitutional validity of section 27 of the amended
Electoral Act which makes the signature requirement applicable to political parties not
represented in the National Assembly or in any provincial legislature as yet. This means
that, if OSA succeeded in its challenge, independent candidates would be required to
obtain 1 000 signatures whereas new political parties would be required to obtain 15%
of the quota of registered voters i n a region. OSA has not sought to justify this
differentiation.

[83] Reference has been made above to Mr Mamabolo’s evidence that over the years
there has been a sharp increase of political parties which participate in elections but do
not get even one seat. This happened even during the period when 500 signatures were
required for the registration of a political party. It is quite clear from this evidence, that
the requirement of 500 signatures of registered voters that used to be required for the
registration before the 1 000 signature requirement which is half of 1 000 signatures
now required for the registration of political parties with the IEC – has not been effective
in preventing political parties that have no credible prospects of obtaining even just one
seat in the elections from being registered with the IEC. The requirement of 1 000
signatures is unlikely to make any difference.
ZONDO CJ
41

[84] The Minister, Parliament and the Commission contend that requiring
independent candidates and political parti es to obtain only 1 000 signatures would not
serve the purpose which the requirement of 15% of the quota of the previous election is
meant to serve. They point out that OSA got its facts wrong with regard to the purpose
of the requirement of 1 000 signatures. They state that the 1 000 signature requirement
applicable to political parties relates to registration and not to contesting elections
whereas the requirement of 15% signatures of the quota of the previous election relates
to contesting elections.

[85] OSA did not deny that the requirement of 1 000 signatures applicable to
political parties relate s to registration and not to contesting elections. Indeed, the
Minister, Parliament and the IEC actually point out that, contrary to OSA ’s suggestion
in its founding affidavit that the 1 000 signature requirement used to apply but is no
longer applicable now, this requirement is still one of the legal requirements for the
registration of political parties with the IEC. They point out that the requirements with
which new political parties are required to comply in order to contest elections are
separate and do not include any requirement for 1 000 signatures.

[86] OSA did not file any affidavit that questioned or challenged or d isputed the
explanation given by Parliament, the M inister and the IEC that the requirement of
1 000 signatures is not a requirement for contesting elections. Even if one were to say
that the requirement of 1 000 signatures is intended to gauge some support for a political
party just like the requirement of 15% signatures of a quota from a previous election,
there is a distinction between the two which would strongly militate against the use of
the requirement of 1 000 signatures in the place of the 15% signatures requirement to
test the serio usness of an independent candidate or of a political party to conte st
elections and to discourage frivolous candidates.

[87] It is legitimate to exclude from contesting elections anyone and any political
party that has no credible prospects to win even one seat during the elections. There is
ZONDO CJ
42
nothing unconstitutional about that. Therefore, it is quite legitimate for Parliament to
legislate a requirement such as a certain percentage of supporting signatures of
registered voters which reflects that such an individual or political party has a credible
chance of obtaining in the election enough votes for a seat. In this regard it m ust be
borne in mind that different people will have different views as to what percentage
would represent a credible prospect for such an individual or political party to obtain a
seat. Accordingly, some deference should be shown to the percentage determ ined by
Parliament unless such percentage is irrational. In this regard the Court should be slow
to impose its own view of what an appropriate percentage should be.

[88] OSA states:

“OSA believes that the amendment Act unjustifiably arbitrarily and
disproportionately sets an independent candidate’s requirement at 15%.
It further submits that the requirement does not fulfil any legitimate
government purpose and that less restrictive means such as the original
1000 signature requirement would achieve the same objective.”

Counsel for OSA conceded, in my view correctly, that the 15% signature requirement
does serve a legitimate government purpose . That is to serve as an indication that the
independent candidate is serious about contesting the election and to eliminate frivolous
contestants. With regard to the first part of the above excerpt I wish to point out that
OSA simply tells us that it believes that the EAA “unjustifiably, arbitrarily and
disproportionately sets an independent candidate at 15%” but it does not provide the
factual basis for its belief nor does it substantiate its statement.

[89] It is trite that in motion proceedings the affidavits serve as both pleadings and
evidence. They define the issues between the parties and provide the evidence.14 This
Court has held that a party that challenges the constitutionality of a provision in a statute
must do so when the legal proceedings are instituted and must lay a proper foundation

14 Molusi v Voges N.O. [2016] ZACC 6; 2016 (3) SA 370 (CC); 2016 (7) BCLR 839 (CC) at para 27 and
Prince v President of the Law Society of the Cape of Good Hope [2002] ZACC 1; 2002 (2) SA 794 (CC);
2002 (3) BCLR 231 (CC) at para 22.
ZONDO CJ
43
for such a challenge in the pleadings. It must place before the court all information
relevant to the determination of the constitutionality of the impugned provisions. This
is necessary to alert the other side to the case it is called upon to meet and to enable the
latter to present factual evidence and legal argumen t to oppose that case so as to leave
no doubt about the nature of the matter, the grounds on which it is brought and the relief
sought.

[90] In its statement OSA also does not identify the things in relation to which the
setting of the 15% signature requirement is disproportionate. For that reason it is
difficult to follow this part of its case. More is required. Something does not become
unjustifiable or arbitrary just because someone says it is unjustifiable or arbitrary. OSA
has failed to substantiate these statements or beliefs or contentions.

[91] The case made out in the founding affidavit is this. OSA’s “chief complaint” is
that the EAA requires political parties and independent candidates to obtain the sa me
number of signatures to register as contestants in the elections, namely 15% of the quota
in the National Assembly or provincial legislature. OSA contends that on its face, this
is a formal equal requirement, but results in a substantively unequal outc ome and
“creates an unfair barrier to entry for independent candidates”. OSA then quotes the
number of signatures required in each province in order to contest the elections for the
National Assembly and provincial legislatures. OSA then points that, before the EAA
came into force, political parties were required to submit 1 000 signatures to register as
a party with the IEC. It then states that t he only reasonable inference is that 1 000
signatures were sufficient to show that political parties were ser ious about contesting
elections. OSA then contends that the EAA “unjustifiably, arbitrarily, and
disproportionately sets an independent candidate ’s requirement at 15% ”, which does
not fulfil any legitimate government purpose. It adds that the original 1 000 signature
requirement would do so.

[92] However, OSA has not put up any facts or evidence in support of its c ontention
that the 15% signature requirement results in an unequal outcome. Neither is there any
ZONDO CJ
44
evidence of the respects in which this requirement creates an unfair barrier for
independent candidates to contest the elections. The same is true for OSA’s assertion
that the signature requirement is arbitrary and disproportionate. There is not a shred of
evidence that it is impossible to meet the signature requirement or that an
independent candidate attempted to do so but failed. On the IEC’s evidence before us
in some province or provinces the 15% signature requirement would require a candidate
to obtain 4 000 signatures. OSA has not said why a serious candidate would not be able
to obtain 4 000 signatures. He or she might have three months to obtain
4 000 signatures. What one is then left with is nothing more than OSA’s say -so as to
why the signature requirement is a barrier.

[93] After making the statements discussed immediately above, OSA submit s that
three rights of independent candidates were “directly and detrimentally affected by the
15% signature requirement”. It stated that these were:
(a) the right of every citizen in section 19 of the Constitution to make political
choices which includes the right to form a political party and to participate
as an independent candidate;
(b) the right provided for in section 19(3)(b) of the Constitution in fav our of
every adult citizen to stand for public office and, if elected, to hold office;
(c) the right to associate or , by extension, not to associate with the
political party system by running as an independent candidate;
(d) an independent candidate’s right to dignity.

[94] After specifying the above as the rights that are “directly and detrimentally
affected by the 15% requirement”, OSA then states:

“Therefore, any requirements that are imposed on independent
candidates to contest elections as prescribed in section 31B should be
balanced in the context of these political rights and ought not to be
treated as a gate-keeping mechanism nor as a barrier to entry.”

ZONDO CJ
45
OSA makes this statement about balancing the 15% signature requirement with the
political rights referred to above but does not say why it contends that the EAA, as it
stands, does not achieve the required balance. After making the statement that it makes
about balancing the 15% signature requirement with the political rights referred to
above, OSA the n proceeds to say that the next question is “whether the limitation is
unjustifiable in terms of section 36(1) of the Constitution ”. OSA then says: “This
question comes down to whether there are less restrictive means to achieve the purpose
which is ultimately a question about threshold”.

[95] The first question that this Court is called upon to determine with regard to the
signature requirement is whether section 31B(3) of the EAA limits or infringes the right
to stand for public office, the right to vote and the right to free and fair elections
entrenched in section 19 of the Constitution. If section 31B(3) limits one or more of
these rights, the next question will be whether the limitation is reasonable and justifiable
in a democratic society based on freedom, equality and human dignity as contemplated
in section 36 of the Constitution.

OSA’s statement that three of independent candidates’ rights have been detrimentally
affected
[96] In its founding affidavit OSA states that the right to make political choices which
includes the right to form a political party and to participate as an independent
candidate, the right of every adult citizen to stand for public office and, if elected, to
hold office, the right to associate or the right not to associate with the pol itical party
system by running as an independent candidate and an independent candidate’s right to
dignity are “directly and detrimentally affected” by the 15% signature requirement.
OSA did not substantiate this assertion at all. It did not provide any reasons why it says
that the 15% signature requirement has detrimentally affected any of their rights.

[97] OSA is the applicant here. It must convince the Court that the
15% signature requirement has detrimentally affected these rights. It cannot do so by
simply making an unsubstantiated statement that the 15% signature requirement has
ZONDO CJ
46
detrimentally affected any of the rights it identified. OSA is the one that has come to
this Court to seek relief. It must substantiate its case. In regard to this statemen t OSA
has failed dismally to substantiate its case. I will return to this point later when I deal
with the test for determining whether a regulatory provision of a statute limits an
entrenched right as articulated by this Court in Affordable Medicines.15

[98] Although the Constitution contains the Bill of Rights, it does not provide any
detailed provisions that may be necessary for the effective exercise of those rights or
for the implementation of those rights. The Constitution leaves the task of providing
such detailed provisions to Parliament. Parliament is, therefore, enjoined to make such
detailed provisions. In New National Party, this Court inter alia said:

“The Constitution recognises that it is necessary to regulate the exercise
of the right to vot e so as to give substantive content to the right”. 16
(Emphasis added.)

In the same case this Court said:

“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.”17

The regulation and limitation of rights
[99] Statutory provisions that affect rights entrenched in the Bill of Rights fall into at
least two categories. The one category would be a provision that constitutes a complete

15 Affordable Medicines Trust v Minister of Health [2005] ZACC 3; 2006 (3) SA 247 (CC); 2005 (6) BCLR
529 (CC).
16 New National Party v Government of the Republic of South Africa [1999] ZACC 5; 1999 (3) SA 191 (CC);
1999 (5) BCLR 489 (CC) at para 13.
17 Id at para 14.
ZONDO CJ
47
or total denial of such a right. Another category is one where the provision permits the
exercise of th e right or prohibits or preclud es it s exercise only conditionally or
temporarily. An example of a statutory provision that constitutes a total denial or a
complete prohibition is section 65(1) of the Labour Relations Act (LRA).18

[100] Section 65 falls within a chapter that deals with strikes. What section 65(1) does
must be viewed against the background that section 23(2)(c) of the Constitution confers
on every worker the right to strike. Section 65(1) reads:

“Limitations to right to strike and recourse to lock-out
(1) No person may take part in a strike or a lock -out or in any
conduct in contemplation or furtherance of a strike or a
lock-out if—
(a) that person is bound by a collective agreement that
prohibits a strike or lock -out in respect of the issue in
dispute;
(b) that person is bound by an agreement that requires the
issue in dispute to be referred to arbitration;
(c) the issue in dispute is one that a party has the right to
refer to arbitration or to the Labour Court in terms of
this Act;
(d) that person is engaged in—
(i) an essential service; or
(ii) a maintenance service.”

[101] Even the heading to section 65 reflects that indeed section 65 is a limitation of
the right to strike or recourse to lock -out. The heading reads: “ Limitations on right to
strike or recourse to lock-out”.


18 66 of 1995.
ZONDO CJ
48
[102] In other words, whereas section 23(2)(c) of the Constitution confers upon every
worker the right to strike, section 65(1) precludes certain categories of workers from
exercising such a right. Those workers include those employed in essential services
and maintenance services. In New National Party this Court repeatedly referred to a
denial of a right as an infringement of a right.19 One then has those statutory provisions
which govern or regulate the exercise of such rights. In the context of the LRA, an
example would be section 64 thereof. These provisions permit workers to exercise the
right to strike but, only if certain conditions or requirements are met or complied with.

[103] There is no heading to section 64. Section 64 reads:

“(1) Every employee h as the right to strike and every employer has recourse to
lock-out if—
(a) the issue in dispute has been referred to a council or to the Commission
as required by this Act, and—
(i) a certificate stating that the dispute remains unresolved has
been issued; or
(ii) a period of 30 days, or any extension of that period agreed to
between the parties to the dispute, has elapsed since the
referral was received by the council or the Commission; and
after that—
(b) in the case of a proposed strike, at least 48 hour s’ notice of the
commencement of the strike, in writing, has been given to the
employer, unless—
(i) the issue in dispute relates to a collective agreement to be
concluded in a council, in which case, notice must have been
given to that council; or
(ii) the employer is a member of an employers' organisation that
is a party to the dispute, in which case, notice must have been
given to that employers' organisation; or

19 See New National Party above n 15 at paras 18, 20, 37, 46 and 47.
ZONDO CJ
49
(c) in the case of a proposed lock -out, at least 48 hours ’ notice of the
commencement of the lock-out, in writing, has been given to any trade
union that is a party to the dispute, or, if there is no such trade union,
to the employees, unless the issue in dispute relates to a collective
agreement to be concluded in a council, in which case, not ice must
have been given to that council; or
(d) the case of a proposed strike or lock -out where the State is the
employer, at least seven days ’ notice of the commencement of the
strike or lock -out has been given to the parties contemplated in
paragraphs (b) and (c).
(2) If the issue in dispute concerns a refusal to bargain, an advisory award must
have been made in terms of section 135(3)(c) before notice is given in terms
of subsection (1)(b) or (c). A refusal to bargain includes—
(a) a refusal—
(i) to recognise a trade union as a collective bargaining agent; or
(ii) to agree to establish a bargaining council;
(b) a withdrawal of recognition of a collective bargaining agent;
(c) a resignation of a party from a bargaining council;
(d) a dispute about—
(i) appropriate bargaining units;
(ii) appropriate bargaining levels; or
(iii) bargaining subjects.”

It will be seen that provisions under section 64(1) and those under section 65(1) do not
fall into the same category. I have no doubt that those of section 64 regulate the right
to strike but those of section 65 limit the right to strike . However, even if those of
section 65 may also be said to regulate the right to strike, they are not the same as the
provisions in section 64, do not serve the same purpose and cannot be treated the same.

[104] In fact section 23(5) of the Constitution appears to recognise that a provision that
regulates a right entrenched in the Bill of Rights may or may not limit such a right. That
is why it provides that national legislation m ay be enacted to regulate collective
ZONDO CJ
50
bargaining. That comes after a sentence that refers to “the right to engage in collective
bargaining”. Then section 23(5) provides: “To the extent that the legislation may limit
a right in this chapter, the limitation must comply with section 36(1)”. This suggests a
recognition that a statutory provision that regulates a right may or may not limit such
right. Where a statutory provision totally or completely prohibits or outlaws the
exercise of a right or where it co nstitutes a complete denial of a right ; it constitutes a
limitation of the right and, therefore, the inquiry would focus on whether the limitation
is reasonable and justifiable in terms of section 36(1).

[105] The jurisprudence of this Court does reflect that a statutory provision that merely
regulates a right entrenched in the Bill of Rights does not limit such a right . Such a
provision must do much more before it can be said to limit a right entrenched in the Bill
of Rights. It must be remembered that the purpose of the LRA includes:

“[T]o advance economic development, social justice, labour peace and the
democratisation of the workplace by fulfilling the primary objects of this
Act, which are—
(a) to give effect to and regulate the fundamental rights conferred by
section 27 of the Constitution”20

[106] In POPCRU21 this Court held, through Nkabinde J, that “[t]he LRA was enacted,
among other things, to regulate the right to strike in conformity with the Constitution”.
Islamic Unity Convention22 related to the right to freedom of expression as entrenched
in section 16 of the Constitution. In that case this Court said about regulation:

“There is accordingly no bar to the enactment of legislation that
prohibits such expression. Any regulation of expression that falls within

20 Section 1(a) of the LRA.
21 South African Police Service v Police and Prisons Civil Rights Union [2011] ZACC 21; 2011 (6) SA 1 (CC) ;
2011 (9) BCLR 992 (CC).
22 Islamic Unity Convention v Independent Broadcasting Authority [2002] ZACC 3; 2002 (4) SA 294 (CC);
2002 (5) BCLR 433 (CC).
ZONDO CJ
51
the categories enumerated in section 16(2) would not be a limitation of
the right in section 16.” (Emphasis added.)

Soon thereafter, this Court said:

“Where the State extends the scope of regulation beyond expression
envisaged in section 16(2), it encroaches on the terrain of protected
expression and can do so only if such regulation meets the justification
criteria in section 36(1) of the Constitution.” (Emphasis added.)

It seems to me that what this Court had in mind here is where a provision goes beyond
mere regulation.

[107] In Garvas23 one of the issues that this Court had to consider was whether
section 11(2) of the Regulation of Gathering s Act24 limited the right to freedom of
assembly and, if it did, whether the limitation was justifiable. In the course of its
consideration of that issue, this Court said through Mogoeng CJ:

“[54] The long title of the Act states that the purpose of the Act is ‘(t)o
regulate the holding of public gatherings and demonstrations at certain places;
and to provide for matters connected therewith’. The Act requires the
appointment of persons responsible for giving and receiving notices to hold
gatherings and to act at consultations or negotiations in relation to the holding
of gatherings on behalf of the organisers, the police and the local authority
involved.
[55] The mere legislative regulation of gatherings to facilitate the
enjoyment of the right to assemble peacefully and unarmed, demonstrate,
picket and petition may not in itself be a limitation. Section 11(2), read with
section 11(1), goes further than simply to regulate the exercise of the right in
order to facilitate its full and appropriate enjoyment by those who organise
and those who participate.” (Emphasis added.)

23 South African Transport and Allied Workers Union v Garvas [2012] ZACC 13; 2013 (1) SA 83 (CC); 2012 (8)
BCLR 840 (CC).
24 205 of 1993.
ZONDO CJ
52

This excerpt reflects that this Court held in Garvas that a regulation has to go beyond
regulation before it can be said to constitute a limitation.

[108] Mogoeng CJ went on to say in the Garvas judgment:

“[56] Section 11(1) holds organisers of a gathering liable for riot
damage subject to section 11(2), which provides a limited defence to a
claim of this kind. The effect of these specific provisions, in the context
of the Act as a whole, is to render holders of a gathering organised with
peaceful intent liable for riot damage on a wider basis than is provided
for under the law of delict. This is all the more so, given the extremely
wide definition of riot damage in the Act. This means that pr oof of
liability will, as indicated earlier, be easier in a large number of cases.
[57] Compliance with the requirements of section 11(2)
significantly increases the costs of organising protest action . And it
may well be that poorly resourced organisations that wish to organise
protest action about controversial causes that are nonetheless vital to
society could be inhibited from doing so. Both these factors amount to
a limitation of the right to gather and protest.” (Emphasis added.)

[109] So, in Garvas this Court held a statutory provision that regulated the exercise of
the right to assemble to constitute a limitation of that right because it found the provision
to go beyond regulating the exercise of the right. The statutory provision was held to
go beyond mere regulation because of the civil liability that was placed on the leaders
or convenors of a public gathering as well as increased costs of organising protests. Had
it not been for that the statutory provision would have been found to cons titute a
limitation. In the present case OS A has not cited cost as a factor in objecting to the
signature requirement. Furthermore, in the present case there is no criminal or
civil liability associated with the signature requirement.

ZONDO CJ
53
[110] In Mlungwana25 this Court had to consider whether the criminali sation by
section 12(1) of the Regulation of Gatherings Act of the failure of the convener of a
gathering of 50 normal persons to give notice or adequate notice limited the right
entrenched in section 17 of the C onstitution peacefully and unarmed, to assemble, to
present petitions and, if it did, w hether such limitation was reasonable and justifiable
under section 36(1) of the Constitution.

[111] In Mlungwana this Court, through Petse AJ, said:

“A convener can be held liable for any riot damage caused by a
gathering or demonstration. This liability is civil in nature. The
convener is presumed to have acted unreasonably if riot damage occurs
as a result of the gathering, but this presumption is rebuttable. If the
convener can show – in essence – that the riot damage was not
reasonably preventable and foreseen, then they can avoid liability.”

This Court rejected a contention that section 12(1) was not a limitation but a mere
regulation. It said:

“[46] The respondents’ argument is unsustainable. Section 12(1)(a)
goes beyond mere regulation. In Garvas, this Court considered
whether section 11(1) and (2) of the Act – which provides for the civil
liability of a convener for riot damage – constituted a lim itation of
section 17. This Court held that “mere regulation” would not
necessarily amount to a limitation of the section 17 right. But the
increased cost of organising protest action and the deterrent effect of
the civil liability did amount to a limitation. Thus, this Court found that
deterring the exercise of the right in section 17 limits that right. The
reason is obvious. Deterrence, by its very nature, inhibits the exercise
of the right in section 17. Deterrence means that the right in question
cannot always be asserted, but will be discouraged from being
exercised in certain instances.

25 Mlungwana v State [2018] ZACC 45; 2019 (1) SACR 429 (CC); 2019 (1) BCLR 88 (CC).
ZONDO CJ
54
[47] In this matter, the criminal sanction in section 12(1)(a) deters
the exercise of the right in section 17. The respondents not only admit
this, but invoke th e self -same deterrent effect to explain
section 12(1)(a)’s purpose and justify its provisions. The possibility of
a criminal sanction prevents, discourages, and inhibits freedom of
assembly, even if only temporarily. In this case, an assembly of 16
like-minded people cannot just be convened in a public space. The
convener is obliged to give prior notice to avoid criminal liability. This
constitutes a limitation of the right to assemble freely, peacefully, and
unarmed. And this limitation not only applies to conveners, but also to
all those wanting to participate in an assembly. If a convener is deterred
from organising a gathering, then in the ordinary course (save for the
rare spontaneous gathering) a gathering will not occur.”
(Emphasis added.)

[112] I pause here to highlight the fact that both in Garvas and Mlungwana this Court
held that a statutory provision that regulates a right entrenched in the Bill of Rights will
only constitute a limitation of that right if it goes beyond mere regulation. The two
judgments may be interpreted to mean that a statutory provision that regulates a right in
the Bill of Rights will only constitute a limitation if it adversely affects that right or if
it deters somebody from exercising that right. In Garvas what led this Court to conclude
that the statutory provision in issue in that case that sought to regulate the right to
assemble constituted a limitation of the right of assembly was the increased costs
associated with convening a gathering as well as the civil liability for riot damage.

[113] I now turn to discuss Affordable Medicines.26 Because of the importance of this
case, I propose to deal with it in some detail. This is so because it shows, among other
things, even what regulatory provisions may deal with. In Affordable Medicines
this Court had to consider whether certain statutory provisions unjustifiably infringed
or limited the right entrenched in section 22 of the Constitution. Section 22 of
the Constitution reads:

26 Affordable Medicines above n 14.
ZONDO CJ
55

“Every citizen has the right to choose their trade, occupation or
profession freely. The practice of a trade, occupation or profession may
be regulated by law.” (Empahasis added.)

[114] Government introduced a licencing scheme by way of legislatio n in terms of
which medical practitioners needed a licence if they wanted to compound and dispense
medicines. The scheme also regulated the premises from which medicines were to be
dispensed by licenced medical practitioners and dentists.

[115] Section 22(c)(1)(a) of the Medicines and Related Substances Act, as amended 27
(Medicines Act) empowered the Director -General to issue licences to medical
practitioners “on the prescribed conditions”. Regulation 18(3) of the Regulations made
under the Medicines Act ,28 set out information that needed to be contained in an
application for a licence while regulation 18(5) set out factors to which the
Director-General had to have regard when considering an application for a licence. The
issuing of a licence was subject to, among other requirements, the successful completion
of a supplementary course determined by the South African Pharmacy Council after
consultation with the South African Nursing Council.

[116] The licencing scheme was directed at addressing bad dispensing and
compounding practices and their consequences which arose from the fact that, prior to
this licencing scheme, the compounding and dispensing of medicines by medical
practitioners and other h ealth practitioners, with the exception of pharmacists, were
either not adequately regulated or not regulated at all. There were no standards, norms
or guidelines to ensure that dispensers of medicines adhered to good dispensing and
compounding practices. The old legislative framework did not prohibit practices such
as pharmaceutical companies giving incentives to medical practitioners nor did they
prohibit practitioners from selling on samples they had received for free from

27 101 of 1965.
28 Regulation 18(3) of the General Regulations made in terms of the Medicines and Related Substances Act, 1965
(Act 101 of 1965) GNR 844 GG 26572, 16 July 2004.
ZONDO CJ
56
pharmaceutical companies. Th is created a conflict of interest between the dispensing
medical practitioner and their patients.

[117] The underlying objective behind the scheme was to increase access to medicines
that were safe for consumption by the public. This was to be achieved by, among other
things, ensuring that healthcare practitioners who dispensed and compounded
medicines were adequately trained in good dispensing practice and maintaining high
standards in the safe and proper storage, labelling, handling and keeping of medicines.
To this end, the respondents in the Affordable Medicines case said that the sale of
medicines, their suitability, the standard of dispensing, the suitability of premises where
medicines were kept and the conditions under which they were kept had to be properly
regulated.

[118] One of the constitutional challenges mounted by the applicants in the
Affordable Medicines case was that, in so far as regulation 11 required that a licence be
“coupled” to specific premises from which medicines would be compounded and
dispensed by a medical practitioner or dentist, it (i.e. regulation 11) fell outside the
purview of section 22 of the Constitution. Section 22 permitted only the practice of a
profession to be regulated by law. They also contended that “coupling” violated other
rights in the Bill of Rights.

[119] This Court recorded the conclusions of the High Court in Affordable Medicines.
The High Court found that the licensing scheme was introduced by the government in
order to achieve its objective of increasing access to medicines that were safe for
consumption. The High Court found that this was a legitimate purpose to pursue. It
held that the Minister did not exceed her powers when making regulation 11 which
linked a licence to compound and dispense medicines to specific premises.
The High Court concluded that the Minister had not breached the principle of legality.
It also held that the licensing scheme did no more than regulate the practice of
dispensing medicines within permissible constitutional limits. It held that the scheme
did not infringe the right of medical practitioners to choose to dispense medicines as
ZONDO CJ
57
part of their practice. It held that the scheme, therefore, did not infringe section 22 of
the Constitution nor did it infringe any of the other constitutional rights that had been
asserted by the applicants in that case.

[120] The conclusions of the High Court as recorded above including the conclusion
that the licensing scheme did no more than regulate the practice of dispensing
medicines within permissible constitutional limits were confirmed by this Court. This
is how Ngcobo J, writing for a unanimous Court, put it:

“For all these reasons, the contention that the Minister exceeded her
powers in making regulations that link a license to dispense medicines
to particular premises cannot be sustained. The finding of the
High Court in this regard must, therefore be upheld. But, the applicants
had another string to their bow. They contended that, if the scheme of
the Medicines Act authorises the linking of the issuing of a license to
dispense medicines to specific premises, it falls outside the purview of
regulation permitted by section 22 of the Constitution .”29 (Emphasis
added.)

[121] To the extent that the licensing scheme of the Medicines Act authorised the
linking of the issuing of a licence to compound and dispense medicines to specific
premises, did it fall outside the p urview of regulation permitted by section 22 of
the Constitution? The applicants in the Affordable Medicines case contended that it did.
The respondents in that case contended that it did not. Section 22 of our Constitution
is based on section 12(1) of Germany’s Basic Law.30 Section 12(1) of Germany’s Basic
Law reads:

“(1) All Germans shall have the right freely to choose their
occupation or profession, their place of work and their place of
training. The practice of an occupation or profession may be
regulated by or pursuant to a law.”

29 Affordable Medicines above n 14 at para 55.
30 Section 12(1) of the Constitution of the Federal Republic of Germany, 1949.
ZONDO CJ
58

Our section 22 of the Constitution reads as follows:

“Every citizen has the right to choose their trade, occupation or
profession freely. The practice of a trade, occupation or profession may
be regulated by law.”

Dealing with article 12(1) of Germany’s Basic Law, the German Federal Constitutional
Court held that article 12(1) contemplated a unitary right of freedom of occupational
activity that embraces both the choice and practice of a profession.31

[122] With regard to the standard for determining whether the regulation of the practice
of a profession falls within the purview of section 22 of the Constitution, this Court
said:

“[80] The standard for determining whether the regulation of the
practice of a profession fall s within the purview of section 22 can
therefore be formulated as follows: if the regulation of the practice of a
profession is rationally related to a legitimate government purpose and
does not infringe any of the rights in the Bill of Rights, it will fall within
the purview of section 22. Where the regulation of a practice, viewed
objectively, is likely to impact negatively on the choice of a profession,
such regulation will limit the right freely to choose a profession
guaranteed by section 22, and must therefore meet the test under
section 36(1). Similarly, where the regulation of practice, though
falling within the purview of section 22, limits any of the rights in the
Bill of Rights, must meet the section 36(1) standard. ”
(Emphasis added.)

[123] This passage says that the test for determining whether a statutory provision is a
regulation within section 22 – in other words it is not a regulation that goes beyond a
regulation and limits the right – is whether “the regulation, viewed objectively, is likely

31 Id.
ZONDO CJ
59
to impact negatively” on the right or the activity that is the subject of the right. If the
answer is that, viewed objectively, the regulation is likely to impact negatively on the
right or the activity that is the subject of the right, then that regulation limits or will limit
the right. In such a case the regulation must meet the test under section 36(1) of
the Constitution. However, where, it follows, the regulation is unlikely to impact
negatively on the right or the activity which is the subject of the right, then the regulation
does not limit or will not limit the right and section 36(1) has no application.

[124] If one were to apply the Affordable Medicines test for determining whether
section 31B(3), in so far as section 31B(3) constitutes a regulation or constitutes a
limitation, one would have to go to OSA’s founding affidavit to see whether OSA did
place any evidence before this Court that tends to show that the signature requirement
was likely to impact negatively on an independent candidate’s right to stand for public
office or right to associate or disassociate or right to make political choices, the answer
would be that OSA simply did not place any such evidence before the Court. Indeed,
that should not be surprising because how can going out to c ommunities to get
registered voters who will support your candidature as an independent candidate
conceivably impact negatively on your candidature or on your right to stand for public
office as an independent candidate? It simply cannot. Therefore , it simply cannot
possibly limit your right. It is, therefore, not a limitation.

[125] Ngcobo J said in paragraphs 92, 93 and 94 of the judgment:

“[93] That said, however, the scope of permissible regulation that we
adopt here is not entirely inconsistent wit h the German approach. It
recognises that it is not always possible to draw a clear line of
distinction between regulation that affects the practice of a profession,
on the one hand, and one that affects choice on the other. It requires
that where, objec tively viewed, the regulation of the practice of a
profession impacts negatively on choice such regulation must be tested
under section 36(1). Such regulation does not fall within the purview
of section 22, and must, therefore, meet, amongst other require ments,
the standard of reasonableness, of which proportionality analysis is an
ZONDO CJ
60
important component. The same standard must be met where the
regulation of the practice of a profession limits any of the rights in the
Bill of Rights. However where, as here, the regulation, objectively
viewed, does not impact negatively on choice, it need only satisfy the
rationality test. In the result, restrictions on the right to practise a
profession are subject to a less stringent test than restrictions on the
choice of a profession.
[94] Where, as here, the Constitution gives the power to regulate a
right, not every regulation of that right amounts to a limitation of the
right in question. But at the same time Parliament may not
unconstitutionally limit the right to practise a profession under the
guise of regulating it. Where the regulation of the right amounts to a
limitation of that right, such a limitation will have to be tested under
section 36(1). In this case we are concerned with regulation that merely
regulates in the sense of facilitating the proper exercise of the right to
practise a profession. It does not limit the right to practise. The
applicants did not contend otherwise.
[95] The question that falls to be determined, therefore, is whether
the linking of a licence to dispense medicines to particular premises is
rationally related to the government purpose of increasing access to
medicines that are safe for consumption. It is to that question that I now
turn.” (Emphasis added.)

[126] As I have said above, this case concerns the right to vote, the right to stand for
public office and the right to free and fair elections. There are at least three cases that
have come before this Court in which this Court was called upon to determine whether
certain statutory provisions limited or infringed the right to vote. These were the
New National Party 32, Democratic Party 33 and Richter.34 In New National Party
this Court enunciated the test for determining whether a statutory provision or an Act
of Parliament limits or infringes the right to vote. That test was endorsed and followed

32 New National Party above n 15.
33 Democratic Party v Minister of Home Affairs [1999] ZACC 4; 1999 (3) SA 254 (CC); 1999 (6) BCLR 607 (CC).
34 Richter v Minister of Home Affairs [2009] ZACC 3; 2009 (3) SA 615 (CC); 2009 (5) BCLR 448 (CC).
ZONDO CJ
61
by this Court in Democratic Party and in Richter. Since the present case is also ab out
the right to vote, the right to stand for public office and the right to free and fair
elections, that test is the most appropriate that should be used. Indeed, I am of the
opinion that it is the test that we should use in a case involving an alleged limitation or
infringement of a right entrenched in section 19 of the Constitution.

[127] As I have already said, this case is about every adult citizen’s rights to vote, every
citizen’s right to free and fair elections and the right to stand for public office and, if
elected, to hold public office. All of these rights are entrenched in section 19 of
the Constitution. These rights are very important in our constitutional democracy
because they are the primary pillars of any democracy. This is because, without the
right to vote, there can be no democracy. The right to vote means nothing if it is
exercised in an election that is not free and fair. The right to vote is meaningless unless
it is accompanied by the right to stand for public office. The right to vote, the right to
stand for public office and the right to free and fair elections are inextricably
intertwined.

[128] Concerning the importance of the right to vote, this Court said through Sachs J
in August:35

“The vote of each and every citizen is a badge of dignity and of
personhood. Quite literally, it says that everybody counts. In a country
of great disparities of wealth and power it declares that whoever we are,
whether rich or poor, exalted or disgraced, we all belong to the same
democratic South African nation; that our destinies are intertwined in a
single interactive polity. Rights may not be limited without justification
and legislation dealing with the franchise must be interpreted in favour
of enfranchisement rather than disenfranchisement.”36


35 August v Electoral Commission [1999] ZACC 3; 1999 (3) SA 1 (CC); 1999 (4) BCLR 363 (CC).
36 Id at para 17.
ZONDO CJ
62
[129] In August prisoners sought an undertaking from the Electoral Commission that
they would be allowed to register to vote and , ultimately, to vote in the 1999 national
and provincial elections. The Commission, which was mandated to manage elections
and to ensure that elections were free and fair, took the position that, if a court made an
order favourable to the prisons, it would comply with it. It is clear from the facts of the
case that the Commission had included prisoners in its plans to conduct election s but
was prepared to do so if the Court ordered that prisoners be included or if the Court
declared that prisoners were entitled to vote. This Court concluded that there was
nothing that disqualified the prisoners from voting. The Court granted an order which
sought to ensure that prisoners were allowed to vote.

[130] In New National Party statutory provisions required every adult citizen who was
otherwise entitled to be registered as a voter and who was otherwise entitled to vote to
be in possession of a bar -coded identity document or a temporary identity certificate
(TIC) in order to be registered as a voter and in order to vote. Many citizens were not
in possession of such identity documents and it was contended that the Department of
Home Affairs lacked capacity to ensure that the citizens who were not in possession of
bar-coded identity documents or TICs would be issued with such identity documents
before the polling day.

[131] In New National Party the New National Party challenged the constitutional
validity of section 1(xii) and section 6(2) read with section 38(6) of the Electoral Act.
Those provisions were, to the extent relevant here, to the effect that South African
citizens who were otherwise entitled to vote could:
(a) register as voters and have thei r names included in the common voters’
roll only if they were in possession of and produced an identity document
(bar-coded ID) issued after 1 July 1986 in accordance with the provisions
of the Identification Act 37 (1986 Act), a TIC issued in terms of
section 6(2) of the Electoral Act; and

37 72 of 1986.
ZONDO CJ
63
(b) vote only if they were registered on the common voters’ roll and in
possession of and produced the bar-coded ID or a TIC.

[132] The New National Party’s complaint against these sta tutory provisions was that
they unjustifiably infringed the right entrenched in section 19(3)(a) of the Constitution
and were, therefore, constitutionally invalid. Section 19(3)(a) reads:

“Every adult citizen has the right—
(a) to vote in elections for any legislative body in terms of
the Constitution, and to do so in secret.”

The Court had this to say about these statutory provisions:

“[10] The aspects of the Electoral Act in issue regulate the way in
which citizens must register and vote. The questi on which must be
answered is whether these requirements constitute an infringement of
the right to vote. This can only be done in the context of an analysis of
the nature, ambit and importance of the right in question, the effect and
importance of other r elated rights, the importance of the need for an
effective exercise of the right to vote and the degree of regulation
required to facilitate the effective exercise of the right .”38
(Emphasis added.)

In New National Party Yacoob J writing for the majority said that there was no point in
belabouring the importance of the right to vote and it was sufficient “to say that the right
is fundamental to a democracy, for without it there can be no democracy. But the mere
existence of the right to vote without proper arrangements for its effective exercise does
nothing for a democracy, it is empty and useless” (emphasis added).39


38 New National Party above n 15 at para 10.
39 Id at para 11.
ZONDO CJ
64
[133] This Court said:

“The right to vote is, of course, indispensable to, and empty without,
the right to free and fair elections: the latter gives content and meaning
to the former . The right to free and fair elections underlines the
importance of the exercise of the right to vote and the requirement that
every election should be fair has implications for the way in which the
right to vote can be given more substantive content and legitimately
exercised.”40 (Emphasis added.)

The Court said that two of the implications referred to in the preceding excerpt are that
each citizen entitled to vote must vote only once in any election and that anybody not
entitled to vote must not be permitted to do so. This Court continued:

“The extent to which these deviations occur will have an impact on the
fairness of the election. This means that the regulation of the exercise
of the right to vote is necessary so that these deviations can be
eliminated or restricted in order to ensure the proper implementation of
the right to vote.”41 (Emphasis added.)

[134] This Court pointed out:

“The Constitution recognises that it is necessary to regulate the exercise
of the right to vote so as to give substantive content to the right .
Section 1(d) contemplates the existence of a national common voters’
roll. Section 46(1), 105(1) and 157(5) of the Constitution all make
significant provisions relevant to the regulation of the exercise of the
right to vote.”42 (Emphasis added.)

I pause here to highlight the fact that in three of the passages quoted above from
New National Party this Court spoke about the impugned provisions in that case being

40 Id at para 12.
41 Id.
42 Id at para 13.
ZONDO CJ
65
provisions which were part of the provisions regulating the right to vote. This suggests
very strongly that this Court accepted that these provisions regulated the right to vote.

[135] In New National Party the effect of the impugned statutory provisions was that
a citizen who was otherwise entitled to register as a voter and have their names included
in the com mon voters’ roll was prevented from registering as voters and having their
names included in the common voters ’ roll unless he or she was in possession of and
could produce a bar-coded identity book or a TIC. Indeed, those provisions also meant
that a Sou th African citizen who was otherwise entitled to vote would be prevented
from voting unless he or she was in possession of, and, could produce a bar -coded
identity book or TIC.

[136] Although OSA did not put its case in this way, it seems to me that its case ca n
be put on the basis that section 31B(3) prevents an adult citizen who otherwise is entitled
to stand for public office and, if elected, to hold office f rom standing for such office
unless he or she obtains 15% signatures of the quota of registered voters of the relevant
region from the previous election. It, therefore, seems to me that the requirement that
was challenged in New National Party had the same effect as the requirement that is
challenged in these proceedings. In New National Party a voter was required to be in
possession of a bar-coded identity book before he or she could be registered as a voter
and before he or she could be allowed to vote. In the present case an independent
candidate or a new political party is required to secure 15% sign atures of the quota of
the previous election in the relevant region before he or she can stand for public office.
In other words, he or she would not be allowed to stand for public office if he or she did
not secure the signatures of registered voters in the relevant region that make 15% of
the quota of the relevant region from the previous election.

[137] In New National Party this Court said:

“[15] The requirement that only those persons whose names appear
on the national voters’ roll may vote, renders the requirement that South
African citizens must register before they can exercise their vote, a
ZONDO CJ
66
constitutional imperative. It is a constitutional requirement of the right
to vote, and not a limitation of the right.
[16] The process of registration and voting needs to be managed
and regulated in order to ensure that the elections are free and fair .
The creation of a Commission to manage the elections is a further
essential though, not sufficient ingredient in this process. In order to
understand the enormity of the problem, one has just to picture the
spectre of millions of South Africans arriving at registration points or
voting stations armed with all manner of evidence that they are entitled
to register or to vote, only to have the registration or electoral officer
sift through this evidence in order to determine whether or not each of
such persons is entitled to register or to vote. It is to avoid this difficulty
that the Electoral Act makes detailed provisions concerning
registration, voting and related matters including the way in which
voters are to identify themselves in order to register on the common
voters’ roll and to vote.”43 (Emphasis added.)

[138] I pause here to say, with reference to the signature requirement in the present
case: in order to understand the enormity of the problem that would arise if a
requirement such as the signature requirement was not put in place, one has just to
picture the spectre of thousands of independent candidates contesting the elections and
voters having to go through numerous pages of the ballot to find their candidates.
Indeed, one has to picture electoral offices having to count votes of thousands of
candidates the majority of whom would not earn even a single seat. It is to avoid these
difficulties that section 31B(3) contains the signature requirement.

[139] In New National Party this Court dealt squarely with the question of when a
statutory provision or an Act of Parliament can be said to constitute an infringement of
the right to vote. The reference to an infringement in the New National Party judgment
is a reference to a limitation of a right. This Court said in effect that, although it is for
Parliament to determine the mea ns by which voters must identify themselves, that did

43 Id at paras 15-6.
ZONDO CJ
67
not mean that Parliament was at large in determining the way in which the electoral
scheme was to be structured. It pointed out that there were important safeguards aimed
at ensuring appropriate protection for citizens who desired to exercise this fundamental
right. This Court said that the first safeguard was that there had to be a rational
relationship between the scheme which Parliament adopted and the achievement of a
legitimate governmental purpose. The Court emphasised:

“Parliament cannot act capriciously or arbitrarily. The absence of such
rational connection will result in the measure being unconstitutional.”44

[140] This Court went on to deal with who bears the onus of proving a limitation or
infringement of a right. It said:

“An objector who challenges the electoral scheme on these grounds
bears the onus of establishing the absence of a legitimate governmental
purpose, or the absence of a rational relationship between the measure
and that purpose.”45 (Emphasis added.)

[141] This Court went on to say:

“[20] 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. The contention in this appeal
is that the impugned provisions of the Electoral Act constitute a denial
of the right to vote to a substantial number of South African citizens .
Any scheme designed to facilitate the exercise of this right carries with
it the possibility that some people will not comply with its provisions.
But that does not make the scheme unconstitutional. The decisive
question which arises for consideration in this case is the following:
when can it legitimately be said that a legislative measure designed to

44 Id at para 19.
45 Id.
ZONDO CJ
68
enable people to vote in fact results in a denial of that right? What a
party alleging that an Act of Parliament has infringed the right to vote
is required to establish in order to succeed will emerge in the process
of answering this question.”46 (Emphasis added.)

[142] This Court pointed out that the exercise to be carried out by a court entails an
evaluation of the consequences of a statutory provision in the process of its
implementation which occurs at some time in the future. It went on to say that it was
necessary, at the outset of the enquiry, to determine the nature of the consequence that
is impermissible. The consequence that would be impermissible in the
New National Party case could best be determined by focussing on the question as to
what Parliament had to achieve. This Court said that Parliament had to ensure that
people who would otherwise be eligible to vote are able to do so if they want to vote
and if they take reasonable step s in pursuit of the right to vote. This Court had this to
say: “More cannot be expected of Parliament. It follows that an impermissible
consequence will ensue if those who wish to vote and who take reasonable steps in
pursuit of the right, are unable to do so”.47 (Emphasis added.)

[143] This Court went on to say:

“It is necessary to determine the circumstances that are to be taken into
account in deciding whether the impugned provisions infringe the right
to vote. There are two possibilities. A court can make an evaluation in
the light of the circumstances pertaining at the time the provisions were
enacted, o r those which exist at some later date when the
constitutionality of the provisions are challenged.”48 (Emphasis added.)

This Court also stated:


46 Id at para 20.
47 Id at para 21.
48 Id at para 22.
ZONDO CJ
69
“Nevertheless, the implementation of an Act which passes
constitutional scrutiny at the time of its enactment, may well give rise
to a constitutional complaint, if, as a result of circumstances which
become apparent later, its implementation would infringe a
constitutional right. In assessing the validity of such a complaint, it
becomes necessary to det ermine whether the proximate cause of the
infringement of the right is the statutory provision itself, or whether the
infringement of the right has been precipitated by some other cause,
such as the failure of a governmental agency to fulfil its responsibilities.
If it is established that the proximate cause of the infringement, in the
light of the circumstances, lies in the statutory provision under
consideration, that provision infringes the right. This is not a departure
from the objective approach to unconstitutionality. It is merely a
recognition of the fact that a constitutional defect in a statutory
provision is not always readily apparent at the time of its enactment, but
may only emerge later when a concrete case presents itself for
adjudication.”49 (Emphasis added)

[144] The test for determining whether a statutory provision or an Act of Parliament
constitutes a limitation or infringement was articulated thus by this Court in New
National Party:

“Parliament must ensure that people who would otherwise be eligible
to vote are able to do so if they want to vote and if they take reasonable
steps in pursuit of the right to vote . More cannot be expected of
Parliament. It follows that an impermissible consequence will ensue if
those who wish to vote and who take reasonable steps in pursuit of the
right, are unable to do so.”50 (Emphasis added.)

In New National Party this Court also said:

“Parliament is obliged to provide for the machinery, mechanism or
process that is reasonably capable of achieving the goal of ensuring

49 Id.
50 Id at para 21.
ZONDO CJ
70
that all persons who want to vote, and who take reasonable steps in
pursuit of that right, are able to do so. I conclude, therefore, that the
Act would infringe the right to vote if it is shown that, as at the date of
the adoption of the measure, its probable consequence would be that
those who want to vote would not have been able to do so, even though
they acted reasonably in pursuit of the right. Any scheme which is not
sufficiently flexible to be reasonably capable of achieving the goal of
ensuring that people who want to vote will be able to do so if they act
reasonably in pursuit of the right, has the potential of infringing the
right. That potential becomes apparent only when a co ncrete case is
brought before a court. The appellant bears the onus of establishing that
the machinery or process provided for is not reasonably capable of
achieving that purpose. As pointed out in the previous paragraph, it
might well happen that the right may be infringed or threatened because
a governmental agency does not perform efficiently in the
implementation of the statute. This will not mean that the statute is
invalid. The remedy for this lies elsewhere. The appellant must fail if
it does no t establish that the right is infringed by the impugned
provisions in the manner described earlier. This Court held in August v
The Electoral Commission that all prisoners would have been
effectively disenfranchised without constitutional or statutory authority
by the system of voting and registration which had been put into place
by the Commission. This case is different, however, because the
alleged disenfranchisement is said to arise from the terms of the statute
and not from the acts or omissions of t he agency charged with
implementing the statute.”51 (Emphasis added.)

[145] In New National Party this Court made it clear that the “aspects of the
Electoral Act in issue regulate the way in which citizens must register and vote”.52 This
is a sufficient indicat ion that the statutory provisions impugned in that case regulated
the right to vote. The provisions that were impugned in New National Party precluded
adult citizens who were otherwise entitled to vote from voting unless they obtained a

51 Id at para 23.
52 Id at para 10.
ZONDO CJ
71
bar-coded identity document or TIC and produced it for registration and for voting. In
the present case the impugned provision precludes an adult citizen who is otherwise
entitled to stand for public office from standing for public office unless he or she has
obtained signatures of registered voters in the relevant region that amount to 15% of the
quota of that region from the previous election.

[146] So, if the provision that was impugned in New National Party was a regulatory
provision regulating the right to vote and the provision impugned in the present case is
a regulatory provision regulating the right of an independent candidate to stand for
public office – each provision precluding the exercise of a right under section 19 unless
something had been or has been done or achieved – there can be no basis for accepting
that the impugned provision in New National Party was a regulatory provision but the
impugned provision in the present case is not a regulatory provision. If the impugned
provision in the present case is accepted as a regulatory provision, there can then be no
doubt that this Court’s decision in New National Party applies and that the judgment of
this Court in Affordable Medicines also applies.

[147] In New National Party this Court said the issue it had to determine was “whether
the measure itself constitute[d] such denial [of the vote] and [was] on that account an
infringement of the right to vote”. This Court then said:

“To establish this, the appellant must show that the machinery,
mechanism or pro cess provided for by the Electoral Act is not
reasonably capable of ensuring that those who want to vote and who
take reasonable steps in pursuit of the right are unable to exercise it.”53

The judgment of this Court in Democratic Party was handed down on the same day as
this Court’s judgment in New National Party but the latter was handed down first. This
can be gathered from the dates of their hand down as they appear in the judgment as
well as from the terms of the judgment in Democratic Party. Goldstone J wrote the

53 Id at para 37.
ZONDO CJ
72
judgment of the Court. Both in New National Party and in Democratic Party the issues
were largely the same. The Democratic Party challenged the constitutional validity of
the Electoral Act in so far as it only allowed t o register to vote and ultimately to vote
only those adult citizens of South Africa who were in possession of bar -coded identity
documents or TIC.

[148] The Democratic Party contended that the Electoral Act infringed the right to vote
in so far as it had this r equirement. This Court rejected this contention for the same
reasons that were advanced in New National Party . In other words, this Court also
concluded in Democratic Party that the Electoral Act did not constitute a limitation or
an infringement of the right to vote. It reached this conclusion for the same reasons that
it had given in New National Party.

[149] In Richter, a unanimous judgment of this Court written by O’Regan J, who had
dissented from the majority in New National Party , this C ourt followed th e approach
adopted in New National Party in determining whether a statutory provision limited the
right to vote. Indeed, in Richter this Court endorsed the test that this Court enunciated
in New National Party.

[150] In Richter what was challenged was the constitutionality of section 33(1)(e) of
the Electoral Act. In terms of the Electoral Act the default position was that, if a citizen
would be out of the country on polling day, he or she would not be able to vote.
However, th e Electoral Act made provision for exceptions to this general rule.
Section 33(1)(e) of the Electoral Act made special provision for certain cl asses of
South African citizens who would be abroad on polling day in 2009 to vote from outside
of the country. The classes of people that s ection 33(1)(e) provided would be allowed
to vote from outside the country on polling day were people who were temporarily
absent from the Republic for purposes of a holiday, a business trip, attendance of a
tertiary institutio n or an educational visit or participation in an international sports
event.

ZONDO CJ
73
[151] The classes of people who would be permitted under section 33(1)(e) to vote
from outside of South Africa did not include persons in Mr Richter’s class. Mr Richter
was temporarily in the United Kingdom (UK) where he was working as a teacher . He
did not fall into any of the classes that would be allowed under section 33(1)(e) to vote
from outside the country. Mr Richter contended that the failure of the Electoral Act to
include people in a similar position as him among those who would be allowed to vote
from outside the country constituted an unjustifiable and unreasonable limitation of
their right to vote.

[152] In Richter, talking about the importance of the right to vote, O’Regan J said:

“[53] The right to vote is symbolic of our citizenship, as Sachs J
declared. In entrenching the right of every citizen to vote, section 19 of
our Constitution affirms that symbolic value. But the right to vote, and
its exercise, has a constitutional importance in addition to this symbolic
value. The right to vote, and the exercise of it, is a crucial working part
of our democracy. Without voters who want to vote, who will take the
trouble to register, and to stand in queues, as millions patientl y and
unforgettably did in April 1994, democracy itself will be imperilled.
Each vote strengthens and invigorates our democracy. In marking their
ballots, citizens remind those elected that their position is based on the
will of the people and will remain subject to that will. The moment of
voting reminds us that both electors and the elected bear civic
responsibilities arising out of our democratic Constitution and its
values. We should accordingly approach any case concerning the right
to vote mindful of the bright, symbolic value of the right to vote as well
as the deep, democratic value that lies in a citizenry conscious of its
civic responsibilities and willing to take the trouble that exercising the
right to vote entails.”54


54 Richter above n 34 at para 53.
ZONDO CJ
74
[153] In Richter this Court pointed out that the right to vote imposes an obligation upon
the state not merely to refrain from interfering with the exercise of the right but to take
positive steps to ensure that it can be exercised. The right to vote necessitates an
electoral system and the calling of elections.55 O’Regan J said:

“[56] Just as the State bears a responsibility to take positive steps to
enable elections to take place, the right to vote itself cannot be exercised
by a citizen unless he or she takes the trou ble to exercise it. The very
process of regulating the elections which requires the composition of a
national voters’ roll, the establishment of voting stations and voting
times will impose burdens upon members of the public who wish to
exercise their rig ht to vote. First, they will have to register in good
time. Then, on polling day, they may have to journey some distance to
a voting station; they will have to be in possession of a bar -coded
identity document; and they may have to stand in a long queue to vote.
These burdens are largely unavoidable.”56 (Emphasis added.)

[154] Paragraphs 57 and 58 of this Court’s judgment in Richter are very important.
They include the test enunciated in New National Party for determining whether a
statutory provision limits or infringes the right to vote. In those paragraphs O’Regan J
said:

“[57] In assessing whether the restrictions or burdens placed on a
voter who wishes to exercise his or her right to vote are inconsi stent
with the constitutional protection of the right to vote, a court will accept
that a voter may not complain if the burden imposed does not prevent
the voter from voting, as long as the voter takes reasonable steps to do
so. As the majority in this court noted in New National Party:
‘Parliament is obliged to provide for the machinery, mechanism
or process that is reasonably capable of achieving the goal of
ensuring that all persons who want to vote, and who take
reasonable steps in pursuit of that righ t, are able to do so. I

55 Id.
56 Id at para 56.
ZONDO CJ
75
conclude, therefore, that the Act would infringe the right to vote
if it is shown that, as at the date of the adoption of the measure,
its probable consequence would be that those who want to vote
would not have been able to do so, even though they acted
reasonably in pursuit of the right.’
[58] In approaching each of the provisions in question in this case,
therefore, I would suggest that to determine whether any provision
constitutes an infringement of s ection 19 of the Constituti on, we must
establish whether the consequence of any of the challenged provisions
is such that, were a voter to take reasonable steps to seek to exercise
his or her right to vote, any of the provisions would prevent the voter
from doing so. In determining what would constitute reasonable steps
for the voter to take, we should bear in mind both the fact that the
process of voting inevitably imposes burdens upon a citizen as well as
the important democratic value of fostering participation in elections
that I discussed above. Should it be found that the provision would
prevent a voter from voting despite the voter’s taking reasonable steps
to do so, the provision will constitute an infringement of section 19. The
next question that will arise is whether the infringement is justifiable in
terms of section 36 of the Constitution.”57 (Emphasis added.)

[155] It is to be noted that in the above exce rpt the test for determining whether a
provision or an Act constitutes a limitation or infringement of the right to vote which
was enunciated in New National Party was formulated slightly differently. In this
excerpt the test is formulated as whether the consequence of any provision is such that,
were a voter to take reasonable steps to seek to exercise his or her right to vote, any of
the provisions would prevent the voter from doing so. If it would, the provision or Act
would constitute a limitation or an infringement. The next stage would then be to do a
section 36 analysis. Obviously, if the answer is in the negative, that would mean that
the statutory provision or Act of Parliament does not constitute an infringement or
limitation of the right to vote or right to stand for public office.


57 Id at paras 57-8.
ZONDO CJ
76
[156] This test requires the person who asserts that the statutory provision or Act
infringes or limits the right to vote or the right to stand for public office to identify what
reasonable steps a voter or independent candidate would have taken which would still
not have been enough for the voter or independent candidate to exercise his or her
section 19 right. The test should be formulated thus: would the provision or Act or
measure in question prevent an independent candidate from standing for public office
even if he or she took all the reasonable steps to exercise that right?

[157] This Court observed – obviously bearing in mind the test enunciated in
New National Party – that, apart from travelling from the UK, Korea and Canada back
to South Africa, there were n o steps that Mr Richter, Mr Tipper or Mrs Moloko – the
other applicants in the Richter matter – could take in order to be able to vote in the 2009
elections other than to travel thousands of kilometres to South Africa to vote.

[158] O’Regan J also said:

“Apart from travelling back to South Africa from the United Kingdom,
South Korea and Canada in order to be present in South Africa on
polling day, 2009, there are no steps that Mr Richter, Mr Tipper or Mr
and Mrs Moloko can take to vote in the 2009 elections. C an it be said
that in requiring them to return home to South Africa to vote, the
election regulations are imposing an obligation of reasonable
compliance upon them? I do not think so. It is acceptable to ask voters
to travel some distances from their home s to a polling station. One
cannot quibble, either, at the fact that delays in casting votes at a polling
station may require voters to queue for considerable periods of time to
vote. It cannot be said, however, that requiring a voter to travel
thousands of kilometres across the globe to be in their voting district on
voting day is exacting reasonable compliance from a voter. This is all
the more so, given that section 33(1)(b) expressly does not require those
working abroad on government service to return home to vote, but
ZONDO CJ
77
provides voting facilities for them at embassies, high commissions and
consulates.”58 (Emphasis added.)

[159] This Court then concluded that section 33(1)(e) constituted a limitation of the
right to vote “by restricting the classes of voters who are absent from the Republic on
polling day from participating in the election ”.59 In my view another way of putting
this is that section 33(1)(e) limited the right to vote of those voters who would be outside
of South Africa by not including them in the special arrangements it made for other
classes of voters who would be outside the country on polling day. This Court t hen
engaged in a section 36 analysis and concluded that the limitation was not reasonable
and justifiable in a democratic society based on freedom, equality and dignity.

[160] The test for determining whether a statutory provision or an Act infringes the
right to vote or the right to stand for public office involves asking the question whether
people who otherwise qualify to vote or to stand for public office or who are eligible to
vote and want to vote will be able to do so if they take reasonable steps in pursuit of the
right to vote. This means that the person challenging the constitutionality of a statutory
provision must show that people who are otherwise eligible to vote or to stand for public
office and want to vote or to stand for public office will not be able to vote or stand for
public office despite taking reasonable steps in pursuit of the right. OSA falls dismally
short here. It has not addressed in its founding p apers what steps independent
candidates would reasonably be expected to take in order to be able to exercise their
right to stand for public office. In New National Party this Court found that the Act did
not infringe or limit the right to vote.

[161] At this stage I wish to summarise the test for determining whether or not a
statutory provision or an Act of Parliament infringes or limits a right entrenched in
section 19 of the Constitution. This may involve some repetition of what I have already
said but I consider it appropriate to do so for emphasis and clarity. I emphasise that the

58 Id at para 68.
59 Id at para 70.
ZONDO CJ
78
test for determining whether a statutory provision or an Act constitutes an infringement
or limitation of a right entrenched in section 19 of the Constitution – which includes the
right to vote, the right to a free and fair election and the right to stand for public office
– is whether the provision or Act would prevent a voter from voting or an independent
candidate standing for public office despite the voter or independent candidate taking
reasonable steps to do so. This test has two legs. The first leg is whether the statutory
provision or Act would prevent the voter from voting or would prevent an independent
candidate from standing for public office, as the case may be. If the answer to this
question is in the negative, that would be the end of the inquiry and the conclusion
would be that the statutory provision or Act does not constitute a limitation or
infringement of the rights in section 19 of the Constitution.

[162] If, however, the answer is in the affirmative – in other words, if the answer was
that that provision or Act would prevent a voter from voting or an independent candidate
from standing for public office, then the inquiry goes into the second leg. The question
to be asked in the second leg of the inquiry is whether the prevention would occur
despite the voter or independent candidate taking reasonable steps to vote or to stand
for public office. If the answer to the question in the second leg is in the negati ve, that
would be the end of the enquiry and the conclusion would be that there is no
infringement or limitation of any rights contained in section 19 of the Constitution.
However, if the answer to the question in the second leg is in the affirmative, the n that
would mean that the statutory provision or Act constitutes a limitation or infringement
of the section 19 rights.

[163] The above test entails that it is not enough for a person who contends that a
statutory provision or Act constitutes a limitation or i nfringement of a section 19 right
to simply state that the statutory provision or Act would prevent a voter from voting or
would prevent an independent candidate from standing for public office because that is
only one leg of the test. Such a person must go further and say that the prevention
would occur despite the voter or independent candidate taking reasonable steps to vote
or to stand for public office. If that person does not go further, he or she has failed to
ZONDO CJ
79
address the second leg of the inquiry and would fail to show that the statutory provision
or Act constitutes an infringement or a limitation of the right to vote or right to stand
for public office.

[164] In OUTA60 Parliament had the constitutional competence to make laws
concerning “road traffic re gulation”. Parliament passed the Administrative
Adjudication of Road Traffic Offences Act ( AARTO).61 The main question in OUTA
was whether the AARTO Act fell within the functional area described as “road traffic
regulation” or “traffic and parking” or “mu nicipal roads” in Schedules 4 and 5 of the
Constitution. This Court referred to the meaning given to the verb “regulate” in the
South African Concise Oxford Dictionary where the verb “to regulate” is given the
meaning “ 1. T o control or maintain the rate o r speed (of a machine or process) ; 2.
Control or supervise by means of rules and regulations”. 62 In OUTA this Court inter
alia said about the meaning of regulation:

“It seems to me that, in relation to the functional area ‘road traffic
regulations’, the Constitution confers upon the national and provincial
spheres of government the concurrent competence of making laws that
control traffic on the roads.”63 (Emphasis added.)

[165] This statement reflects that this Court accepted that the concept of regulation
entails the control of something.64 In Beerman65 the Court said:


60 Organisation Undoing Tax Abuse v Minister of Transport [2023] ZACC 24; 2023 JDR 2533 (CC); 2023 (10)
BCLR 1189 (CC) (OUTA).
61 46 of 1998.
62 OUTA above n 60 at para 108.
63 Id at para 114.
64 Id.
65 Rex v Beerman 1947 (2) SA 1028 (C).
ZONDO CJ
80
“The meaning of ‘to regulate’ is to control or govern and ‘regulated’
means governed by rule; properly c ontrolled or directed; adjusted to
some standard.”66

In De Freitas67 a Full Court said:

“The two main purposes of the Act are to regulate the right of advocates
and attorneys to appear in courts and to extend the existing right of
attorneys to appear in courts. To regulate means to ‘control’ or
‘govern.’”68

[166] There is a very helpful discussion of the meaning of the verb “regulate” by
De Villiers J in The Magic Company 69 in which he also refers to some cases of the
Appellate Division (now Supreme Court of Appeal) and the Privy Council. He referred
to Perumal70 at para 528A where, relying on Feinstein,71 Didcott J said that ‘“it is surely
inherent in the idea of regulation that something or other may perforce be forbidden’.
‘It follows that the crucial question about such legislation is not whether it contains a
prohibition affecting the activity liable to be regulated, but whether any ban embodied
in it has such a character and extent that the activity itself has been substantially
prohibited. That question is obviously one of degree in each individual case’”.

[167] Referring to Didcott J, De Villiers J said:

‘“Referring to the case of R v Williams 1914 AD 460, the learned judge continued to
say that some assistance is to be gained, when one attempts to answer it in a given

66 Id at 1030.
67 Society of Advocates of Natal v De Freitas 1997 (4) SA 1134 (N).
68 Id at 1122.
69 The Magic Company (Pty) Ltd v Moleketi, the Member of the Executive Council of the Gauteng Province
Responsible for Finance and Economic Affairs 1999 JDR 0521 (T).
70 S v Perumal 1977 1 SA 526 (N).
71 Feinstein v Baleta 1930 AD 319 at 323 and 329.
ZONDO CJ
81
situation from the judgment of Wessels AJA in that case. The judgment continues as
follows:
“A provincial Ordinance was then in issue. It had forbidden betting
on horse races by any means but the totalisator, with the result that
ordinary betting through bookmakers and with others was abolished.
The Provincial Council had however lacked the power to do more than
regulate betting on horse races. The relevant part of the Ordinance was
consequently declared ultra vires. Wessels AJA had this to say, at
pp468, 469-470:
‘By giving a subordinate legislature the rig ht to make
regulations the superior legislature does not as a rule
intend the subordinate legislature to sweep away
completely or substantially the matter it is required to
regulate. To regulate is not to prohibit either entirely or
substantially. . . . You cannot, under the guise of
regulating, in fact prohibit the usual and prevalent method
of betting. You cannot so frame your regulations that you
virtually regulate the subject matter out of existence. No
doubt a power to regulate implies a power to restrict, and
therefore even to prevent, but only in a small degree, and
not so substantially that very little of the former rights
remains. No hard and fast rule can be laid down, and all
that we can say is that, if a regulation takes away a
substantial a nd important portion of existing rights, it
becomes in fact a prohibition. . . . I therefore ask myself
whether the Provincial Council merely regulates betting,
or whether it prohibits it, by substituting a bet on a
totalisator for the well -known methods of betting with
bookmakers and between the members of the public, and I
can come to no other conclusion tha n that it does away
with such substantial and important kinds of betting that it
must be said rather to prohibit than to regulate betting.’
That is t he background against which bylaw must be considered. It
undoubtedly incorporates a prohibition. What must be examined is the
prohibition’s precise effect.”
ZONDO CJ
82
Earlier in the judgment, at 527F-H the learned judge pointed out that the sources of the
principle set out in R v Williams were judgments by the Privy Council in two cases:
“The one was Municipal Corporation of the City of Toronto v Virgo
1896 AC 88, in which Lord Davey said (at p93):
‘There is marked distinction to be drawn between the
prohibition or prevention of a trade and the regulation or
governance of it, and indeed a power to regulate and
govern seems to imply the continued existence of that
which is to be regulated or governed.’
In the other, Attorney-General for Ontario v Attorney-General for the
Dominion 1896 AC 348, the point was made thus by Lord Watson, at
p363:
‘A power to regulate naturally, if not necessarily, assumes,
unless it is enlarged by the context, the conservation of the
thing which is to be made the subject of regulation.’
Both these statements were endorsed and adopted in R v Williams ,
supra at pp465, 469.”
In Fouche v McDonald 1963 4 SA 968 (C) at 971A-B Diemont J (as he then was) said:
“[O]ne must bear in mind that many regulations must of necessity
contain a limited prohibition.”
In S v Schoenfeld 1963 4 SA 77 (T) at 79B-E Colman AJ (as he then was) stated:
“Counsel for the appellant, having brought these provisions to our
notice, referred to the case of R v Williams 1914 AD 460, which is
frequently cited in support of the proposition that ‘a power to regulate
is not a power to prohibit ’. He therefore suggested that by -law 14,
which prohibits certain conduct, was ultra vires the regulatory powers
conferred by the enabling legislation.
But to summarise the principle referred to in R v Williams supra, and
the authorities cited therein in the words ‘a power to regulate is not a
power to prohibit ’ can be misleading. The principle is that when a
subordinate law-making body is empowered to regulate a certain type
of activity it is not thereby authorised to prohibit that type of activity
altogether; it is not true to say that under a power to regulate, nothing
ZONDO CJ
83
may be prohibited. It is clear from such cases as Brown v
Pretoria Municipality 1926 TPD 59, that in the proper exercise of
power versus regulation a local authority may prohibit certain conduct
incidental to or connected with the activity which is being
regulated.””’

[168] In Virgo,72 the Privy Council heard an appeal from a decision of the
Supreme Court of Canada. The question before the Privy Council was whether a section
of a by-law was completely and validly made by the corporation of the city of Toronto.
However, the Privy Council said that it took the view that the real question was whether
under a power to pass by-laws “for regulating and governing” hawkers, etc. the Council
may prohibit hawkers from plying their trade at all in a substantial and important portion
of the city.

[169] The Privy Council had this to say about what regulation means or does not mean:

“No doubt the regulation and governance of a trade may involve the
imposition of restrictions on its exercise both as to time and to a certain
extent as to place where such restrictions are in the opinion of the
public authority necessary to prevent a nuisance or for the maintenance
of order. But their Lordships think there is marked distinction to be
drawn between the prohibition or prevention of a trade and the
regulation or governance of it, and indeed a power to regulate and
govern seems to imply the continued existence of that which is to be
regulated or governed. An examination of other sections of the Act
confirms their Lordships’ view, for it shews that when the Legislature
intended to give power to prevent or prohibit it did so by express
words.” (Emphasis added.)

This statement was approved by Lord Watson in Attorney-General for Ontario.73


72 Municipal Corporation v Virgo [1896] A.C. 88 (1).
73 Attorney-General for Ontario v Attorney-General for the Dominion 1896 A.C. 348 at para 363.
ZONDO CJ
84
[170] It is quite clear from the authorities discussed above that the power to regulate
includes the power to control, to govern, to set norms and standards, to prohibit at least
partially and not totally or substantially, to prescribe conditions and qualifications and
to impose some restrictions. In this case section 31B(3) – which contains the signature
requirement – provides that an independent candidate must obtain and produce
supporting signatures of registered voters equal to 15% of the quota of the relevant
region in the previous election in order to contest an election. That is a provision that
falls within a regulation and not a limitation of the right to stand for public office.

[171] In New National Party this Court held that a statutory provision which facilitates
the exercise of a right to vote did not constitute a limitation . Section 31B(3) is not a
limitation of any right. The second judgment’s conclusion to the contrary errs. The
signature requirement does not adversely affect anybody. Section 31B(3) does not in
any way adversely affect an independent candidate. It does not require an independent
candidate to do anything unconnected with his or her personal goals or ambitions. It
requires him or her to do that which he or she was bound to do sooner or later, namely,
it requires him or her to go out and canvass and obta in the support of only 15% of the
registered voters he or she was going to canvass anyway. It requires an independent
candidate and a new political party to start campaigning for votes earlier. This has no
adverse consequences for the independent candida te or his or her right to stand for
public office. In fact the signature requirement entails that he or she must advance his
or her ambition of obtaining a seat in the National Assembly or in a provincial
legislature. The signature requirement of 15% can therefore not be said to be a
limitation of this right to stand for public office. The requirement is ultimately
beneficial to the candidate’s candidature. Any suggestion that it limits his or her right
is to misconstrue the requirement.

Second judgment
[172] I have read the judgment (second judgment) by my Colleague, Kollapen J. It
strongly disagrees with my conclusion that my use of the test adopted by this Court in
New National Party, Democratic Party and Richter for the determination of whether a
ZONDO CJ
85
statutory provision limits the right to vote and with my conclusion that the signature
requirement doesn’t constitute a limitation. It takes the view that the test in
New National Party does not apply in this case. In my view the second judgment errs.

[173] The New National Party case concerned the right to vote and the right to free
and fair elections entrenched in section 19 of our Constitution. The present case is a
case concerned with the right to vote, the right to stand for public office and the right to
free and fair elections. The right to vote, the right to stand for public office and the right
to free and fair elections are inextricably intertwined. Although, generally speaking it
can be said that facilitating the exercise of the one may facilitate the ex ercise of the
others, this does not mean that the exercise of each one of them may be facilitated in
the same way. There may be facilitation that is peculiar to the right to vote which might
not necessarily apply to the facilitation of the exercise of the right to stand for public
office. The signature requirement is a form of facilitation of the right to stand for public
office which would not apply to the right to vote. Indeed, section 31B(3) constitutes
regulation of the right to stand for public offi ce. The statutory provisions with which
this Court had to deal in New National Party were also part of the regulation of the right
to vote. That is why when one reads this Court’s judgment in that matter, one finds
many instances where it expressly refer s to regulation and many instances where,
although it does not say so in so many words, it is quite clear that it considered the
provisions to regulate the right to vote. The test in New National Party was followed
in Democratic Party and in Richter.

[174] Quite apart from the test in New National Party , if one accepts, as I do, that
section 31B(3) constitutes regulation, one can also use the test used by this Court in
Affordable Medicines. I have discussed the decision of this Court in
Affordable Medicines at some length. Even on that test, too, the signature requirement
does not constitute a limitation. OSA does not satisfy the New National Party test nor
does it satisfy the Affordable Medicines test.

ZONDO CJ
86
[175] Even applying the test that the second judgment appli es, the result will be that
the signature requirement does not constitute a limitation. Section 31B(3) does not
constitute a denial of the right to stand for public office. The signature requirement has
no adverse or prejudicial effects on the right of a n independent candidate to stand for
public office. Instead, it gets the candidate to go out and persuade registered voters to
support his or her candidature – something that seeks to advance his or her political
ambitions. Once he or she has secured the support of 15% of the quota of registered
voters in a region, the candidate does not have to go back to those voters when the
election campaign starts because those have already said that they support him or her.
So, there is absolutely no basis for the proposition that section 31B(3) constitutes a
limitation. How does something that advances an independent candidate’s right to stand
for public office suddenly become a limitation of that right? The 15% registered voters
are voters that the independent candidate was going to have to approach anyway during
his or her election campaign even if the signature requirement was not there.

[176] The second judgment has rejected Parliament’s decision to use a percentage
rather than a fixed number for the signatures of registered voters that must be secured
by an independent candidate to stand for public office. There is absolutely no
justification for the second judgment to effectively set aside that decision of Parliament
when it was not one of the decisions that OSA sought to have set aside. This Court
should defer to Parliament on such a matter. The second judgment had to take such an
unwarranted step because it was faced with the problem of what relief it could grant to
OSA after reaching the conclusion that secti on 31B(3) constitutes an unjustifiable and
unreasonable limitation of the right to stand for public office. The second judgment
itself provides no proper basis for effectively usurping Parliament’s power to make such
a decision that rightly is that of Parliament.

[177] The second judgment relies heavily on the decision of the majority in Moloto74
in this Court to say that the test in New National Party does not apply. In the paragraph

74 South African Transport and Allied Workers Union (SATAWU) v Moloto [2012] ZACC 19; 2012 (6) SA 249
(CC); 2012 (11) BCLR 1177 (CC).
ZONDO CJ
87
in which the majority in Moloto referred to in New National Party , it said th at
New National Party applies to cases where the statutory provision facilitates the right
to vote. That is what I have said section 31B(3) in this case does. However, the majority
in Moloto seems not to have appreciated the difference between statutory provisions
that constitute a complete or total denial of a right such as those in section 65(1) of the
LRA and those that permit the exercise of a right provided certain conditions are met or
provisions that regulate rather than totally limit a right. I s ay this because a mere look
at section 65 of the LRA would have revealed that the provisions in section 64 were not
limitations of the right to strike but sought to regulate and that those of section 65 were
the ones which were limitations. Both the major ity in Moloto and the
Labour Appeal Court in CWIU75 on which the majority in Moloto relied fell into the
same error. The second judgment also seems to fall into the same error.

[178] The second judgment decides that the signature requirement of 15% of the quota
of the region from the previous election should be replaced by a requirement of 1 000
signatures. I have already said above that it is Parliament’s decision whether to use a
percentage or a fixed number and t hat there is no proper basis for this Court to
second-guess Parliament on this issue. Furthermore, the purpose of the 15%
requirement was to g auge the support of a candidate to see whether he or she has
credible prospects of getting a seat in the election s if he or she were to contest the
election. By determining that the threshold should be a fixed number i.e. 1 000
signatures as opposed to a percentage of a quota of a region the second judgment has
decided to use a one-size-fits-all approach despite the fact that it is common cause that
the different regions or provinces have vastly different sizes which is why the figures
reflected both in OSA’s affidavit as well as the IEC’s affidavit relating to the numbers
that would represent 15% of different quotas are so vastly different.

[179] One thousand (1 000) signatures may have been able to give an indication of the
support for an independent candidate or new political party if the regions were of the

75 Chemical Workers’ Industrial Union v Plascon Decorative (Inland) (Pty) Ltd [1998] ZALAC 27 (CWIU).
ZONDO CJ
88
same size and the quotas were not above, for example, 20 000 registered voters but once
the quotas reach 40 000 and above, that an independent candidate was able to get 1 000
supporting signatures will not be an indication of anything really. Furthermore , there
is a great risk that, if this threshold of 1 000 signatures is used, the ballot paper will be
unmanageable because of the great numbers of independent candidates that may contest
the elections when they have no credible prospects of obtaining enough votes in the
elections to get a seat. Indeed, this may pose a threat to the fairness and freeness of the
election.

[180] The second judgment relies on legislation applicable in other countries to support
its conclusion against the respondents. Those are pieces of legislation which
Parliament, the IEC and the Minister have not been given an opportunity to deal with.
This is unfair to them and it violates their right to a fair public hearing guaranteed by
section 34 of the Constitution. If they had been afforded an opportunity to comment on
such legislation, they may have disputed their relevance or applicability. In fact they
may have produced a number of pieces of legislation from other countries which reflects
a different picture.

[181] In the circumstances OSA has failed to show that the 15% signature requirement
constitutes a limitation of any rights. That being the case, we do not reach the
section 36 analysis. Accordingly, OSA’s contention that section 31(B) (3) of the
Electoral Act as amended by the EAA is inconsistent with the Constitution and invalid
falls to be rejected.

OSA’s second constitutional challenge
[182] The participation of independent candidates in the next election in South Africa
will raise some issues that have never arisen in national and provincial elections in this
country in the past 29 years since the advent of democracy. The difference between the
previous national and provincial elections and next year’s elections is that in the past
29 years our elections did not involve independent candidates whereas, following upon
the EAA, next year’s elections will involve independent candidates.
ZONDO CJ
89

[183] When seats became vacant in the National Assembly or provincial legislature in
the past 29 years as a result of the death or resignation of a member of the
National Assembly or a provincial legislature, that did not give rise to a need for any
recalculation of votes and reallocation of a seat or seats. This was because our electoral
system did not involve independent candidates. All members of the National Assembly
and of provincial legislatures were members of one or other political party represented
in the National Assembly or provincial legislature. When a vacancy arose from the
resignation or death of one of its members, the party concerned would simply nominate
their party member next on the party’s list for th e National Assembly or
provincial legislature.

[184] Now that in next year’s elections and beyond independent candidates will contest
elections and some will win and be allocated seats either in the National Assembly or
provincial legislatures, the EAA had to provide answers to the following questions:
(a) what will happen when an independent candidate wins a seat in
the National Assembly and a provincial legislature since he or
she cannot occupy both seats?
(b) what will happen when an independent candidate who was
allocated a seat in either the National Assembly or a provincial
legislature resigns or dies and his or her seat becomes vacant?
(c) what will happen when an independent candidate for the
National Assembly, and who competes in a single region, gets
enough votes to win two or more seats?
(d) what will happen when an independent candidate for the
National Assembly, and who competes in more than one region,
wins a seat in more than one region?
(e) what will happen when an independent candidate for a
provincial legislature gets enough votes to win two or more
seats?
ZONDO CJ
90

[185] Item 11(f) of the amended Schedule 1A deals with the position in regard to a seat
in a provincial legislature and item 5(i) of the amended Schedule 1A deals with a
regional seat in the National Assembly. Item 23 of Schedule 1A of the EAA deals with
vacancies that occur when an independent candidate dies or resigns and a vacancy then
arises.

[186] Item 11(f) of the amended Schedule 1A reads:

“If no recalculation of provisional allocations for a province concerned
is required in terms of item 12, the provisional all ocation of seats in
respect of that province in terms of p aragraph (d), becomes the final
allocation of such seats to the various parties and independent
candidates, and if such a recalculat ion is required the provisional
allocation of such seats as adjusted in terms of item 12 becomes the
final allocation of such seats to the various parties and independent
candidates.”

[187] Item 5(i) of the amended Schedule 1A reads as follows:

“If no recalculation of provisional allo cations is required in terms of
item 7, the provisional allocation of such seats in terms of paragraphs
(e), (f), (g) and (h) becomes the final allocation of such seats to the
various parties and independent candid ates, and if a recalculation is
required, the provisional allocation of s uch seats, as adjusted in terms
of item 7, becomes the final allocati on of such seats to the various
parties and independent candidates.”

[188] Item 23 of the amended Schedule 1A reads:

“(1) In the event of a vacancy in a region or provincial legislature
with respect to a seat allocated to an independent candidate, the
ZONDO CJ
91
chief electoral officer must in writing allocate the seat by
recalculating the result as follows:
(a) disregarding the votes allocated to the independent
candidate causing the vacancy;
(b) disregarding the votes and seats allocated to the
independent candidates already in office; and
(c) recalculating the result for the region or provincial
legislature in terms of the provisions in sub-item (3).
(2) The vacant seat is awarded to an eligible independent candidate
or party that contested the preceding election in terms of
sub-item (1)(c).
(3) (a) An amended quota of votes per seat must be
determined in respect of such region or province by
dividing the total number of votes cast in the region or
province, minus the number of votes cast in the region
or province in favour of the independent candidate
causing the vacancy, minus the votes cast in such
region or province in favour of independent candidates
already allocated one seat, by the number of seats, plus
one, determined in terms of item 4 or item 8 in respect
of the region or province concerned, minus the seats
held by independent candidates.
(b) The result plus one, disregarding fractions, is the
amended quota of votes per seat in res pect of such
region or province for purposes of the said
recalculation.
(c) The number of seats to be awarded for the purposes of
paragraph (e) in respect of such region or province to
a party or independent candidate participating in the
recalculation mus t, subject to paragraph (d), be
determined by dividing the total number of votes cast
in favour of such party or independent candidate in
such region or province by the amended quota of votes
per seat indicated by paragraph (b) for such region or
province.
ZONDO CJ
92
(d) Where the result of the calculation referred to in
paragraph (c) yields seats not absorbed by the number
awarded to parties or independent candidates, the
surplus of votes accruing to any party, parties or
independent candidates participating in the
recalculation, competes for the remaining seats in
sequence of the highest surplus of votes.
(e) The aggregate of such a party ’s awards in terms of
paragraphs (c) and (d) in respect of such region or
province, subject to paragraph (f), indicates that a
party’s or independent candidate ’s final allocation of
the seats determined under item 4 or item 8 in respect
of that region or province.
(f) In the event of a party being allocated an additional
number of seats in terms of this item and if its list in
question then does not contain the names of a sufficient
number of candidates as set out in item 7(1) or item
12(1), the process provided for in item 24 must be
repeated with the changes required by the context until
all seats have been allocated.”

[189] In considering OSA’s second constitutional challenge it is important to bear in
mind the provisions of section s 46(1)(d) and 105(1) (d) of the Constitution.
Section 46(1)(d) relates to the National Assembly whereas section 105(1)(d ) relates to
provincial legislatures. Section 46(1)(d) reads:

“(1) The National Assembly consists of no fewer than 350 and no
more than 400 women and men elected as members in terms of
an electoral system that—
. . .
(d) results, in general, in proportional representation.”

Section 105(1)(d) reads:

ZONDO CJ
93
“(1) A provincial legislature consists of women and men elected as
members in terms of an electoral system that—
. . .
(d) results, in general, in proportional representation.”

This is a very important constitutional requirement because, when all is said and done,
our Constitution requires that the outcome of our elections must result, in gener al, in
proportional representation.

[190] OSA challenges the recalculation of seats when seats are forfeited in the
National Assembly or a provincial legislature or when vacancies arise in a legislature.

[191] In his affidavit Mr Mamabolo gives an overview of the forfeiture of seats and
vacancies. Mr Mamabolo points out that the forfeiture of seats arises during the initial
allocation of seats following an election. He points out that there are three instances in
which the forfeiture of seats occurs in the National Assembly. These are the following:
(a) If a party has submitted a national or regional list with fewer
names of party candidates than the number of seats to be allocated
to it, the number of seats for which it has no listed candidates is
forfeited.
(b) If an independent candidate stands to be allocated more than one
seat in a region, any excess seats won by the candidate are
forfeited since an individual candidate can, by definition, only
hold one seat.
(c) If an independent candidate contests the election in more than one
region and wins a seat in more than one region, that candidate will
be allocated a seat in the region where he or she received the
highest proportion of votes and any excess seats are forfeited.

[192] Mr Mamabolo points out that, where a forfeiture occurs, the initial or provisional
allocation of seats is recalculated in the affected region(s) in accordance with the
ZONDO CJ
94
method prescribed in item 7(3) to (6) of Schedule 1A, which is applicable to the
National Assembly and item 12(3) to (5) in respect of provincial legislatures.

[193] According to Mr Mamabolo the recalculation entails the following:

“48.1 The party or independent candidate forfeiting seats is
disregarded in the recalculation. Its provisional allocation of
seats for the region (or province), minus the number of seats
forfeited by it, becomes its final allocation of seats.
48.2 An amended quota of votes per seat is determined for the region
(or province) by dividing the total number of votes cast minus
the number of votes cast in favour of the party or candidate
whose seat(s) has been forfeited, by the number of seats, plus
one, reserved for the region (or in the province) minus the
number of seats finally allocated to the said party or
independent candidate. Simply put, the votes cast for the party
or candidate that has forfeited a seat are not counted in
determining the amended quota, and nor are the seats that have
already been finally allocated to that party or candidate.
48.3 The result plus one, disregarding fractions , is the amended
quota of votes per seat for the recalculation.
48.4 Seats are then allocated to the parties and independent
candidates by applying the amended quota to the vote count:
i.e., by dividing the total votes cast in support of each party or
independent candidate by the amended quota.
48.5 With this calculation, not all the seats may be allocated. This
is because the contesting candidates may not, between them,
obtain sufficient votes to meet the full quota of seats. The
unallocated seats are aw arded based on ‘the largest remainder
method’: the surplus votes accruing to each party or
independent candidate (i.e., the remainder votes they obtained
that did not meet the threshold of the quota for a seat) compete
for the remaining seats, in sequence of the highest number of
surplus votes.”

ZONDO CJ
95
[194] Mr Mamabolo then deals with vacancies that arise when an independent
candidate dies or resigns from the National Assembly or a provincial legislature. He
points out that such vacancies are dealt with in items 22 to 24 of Schedule 1A. Item 22
deals with vacancies that arise when a member of the National Assembly or provincial
legislature who is a party representative dies or resigns. Item 23 deals with vacancies
which occur when an independent candidate dies or re signs as a member of the
National Assembly or provincial legislature. Since an independent candidate is not
associated with any other person or party, his or her seat cannot just be reallocated to
another person. Mr Mamabolo points out that instead there needs to be a recalculation
of the votes cast to determine which party or independent candidate who contested the
preceding elections is entitled to be allocated the vacated seat.

[195] Mr Mamabolo points out that the recalculation method is “akin to that appl ied
when seats are forfeited”. He then explains the recalculation thus:

“49.3.1 The votes allocated to the independent ca ndidate causing the
vacancy are disregarded.
49.3.2 The votes and seats allocated to independent candidates already
in office are also disregarded. Since individual ca ndidates
cannot, by definition, hold a second seat, independent
candidates already in office are not eligible to be allocated the
vacant seat an d are accordingly excluded from the
recalculation.
49.3.3 An amended quota of votes per seat is determined for the region
(or province) by dividing the total number of votes cast minus
the number of votes cast in favour of the individual candidat e
that caused the vacancy and minus the number of votes cast in
favour of individual candidates that have already been allocated
a seat, by the number of seats, plus one, reserved for the region
(or in the province) minus the number of seats held by
independent candidates. Simply put, the votes cast for the
independent candidate that caused the vacancy are not counted
in determining the amended quota, and nor are the votes c ast
ZONDO CJ
96
for, and the seats held by, independent candidates that already
hold office in the legislature.
49.3.4 The result plus one, disregarding fractions , is the amen ded
quota of votes per seat for the recalculation.
49.3.5 The number of seats to be allocated to the parties and
independent candidates participating in the recalculation is then
determined by applying the amended quota to the vote count:
i.e., by dividing the total votes cast in support of each party or
independent candidate by the amended quota.
49.3.6 Any unallocated seats are awarded based on the largest
remainder method.
49.3.7 Importantly, item 24 provides that if any party or independent
candidate should stand to lose a seat following the recalculation
contemplated in item 23, the party or independent candidate
will retain that seat, and a further recalculation must be done by
the method prescribed in item 24. This ensures that a party or
candidate may never lose a seat already allocated to it as a result
of a recalculation conducted to fill a vacancy.”

[196] OSA’s gripe with the recalculation provided for by the EAA is that the
recalculation of outcomes in the event of forfeiture and vacancies should not disregard
the votes cast for independent candidates as that affects the proportionality between the
votes cast and allocated seats and advantages the largest party. In short, OSA contends
that upon recalculation in terms of items 7, 12 and 23 as the c ase may be, the quota to
be awarded a seat becomes progressively smaller. Mathematically, that is correct.

[197] Mr Mamabolo explains that the rationale for disregarding independent
candidate’s votes in the recalculation of seats is to ensure that the election of candidates
who are not eligible to hold a seat or have chosen to vacate a seat does not continue to
influence the outcome of the re -allocation of that seat. He points out that in this way
the recalculation method is also fundamentally about respecting voters’ choices: votes
for independent candidates that are not eligible to hold a seat are not factored into the
recalculation of seats for political parties or other eligible independent candidates. He
ZONDO CJ
97
also says that, by the same token, if a vacancy arises, the votes in favour of the previous
incumbent should not influence the determination of the new incumbent to fill the seat.
He adds:

“The fact that the independent candidate has vacated the seat implies
that the votes cast originally for that candid ate must be disregarded
since the retention of the votes for that candidate would result in the
same outcome, electing the same person.”

[198] OSA’s objection to the recalculation includes the suggestion that a vote for an
independent candidate is not a vote fo r that individual but a vote that rejects party
politics. As I understand the position, OSA criticises the notion that votes that were
cast for an independent candidate can end up being awarded to a political party. There
is no merit in this suggestion b y OSA. A voter can vote for an independent candidate
in regard to a provincial legislature and for a party in respect of the National Assembly.
Therefore, just because a voter has voted for an independent candidate does not mean
that he or she necessarily does not want political parties.

[199] Furthermore, the votes in favour of the independent candidate whose seat is
forfeited or vacated is not awarded to a political party – those votes are removed
altogether in the calculation. If OSA is suggesting that the se discarded votes should
instead be reallocated to other independent candidates, the suggestion is untenable. The
fact that a voter chose to vote for independent candidate X does not mean that, if the
vote for X has to be disregarded, the voter would wan t his or her vote instead to go to
independent candidate Y. There is no way of knowing what the voter’s “second choice”
would have been – it might have been for another independent candidate or for a
political party, or the voter might have preferred to spoil his or her paper.

[200] OSA also criticises the recalculation method provided for by the EAA and sa ys
it entails that political parties have “a second bite at the cherry ”. This refers to the
occasion when the amended quota is applied. Mr Mamabolo reject s this criticism by
OSA and provides this explanation:
ZONDO CJ
98

“54.1 In the case of forfeiture , the recalculation is done after a
provisional allocation of seats. If the provisional allocation results in the
allocation of excess seats – to parties or independent candidates – those
seats are forfeited, and the allocation is amended (by applying the amended
quota). The allocation is only made final on the recalculation. The votes
cast in favour of political parties are thus only counted once for the purpose
of the final allocation; the only difference is the application of the vote
count to the amended quota.
54.2 If a vacancy arises, all the votes cast for eligible candidates in the
previous election – parties and unrepresented independent candidates alike
– are s imply applied to the amended quota to ascertain the eligible
candidate to acquire the vacated seat.”

[201] Mr Mamabolo explains that disregarding the votes c ast for independent
candidates in the event of forfeiture or vacancies ensures that inter-party proportional
representation is maintained in the legislature. He sa ys that the votes for any
independent candidate in an electoral system based on proportional representation
inevitably affects inter -party proportional representation in the legislat ure as
independent candidates can only hold one seat irrespective of how many votes they win.
Mr Mamabolo states that by disregarding the votes cast for independent candidates who
are not eligible to be allocated the seat in the recalculation, that influence on inter-party
proportionality is limited.

[202] Mr Mamabolo also responds to OSA’s contention that the recalculation method
provided for by the EAA unfairly benefits larger political parties. He states that the
Commission accepts that there is a numerical bias in favour of parties with more overall
votes – which is a reference to larger parties under the amended quota. Mr Mamabolo
says that this is not necessarily inevitable. He states that it is by no means the case that
smaller parties and eligible independent candidates (that is, independent candidates who
did not initially w in a seat) cannot gain seats under the recalculation method. He
attaches annexure “PSM 1” to his affidavit which seeks to show that it is by no means
ZONDO CJ
99
the case that smaller parties and eligible independent candidates cannot gain seats under
the recalculation method. With reference to that annexure, Mr Mamabolo says:

“57.1 The analysis compares the allocation of votes on the basis of a
regional quota without forfeiture (Quota 1) and on the amended quota
to account for forfeiture of a single seat by an independent (Quota 2) in
three hypothetical scenarios.
57.2 The first two scenarios demonstrate that where there is a
substantially larger party (Party A), it will tend to gain on reallocation.
57.3 However, as illustrated in the third scenario, where there is a
more even spread of the votes (such that Party A is only marginally
larger), a smaller party (Party C) may be allocated the forfeited seat. In
this scenario, where there is less of a disparity in the overall vote count,
the highest remainder of surplus votes becomes important.”

[203] Mr Mamabolo goes on to say that it is also not mathematically impossible for a
smaller party to gain a seat on recalcu lation – even when a larger party dominates.
Moreover, sa ys Mr Mamabolo, the outcome of the amended quota is simply the
mathematical consequence of taking into consideration the overall votes while
disregarding the votes c ast for eligible candidates in re calculating the seats. He says
that this is not an unfair or disproportionate outcome, as it gives effect to the overall
support for the respective participants.

[204] It is for a reason that I have referred extensively to Mr Mamabolo’s explanatory
affidavit. OSA filed its papers in which it attac ked the recalculation method provided
for in the EAA. After some time the IEC filed its explanatory affidavit deposed to by
Mr Mamabolo, the Minister filed his answering affidavit and the Speaker and the
Chairperson of the NCOP filed their answering affidavit deposed to by Mr Chabane. In
these affidavits these parties provided explanations for some of the things for which
OSA criticised the EA A. Indeed, in some cases they provided answers to the
contentions contained in OS A’s founding affidavit but OSA did not file any replying
affidavit to deal with the contents of these affidavits.
ZONDO CJ
100

[205] I accept that the Rules of this Court do not give an applicant in an application for
direct access the right to file a replying affidavit. However, as I put it to OSA’s Counsel
at the hearing, if, after receiving all these affidavits, OSA took the view that it was
necessary to respond to any matters in the answering affidavits and explanatory
affidavit, it would have been open to OSA to approach the Court or the Chief Justice
and ask to be granted leave to file a replying affidavit but OSA never did that. The
result is , therefore, that the explanation s given in these affidavits have not been
challenged or refuted. That being the case, OSA’s case is weakened by the absence of
the replying affidavits.

[206] In my view the explanatory affidavit filed by the IEC and the
answering affidavits filed by the Minister and Parliament have adequately shown
OSA’s case t o have no merits and have provided explanations which have not been
challenged in any way by OSA. Another point which must be emphasised which shows
that OSA’s application has no merit is this. Mr Atkins, who prepared a report on which
OSA relied for its case or part of its case , proposed an alternative to the recalculation
method provided for in the EAA. That alternative is supported by OSA.

[207] The alternative proposed by OSA and Mr Atkins is proposed as an interim
measure for the forthcoming 2024 elections in case this Court upheld OSA’s
constitutional challenges . The alternative proposal applies “the largest remainder
method” to allocate excess seats or vacated seats. This alternative is to the effect that
the seat is awarded to the party that, or independent candidate who, has the highest
remainder of the votes that did not meet the threshold of the quota for receipt.
Mr Mamabolo expresses the view that one difficulty with this alternative proposal is
that the votes cast in favour of a candidate who forfeit s a seat or causes a vacancy will
continue to influence the final outcome of an election.

[208] Another difficulty with OSA’s alternative proposal is that a forfeited or vacant
seat could be awarded to a party or independent candidate with a lo w absolute number
ZONDO CJ
101
of votes. Counsel for the Minister made this point , too, as did Mr Mamabolo.
Mr Atkins has acknowledged the difficulty with the alternative proposal. In his report
Mr Atkins states that “this is a valid concern, and one that should be balanced carefully
with the problem of unfairness and disproportionality” . Of course, contrary to
Mr Atkins’ view, there is no unfairness and unacceptable disproportionality that the
recalculation method provided for in the EAA causes.

[209] As Counsel for the Minister put it in their written submissions, OSA’s proposed
alternative contains “the seeds of its own destruction”. He submitted that OSA ’s
alternative proposal produces a problem of “low remainders” and allows for a solution
where a seat may be allocated to a party or candidate with low overall voter support.
He submitted, in my view correctly, that that is patently antithetical to the principles of
the Constitution including the obligation to achieve proportional representation.

[210] Counsel for the Minister drew our attention to the fact that, despite the fact that
OSA knew from Mr Atkins own report that Mr Atkins acknowledged this difficulty
with the alternative proposal they advocated, they did not deal with this difficulty in
their founding affidavit. Indeed, Counsel for the Minister said that the Minister pointed
this out in his answering affidavit and yet even the written submissions filed on behalf
of OSA were silent on this difficulty . This was despite the fact that the Minister had
drawn attention in his answering affidavit to the fact that in its founding affidavit OSA
was silent about this difficulty of which it was aware. The Minister, the IEC and
Parliament raised this difficulty in their written submissions and yet when Counsel for
OSA presented her main address at the hearing, she did not deal with this difficulty. In
fact, she did not argue OSA’s second constitutional challenge at all in her main address.
This led Counsel for Parliament and Counsel for the Minister to indicate , when they
presented their respective oral argument , that they believed that OSA had abandoned
the second constitutional challenge. In reply , Counsel for OSA only said something
about OSA’s second challenge when I asked her whether she had abandoned the point.
Even then, she said very little.

ZONDO CJ / KOLLAPEN J
102

[211] Having regard to all the above I am satisfied that OSA’s second challenge is
devoid of any merit and falls to be rejected. In the result I would have dismissed OSA’s
application in regard to the signature challenge with no order as to costs , regard being
had to Biowatch.76 OSA’s application in regard to the recalculation challenge falls to
be dismissed with no order as to costs in the light of Biowatch.

Order
[212] The following order is made:
1. The applicant is granted direct access to this Court.
2. The application in respect of the recalculation challenge is dismissed.
3. There is no order as to costs.



KOLLAPEN J (Maya DCJ, Mhlantla J concurring in respect of the whole judgment.
Theron J and Rogers J concurring in respect of only direct access and the signature
requirement point):

Introduction
[213] This case is about the constitutional right to stand for and hold public office . In
particular, it is about the boundaries within which that right may be exercised and given
effect to. Much has been said in this matter regarding independent candidates ’ ability
to contest elections successfully. However, contestation requirements ought not to be
based solely on a candidate’s ability to win or lose. Elections serve a variety of purposes
beyond securing office. They must consider a further, equally important purpose:
expression. In the diversity that represents South Africa and in the commitment to
inclusion that characterises our Constitution, a multiplicity of views enriches rather than
diminishes the quality of our democracy.


76 Biowatch Trust v Registrar, Genetic Resources [2009] ZACC 14; 2009 (6) SA 232 (CC); 2009 (10) BCLR 1014
(CC).
KOLLAPEN J
103
[214] Simply put, democratic elections are not solely a matter of arithmetic, nor are
they only about winners and losers. Instead, electi ons have an intrinsic value wholly
separate from the outcome of the vote. They nurture a regular forum for the exchange
of ideas and views which provide the basis for what Nobel Laureate Amartya Sen calls
“government by discussion”. 77 This Court – in Mogale78 – and courts abroad have
endorsed this value. As articulated by the Canadian Supreme Court, “the right to run
for office provides . . . the opportunity to present certain ideas . . . as a viable policy
option; the right to vote provides . . . opportunity to express support for [those] ideas”.79
The focus is not exclusively on which candidates win or lose but also on the voting
public – which definitively loses if ideas are not freely ventilated. 80 It is in th is
constitutional and political context that the signature requirement for contestation will
be considered.

[215] I have read and considered the comprehensive judgment of the Chief Justice
(first judgment). I agree with the conclusion reached therein that a proper case is made
out for direct access with regard to the interests of justice. I also agree with the
reasoning and conclusion reached on the recalculation challenge. I differ, however,
with the conclusion in the first judgment that the applicant’s signature requirement
challenge falls to be dismissed. Instead, a proper case has been advanced in support of
a declaration of unconstitutionality of section 31B(3) of the Electoral Act (Act) as
inserted by the Electoral Amendment Act (Amendment Act). This conclusion comes
by way of two related propositions: (a) the impugned provision limits the rights in
section 19 to stand for public office and other associated rights, and (b) the state
respondents have not shown that the limitation is reasonable and justifiable in an open
and democratic society based on human dignity, equality, and freedom.

77 Sen The Idea of Justice (Belknap Press, Cambridge 2009) at 324.
78 Mogale v Speaker of the National Assembly [2023] ZACC 14; 2023 JDR 1816 (CC); 2023 (9) BCLR 1099 (CC)
at para 4.
79 Figueroa v Canada (Attorney General) 2003 SCC 37; [2003] 1 SCR 912 at 936.
80 In Illinois State Board of Elections v Socialist Workers Party 440 US 173 (1979) (Illinois State Board) at 185-6,
the United States Supreme Court similarly expressed the dual role of elections: “An election campaign is a means
of disseminating ideas as well as attaining political office. Overbroad restrictions on ballot access jeopardize this
form of political expression”.
KOLLAPEN J
104

[216] Because the first judgment fully sets out the background facts, this judgment
reiterates them only as necessary.

Legislative amendment process
[217] Before the Amendment Act’s enactment, political parties were subject only to a
registration requirement. Political parties that wished to register for elections were
required to submit 1 000 signatures in support of their registration (registration
requirement). The legislative amendment process gave rise to an additional contestation
signature requirement in various forms (signature requirement).

[218] Initially, the signature requirement had two relevant aspects: (a) the third
respondent, the Independent Electoral Commission (IEC), was to prescribe the number
of signatures required and (b) the requirement was to apply only to independent
candidates. After deliberation, the position changed in both respects. Parliament
elected to prescribe the percentage in the legislation, which started at 50% of t he
relevant voting quota and settled at 15%. The signature requirement was also extended
to unregistered political parties . Below, I discuss the parliamentary deliberations on
these two aspects before the National Assembly and National Council of Provinces
(Parliament), the fourth and fifth respondent in this matter, respectively.81

[219] When Parliament began to consider a prescribed percentage for the signature
requirement, the IEC made a presentation on 5 July 2022 to the National Assembly’s
relevant Portfolio Committee. The IEC indicated that the regional quota from the 2019
elections ranged from 68 474 to 92 601 across the nine regions. Further, it set out the
number of signatures that would be required at 50% (34 237 to 46 301); 30% (20 542
to 27 780); 20% (13 695 to 18 520); and 15% (10 271 to 13 890).


81 The parliamentary deliberations can be found here: Parliamentary Monitoring Group Electoral Amendment Bill
(B1-2022), available at: https://pmg.org.za/bill/1054/.
KOLLAPEN J
105
[220] On 12 July 2022, the Portfolio Committee resolved that the signature
requirement should be 50% of the relevant regional quota in the previous election. In a
responsive opinion dated 21 July 2021, Counsel expressed the view that 50%
(approximately 22 000 votes) would be an unjustifiable limitation and advised that
Parliament prescribe no more than 20%. In making this calculation, Counsel relied on
the previous quota for a seat in the National Assembly of approximately 44 000. I pause
to mention that this quota was incorrect. The figure of 44 000 differs from the regional
quota range of 68 474 to 92 601 in the IEC’s presentation and was the approximate
quota for compensatory seats, based on the 201 4 and 2019 elections. I return to this
issue later.

[221] Consequently, on 4 August 2022, the Portfolio Committee decided to lower the
percentage of signatures required to 30% , approximately 14 667 signature s. In a n
opinion dated 26 September 2022, Counsel once again cautioned against the proposal
as an unjustifiable limitation which might pose an impossible hurdle for independent
candidates. Counsel calculated that 30% of the 44 000 quota would require
approximately 14 667 signatures. This represented an er ror by Counsel and the
Portfolio Committee in two ways. First, they incorrectly relied on the 44 000 figure.
Then, they mistakenly asserted that 30% of 44 000 is 14 667. It is not. Thirty percent
of 44 000 is 13 200. One-third of 44 000 is 14 667.

[222] Then, on 4 October 2022, the Portfolio Committee opted to yet again lower the
percentage to 20% . Parliament also chose to extend the requirement to unregistered
political parties. In doing so, Counsel calculated that 20% of the 44 000 quota would
be 8 800 signatures and expressed concern that it would be more than eight times the
1 000 signatures required of unregistered political parties. On this basis, Counsel agreed
with Parliament’s decision to extend the requirement to unrepresented political parties.
Again, while 44 000 was the incorrect figure, Parliament used it consistently.

[223] On 2 February 2023, in a presentation to the Portfolio Committee, the IEC
addressed the issue of the incorrect quota, pointing out that it had provided the correct
KOLLAPEN J
106
figures for the regional quotas to the Portfolio Committee in its previous presentation
and did so again. Seemingly prompted by the IEC’s clarification, in an opinion dated
1 February 2023, Counsel accepted that the regional quota may ultimately be higher
than 44 000 but emphasised that the focus was the percentage , not raw numbers.
Counsel supported 20% but advised that 15% could still be considered if concerns
persisted. On 7 February 2023, the Portfolio Committee resolved to lower the signature
requirement to 15%. There it stayed.

The parties’ submissions
[224] The applicant submits that th e signature requirement gatekeeps elections and
serves as a barrier to entry for independent candidates. As such, the applicant argues
the requirement limits the following rights contained in our Constitution:
(a) the section 19(3) right of citizens to stand for public office and if elected
to hold office;
(b) the section 18 right to associate, as well as the right not to associate , by
running as an independent candidate;
(c) the section 19(1) right of every citizen to be free to make political choices;
and
(d) the independent candidates’ section 10 right to dignity.

[225] Prior to the signature requirement under consideration, political parties were only
required to solicit 1 000 signatures to register as a political party with the IEC. Th is,
the applicant says, sufficed as a contestation requirement, if one was needed at all. For
political parties represented in Parliament, there is no signature requirement for
contestation. Parliam ent has, however, introduced a signature requirement for
contestation equaling 15% of the regional quota for independent candidates as well as
for political parties that are not represented in Parliament.

KOLLAPEN J
107
[226] In raw numbers, the applicant previewed what the s ignature requirement would
entail by region. The requirement for contesting a seat in the National Assembly ranges
from 10 271 in the Northern Cape to 13 890 in Gauteng. This is no small feat, in part,
because the label “signature” requirement does not c apture the fact that independent
candidates must also record signatories’ names and identity numbers. An independent
candidate will be required to provide the names, identity numbers , and signatures of
that number of voters in support of their candidacy as a pre-emptive requirement.

[227] The applicant submits that the signature requirement is arbitrary because it was
based on an incorrect quota of 44 000 and because the reduction of percentage over time
was done without any demonstrated rationale. The applica nt also claims that the
numbers required are not only exorbitant but are out of kilter with the rest of the
electoral syste m. By comparison, the 1 000 signature registration requirement
represents an ordinary, feasible and reasonable measure.

[228] The applicant adds in summary that the impugned signature requirement sets an
unjustifiable and disproportionate threshold. One thousand signatures, however, would
achieve the object of a reasonable and justifiable threshold . It would also balance the
attainment of the political rights the applicant relies on with the need, if any, for a
contestation threshold.

[229] The respondents oppose the relief sought on many fronts, including that:
(a) the signature requirement does not constitute a limitation of any right but
rather a measure necessary for the effective management of elections;
(b) the applicant has not offered any evidence to suggest that it s members
who are independent candidates will be unable to meet the signature
requirement and has therefore not proved a limitation of any right; and
(c) even if it does constitute a limitation of rights, the signature requirement
is reasonable and justifiable as contemplated in section 36 of
the Constitution.
KOLLAPEN J
108

Issues to be determined
[230] This judgment follows the well -established two -stage approach to determine
whether there has been an unjustified infringement of a right. First, this approach asks
whether there has been a limitation of fundamental rights. I f so, it asks whether that
limitation is reasonable and justifiable in an open and democratic society based on
human dignity, equality, and freedom, taking into account all relevant factors, including
those listed in section 36(1). This judgment determines that there has indeed been an
unjustified violation of a right and does so as follows:
(a) The primary point of departure from the first judgment is the application
of the standard set by New National Party .82 Contrary to the first
judgment’s analysis, the New National Party standard exists in a carefully
limited and delineated space. It applies when government takes positive
steps to facilitate or give effect to a right . Further, it has subsequently
been applied and clarified in non -election cases. In particular,
this Court’s judgments in Moloto83, Garvas,84 and Mlungwana85 indicate
that the regulation of the exercise of a right will amount to a limitation
when it goes beyond regulation and has a limiting and deterring effect.
(b) This disagreement with the first judgment is fund amental as it leads to a
different conclusion on the first stage, whether the signature requirement
constitutes a limitation.
(c) Two inquiries form part of the first stage as set out in Walters,86 (i) the
content and scope of the rights; and (ii) the meaning a nd effect of the
signature requirement and whether it constitutes a limitation of the

82 New National Party above n 15.
83 Moloto above n 74.
84 Garvas above n 23.
85 Mlungwana above n 25.
86 Ex Parte Minister of Safety and Security: In re S v Walters [2002] ZACC 6; 2002 (4) SA 613 (CC); 2002 (7)
BCLR 663 (CC) at para 26.
KOLLAPEN J
109
content and scope of the right s. The first Walters inquiry covers rights
enumerated in sections 18(1), 19(1) , and 19(3). The second Walters
inquiry evinces that the signature requirement goes beyond mere
regulation and has a limiting purpose. As such, I conclude that the
signature requirement constitutes a limitation of the applicant’s rights to
freedom of association, freedom to make political choices and to stand for
and hold public office, if elected.
(d) The second stage probes the state’s justification for the limitation. In
short, the limitation is not justifiable. I reach this conclusion by applying
section 36 and taking into account the following : (i) the purpose of the
limitation is of low importance; (ii) the nature and extent of the limitation
is arbitrary and extensive; (iii ) there was a limited relation between the
limitation and its purpose whe re P arliament consistently r elied on the
wrong quota; and (iv ) the 1 000 signature registration requirement
constituted less restrictive means to achieve the same purpose.

[231] Having set out the issues to be determined in this judgment, I now turn to expand
on the primary point of departure with the first judgment. As a starting point, it would
be germane to recall the two -stage test laid down in our law for determining a right’s
violation.

Determining a right’s violation
[232] In New National Party, this Court said that it was pre-eminently the function of
Parliament to determine the structure of the electoral scheme but that it was not at liberty
to do so without any constraints. It said:

“It is to be emphasised that it is for Parliament to determine the means by which voters
must identify themselves. This is not the funct ion 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 to exercise this foundational right. The first of the constitutional
KOLLAPEN J
110
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 caprici ously or arbitrarily. The absence of such 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.”87

[233] I pause to mention that the two constraints referred to are separate and distinct
as set out in New National Party above. I focus exclusively on the latter and find that,
however rationally linked the signature requirement may be to a legitimate government
purpose, it constitutes an unjustifiable limitation on a right. The conclusion reached
arising out of the rationality enquiry is not dispositive of the infringement enquiry. They
represent separate tests.

[234] The two-stage approach to determine a rights infringement is well-established in
South African constitutional law.88 The first stage considers whether the impugned law
limits a right. The second stage assesses whether the limita tion is reasonable and
justifiable in terms of section 36 of the Constitution. The court only moves to the second
stage if a rights limitation is established at the first.

[235] At the first threshold, Walters requires this Court to—


87 New National Party above n 16 at paras 19-20.
88 S v Zuma [1995] ZACC 1; 1995 (2) SA 642 (CC); 1995 (4) BCLR 401 (CC) at para 21; S v Makwanyane [1995]
ZACC 3; 1995 (3) SA 391 (CC); 1995 (6) BCLR 665 (CC) at paras 100-2; Ferreira v Levin N.O. [1995] ZACC
13; 1996 (1) SA 984 (CC); 1996 (1) BCLR 1 (CC) at para 44; S v Mamabolo [2001] ZACC 17; 2001 (3) SA 409
(CC); 2001 (5) BCLR 449 (CC) at para 1.
KOLLAPEN J
111
“determin[e] whether or not the enactment in question constitutes a limitation of 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).”89

Accordingly, this first stage poses two inquiries: what are the boundaries of the right
with regard to its scope and content (the first inquiry); and did the effect of the impugned
provision cross those boundaries (the second inquiry)?90

[236] The first judgment finds that the signature requirement is not a limitation. In
doing so, it does not apply Walters, but instead, impermissibly relies on
New National Party. It finds that the applicant fell short in addressing what steps
independent candidates would reasonably be expected to take in order to be able to
exercise their right to stand for office. As such, it concludes that the applicant has failed
to show that the 15% signature requirement constitutes a limitation of any rights. I take
a different view on limitation and find that New National Party is not applicable.
Accordingly, before analysing whether there has been a limitation, it is necessary to
deal with this primary point of departure.

What is the New National Party reasonable steps standard?
[237] In New National Party, when assessing whether a bar-coded ID requirement was
an infringement of the right to vote, this Court held as follows:

“Parliament must ensure that people who would otherwise be eligible to vote are able
to do so if they want to vote and if they take reasonable steps in pursuit of the right to
vote. More cannot be expected of Parliament. It follows that an impermissible
consequence will ensue if those who wish to vote and who take reasonable steps in
pursuit of the right, are unable to do so.
. . .

89 Walters above n 86 at para 26.
90 Cheadle et al South African Constitutional Law: The Bill of Rights Service 34 (2023) at 30-3 to 30-4. See also
Woolman and Botha “Limitations” in Woolman et al (eds) Constitutional Law of South Africa Service 5 (2013)
at 34-3 to 34-6; Currie and De Waal Bill of Rights Handbook 6 ed (Juta & Co Ltd, Cape Town 2013) at 145-6.
KOLLAPEN J
112
Parliament is obliged to provide for the machinery, mechanism or process that is
reasonably capable of achieving the goal of ensuring that all persons who want to vote,
and who take reasonable steps in pursuit of that right, are able to do so . I conclude,
therefore, that the Act would infringe the right to vote if it is shown that, as at the date
of the adoption of the measure, its probable consequence would be that those who want
to vote would not have been able to do so, even though they acted reasonably in pursuit
of the right.”91 (Emphasis added.)

[238] This Court found that a limitation would be established if the applicants coul d
show that they were unable to exercise their right to vote, even if they acted reasonably
in pursuit of the right to vote. The New National Party approach was followed in
Democratic Party92 and Richter.93

[239] Democratic Party concerned a challenge to the requirement that a potential voter
had to have a bar-coded identity document in order to vote on the basis that it was
discriminatory. This Court dismissed the challenge and held as follows:

“No more recent evidence of the e ffect of the provisions has been furnished . On the
assumption that the opinions expressed in the [Human Sciences Research Council] 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 t he impugned provisions have had the effect
suggested by the [Democratic Party], it cannot be found that the provisions, on that
account, were unconstitutional.”94 (Emphasis added.)


91 New National Party above n 16 at paras 21 and 23.
92 Democratic Party above n 33.
93 Richter above n 34.
94 Democratic Party above n 33 at para 12.
KOLLAPEN J
113
[240] Richter dealt with regulations which precluded certain categories of citiz ens
from voting in national elections while they were abroad. Unlike New National Party
and Democratic Party, this Court found that there was a limitation. It stated:

“In approaching each of the provisions in question in this case, therefore, I would
suggest that to determine whether any provision constit utes an infringement of
section 19 of the Constitution, we must establish whether the consequence of any of
the challenged provisions is such that, were a voter to take reasonable steps to seek to
exercise his or her right to vote, any of the provisions would prevent the voter from
doing so. In determining what would constitute reasonable steps for the voter to take,
we should bear in mind both the fact that the process of voting inevitably imposes
burdens upon a citizen as well as the important democratic value of fostering
participation in elections that I discussed above. Should it be found that the provision
would prevent a voter from voting despite the voter’s taking reasonable steps to do so,
the provision will constitute an infringement of section 19. The next question that will
arise is whether the infringement is justifiable in terms of section 36 of
the Constitution.”95 (Emphasis added.)

[241] While the New National Party standard is clear, I am not convinced that it finds
application in this matter. This view rests on two observations: (a) its application by
this Court in subsequent cases; and (b) the delineated space within which it emerged.

Point of departure
Subsequent case law
[242] In addition t o the above cases, our courts have subsequently applied the
New National Party standard in non -election matters and clarified its application.
This Court in New National Party recognised that some regulation may be required to
facilitate the effective exercise of the right to vote.96 The subsequent case law is useful
for this matter. It tells us that , even if intended solely to regulate the right, laws or
conduct that effectively restrict the right constitute a limitation. It is not the case that

95 Richter above n 34 at para 58.
96 New National Party above n 16 at paras 10-3.
KOLLAPEN J
114
any law that purports to regulate the exercise of a right is permanently shielded from
constitutional scrutiny as a limitation of that right. An overview of the following cases
demonstrates that.

[243] This overview commences with Affordable Medicines 97 and South African
Diamond Producers Organisation .98 These two cases give guidance as to how to
determine where a regulation of a right may limit it. In Affordable Medicines, this Court
considered whether a requirement to dispense medicines only from licensed premises
limited the section 22 right to choose a profession. It said that the “question is whether
the law which purports to regulate the practice of a profession, viewed objectively,
would impact negatively on the choice of a profession”.99 Notably, the Court said that
“Parliament may not unconstitutionally limit the right to practise a profession under the
guise of regulating it”.100

[244] The standard set by Affordable Medicines was further ventilated in South African
Diamond Producers Organisation , where this Court found that the section 22 right to
choose a trade, occupation and profe ssion would be limited either by a “legal barrier”
that prohibits persons from entering a trade or excluding persons from continuing to
practice that trade; or by an “effective limit” that renders “the practice of that trade or
profession so undesirable, d ifficult or unprofitable that the choice to enter into it is in
fact limited”.101 However, neither of these cases refers to New National Party . This
occurred in Moloto.

[245] In Moloto, Yacoob J grappled with the relationship between the right to strike
and its implementing legislation – section 64 of the LRA. Moloto cautioned that any
interpretation of such legislation should not restrict the right more than required by the

97 Affordable Medicines above n 15.
98 South African Diamond Producers Organisation v Minister of Minerals and Energy N.O. [2017] ZACC 26;
2017 (6) SA 331 (CC); 2017 (10) BCLR 1303 (CC) (South African Diamond Producers Organisation).
99 Affordable Medicines above n 15 at para 68 (emphasis added).
100 Id at para 73.
101 South African Diamond Producers Organisation above n 98 at paras 66-71.
KOLLAPEN J
115
Act’s language. This is especially true when reading -in an implied limitation. It said
that:

“Constitutional rights conferred without express limitation should not be cut down by
reading implicit limitations into them and when legislative provisions limit or intrude
upon those rights they should be interpreted in a manner least intrusi ve of the right, if
the text is reasonably capable of bearing that meaning.
. . .
The point of departure in interpreting section 64(1) is that we should not restrict the
right to strike more than is expressly required by the language of the provision, unless
the purposes of the Act and the section on ‘ a proper interpretation of the statute . . .
imports them.’”102

[246] The Court also distinguished the concepts of regulating or giving effect to a right
from limiting a right and distinguished the matter from New National Party. It said:

“This Court, in New National Party, held that where legislative provisions facilitate the
exercise of a constitutional right it cannot be said that they are limitations of that right
that need justification under section 36 of the Constitution. Workers, however, can go
on strike in the sense of withholding work without needing section 64 to enable them
to do it.
From this it can be seen that procedural pre -conditions for the exercise of a
constitutional right place some limit on that right. This limitation would then have to
be justified under section 36 of the Constitution. One of the considerations in the
justification analysis is whether less restrictive means could achieve the same
purpose.”103 (Emphasis added.)

[247] While the Court referred to the New National Party standard, it found that
section 64 was not a legislative provision that “facilitate[d] the exercise of a
constitutional right” as “[w]orkers . . . can go on strike in the sense of withholding work

102 Moloto above n 74 at paras 52 and 54.
103 Id at paras 69-70.
KOLLAPEN J
116
without needing section 64 to enable them to do it”. 104 Instead, section 64 “expressly
requires minimal procedural pre-conditions for the statutorily protected exercise of that
right”.105

[248] Moloto also made reference to the Labour Appeal Court’s judgment in CWIU.106
There, that Court considered the issue of procedural pre -conditions in relation to the
right to strike and material conditions as to whom may strike.107 It concluded that such
conditions limited the right to strike, which was conferred without express limitation in
the Constitution.108 This Court in Moloto agreed with this analysis that the effect of
procedural pre-conditions and material condition s as to who may strike constituted a
limitation.109

[249] This judgment is rel evant for three pertinent reasons. First, it considered the
New National Party framework outside the exclusive context of voting rights. Second,
it closely examined legislation to determine whether or not it facilitated the exercise of
the right to strike. Third, because the Court found that the legislation did not facilitate
the right, its analysis fell outside the New National Party framework and did not apply.

[250] Garvas110 is also instructive. There, the Court addressed the right to assemble
and the liability of organisers for damages arising from gatherings.111 The crisp issues
relevant for the purpose of this application were : the characterisation of the right to
assemble (not restricted in the Constitution) , including measures to regulate the right;

104 Id at para 69.
105 Id.
106 CWIU above n 75.
107 Id at para 29.
108 Id at para 21.
109 Moloto above n 74 at para 69.
110 Garvas above n 23.
111 Id at paras 61 and 71.
KOLLAPEN J
117
and whether those measures went beyond regulation. 112 This Court , as in
New National Party and Moloto, held that measures to regulate the right may not in
themselves constitute a limitation.113 It examined the provisions of section 11(2) of the
Regulation of Gatherings Act which provided a defence to the claim for liability for riot
damage and concluded that the section went beyond regulating the right. 114 It had the
effect of limiting the right to assemble, found the Court, which then immediately
proceeded to the justification analysis.115

[251] The Court dealt with the difficulty such a defence would introduce in terms of
costs and effect in a general sense but said those were matters for consideration in the
justification stage of the analysis.116 It did not embark on the reasonable steps standard
that the Court adopted in New National Party and largely for the same reason advanced
in Moloto,117 namely, the distinction between measures to give effect to a right and those
that limit a right.

[252] Finally, in Mlungwana this Court again had to consider the constitutionality of
provisions of the Gatherings Act which sought to regulate the right to assemble. The
impugned provision resulted in the risk of criminalisation of an organiser for failing to
give notice of the gathering. Confirming Garvas, this Court held as follows:

“Section 12(1)(a) goes beyond mere regulation. In Garvas. . ., this Court found that
deterring the exercise of the right in section 17 limits that right. The reason is obvious.
Deterrence, by its very nature, inhibits the exercise of the right in section 17.

112 Id at para 32.
113 Id at para 55. Here, the Court said:
“The mere legislative regulation of gatherings to facilitate the enjoyment of the right to assemble
peacefully and unarmed, demonstrate, picket and petition may not in itself be a limitation.
Section 11(2), read with section 11(1), goes further than simply to regulate the exercise of the
right in order to facilitate its full and appropriate enjoyment by those who organise and those
who participate.”
114 Garvas above n 23 at paras 34-43.
115 Id at para 60.
116 Id at para 59.
117 Id at para 57.
KOLLAPEN J
118
Deterrence means that the right in question cannot always be asserted, but will be
discouraged from being exercised in certain instances.
In this matter, the criminal sanction in section 12(1)(a) deters the exercise of the right
in section 17. The respondents not only admit this, but invoke the self-same deterrent
effect to explain section 12(1)(a)’s purpose and justify its provisions. The possibility
of a criminal sanction prevents, discourages, and inhibits freedom of assembly, even if
only temporarily. In this case, an assembly of 16 like -minded people cannot just be
convened in a public space. The convener is obliged to give prior notice to avoid
criminal liability. This const itutes a limitation of the right to assemble freely,
peacefully, and unarmed. And this limitation not only applies to conveners, but also to
all those wanting to participate in an assembly. If a convener is deterred from
organising a gathering, then in t he ordinary course (save for the rare spontaneous
gathering) a gathering will not occur.”118

[253] In Garvas and Mlungwana, the legislation in question was enacted for the
purpose of giving effect to the right to strike and the right to assemble. 119 In Garvas,
the Court found its effect was to limit the right and proceeded with the justification
analysis.120 The Court did not interrogate whether the gathering’s organisers could
show that, even if they had taken reasonable steps, they could not have enjoyed the right
to assemble (the New National Party test).121 That question did not arise because the
Court, having distinguished measures to give effect to a right and measures that limited
the right, found that the measure in question limited the right. 122 That concluded the
limitation inquiry. In Mlungwana, the impugned provision was also found to go beyond
regulation. No inquiry into reasonable steps was taken. The mere possibility, even
temporary, of criminal sanction was seen as having a deterring effect that li mited the
right.


118 Id at paras 46-7.
119 Id at para 130 and Moloto above n 74 at para 20.
120 Garvas above n 23 at para 130.
121 Id at paras 81-4.
122 Id at para 57.
KOLLAPEN J
119
[254] The conclusion arising from this Court in these three matters could not be clearer:
when a measure gives effect to a right, courts apply the reasonable steps test from
New National Party; when a measure limits a right, courts proceed with the justification
analysis.

[255] A second strong thread emerges from these cases. They highlight that the
threshold for determining when a regulation goes beyond giving effect to a right is
relatively low. This is, in part, because the boundaries of the right are to be interpreted
generously and the determination on limitation is not dispositive to a constitutional
challenge. A conclusion on limitation still affords the state the opportunity to show that
the limitation is reasonable and justifiable in an open and democratic society.

[256] In these cases, mere deterrence – measures discouraging or inhibiting the
exercise of the right – sufficiently extended the regulation into the arena of limitation.
The impugned measures were not outright legal barriers and, on the face of it, were not
particularly burdensome. Still, the first judgment claims that limitations must be “total
or complete denial [s] of the right or prohibition[s] of the right ”.123 This raises the
threshold way beyond where this Court has said it should be located.

[257] In Mlungwana, this Court, cited in support of its reasoning the approach taken
by the European Court of Human Rights that interference with the right did not need to
amount to an outright ban, but could occur as a result of other condition s or
requirements. It concluded that criminalising a failure to give notice of a gathering
constituted a limitation of the right of assembly. As was noted in Mlungwana:

“Similarly, regarding the right to freedom of assembly under the European Convention
of Human Rights (ECHR), the Grand Chamber of the European Court of Human Rights
(Grand Chamber) held that ‘the right to freedom of assembly is a fundamental right in
a democratic society and, like the right to freedom of expression, is one of the

123 First judgment at [177]. See also at [104].
KOLLAPEN J
120
foundations of such a society. Thus, it should not be interpreted restrictively ’. The
Grand Chamber went on to find that—
‘the interference [with the right in article 11(1)] does not need to
amount to an outright ban, legal or de facto, but can consist in various
other measures taken by the authorities. The term “restrictions” in
article 11(2) must be interpreted as including both measures taken
before or during a gathering and those, such as punitive measures,
taken afterwards. For instance, a prior ban can have a chilling effect
on the persons who intend to participate in a rally and thus amount to
an interference, even if the rally subsequently proceeds without
hindrance on the part of the authorities. A refusal to allow an
individual to travel for the purpose of attending a meeting amounts to
an interference as well. So too do measures taken by the authorities
during a rally, such as dispersal of the rally or the arrest of participants,
and penalties imposed for having taken part in a rally.’”124 (Emphasis
in the original.)

[258] To summarise:
(a) New National Party established the standard that where impugned law or
conduct gives effect to the exercise of a right to vote, it only limits the
right if it is shown that its probable consequence is that those who want
to exercise the right will be unable to do so , even though they a ct
reasonably in pursuit of the right.
(b) Moloto is authority for the proposition that the New National Party
standard is not confined to political rights cases.
(c) Garvas and Mlungwana confirm that an enactment that goes beyond
giving effect to a right and creates a barrier to the exercise of the right,
will amount to a limitation.

[259] Accordingly, o ur courts have developed the principle that a limitation arises
where the impugned law goes beyond regulation and creates a legal barrier through

124 Mlungwana above n 25 at para 51.
KOLLAPEN J
121
explicit exclusion or in effect through a deterrent which negatively impacts the right.
Both are considered to exceed the prescripts of mere regulation. New National Party
only applies when the G overnment takes positive steps to give effect to a right and
thereby creates reciprocal duties. It does not apply when those steps go beyond giving
effect to the right and result in a limitation of the right.

[260] Given the clear distinction between (a) measures which give effect to and limit
a right and (b) the cons equences which flow from either (as established in
New National Party and supported in Moloto, Garvas and Mlungwana), the second
Walters inquiry directs this Court to determine the impugned provision’s nature.125

The context of New National Party
[261] I also differ from the first judgment in part because it fails to properly take into
account the context within which New National Party arose.

[262] New National Party’s reasonable steps standard arose in the specific context of
election cases and has been the subject of considerable debate. 126 The ability to
meaningfully exercise a right will , in many instances, create an obligation on the part
of government to take positive steps to realise or give effect to that right.127 This
obligation may be expressly stated in the Constitution as is the case with section 32 (the
right of access to information) and section 33 (the right to just administrative action).128
At other times, the obligation may be necessarily implied as is the case with the rights
to vote and stand for publ ic office. 129 It is largely within this context that the

125 Walters above n 86 at para 27.
126 Brickhill and Babiuch “Political Rights” in Woolman et al (eds) Constitutional Law of South Africa Service 5
(2013) vol 3 at 45-18 and 45-19; Rautenbach “Political Rights” in Bill of Rights Compendium Service 37 (2018)
at 1A67.4.
127 Brickhill and Babiuch above n 128 at 45-21.
128 Sections 32 and 33 of the Constitution.
129 New National Party above n 16 at para 118.
KOLLAPEN J
122
New National Party standard applies, namely where positive government action must
be coupled with reasonable exercise by the rights-bearers.

[263] New National Party is confined to this constitutional space. The positive
obligations placed on the state are to facilitate or give effect to the enjoyment of the
right.130 This may still require reasonable steps on the part of the citizen in return. It
is, in part, about the reciprocity of obligations which may characterise the relationship
between state and rights-bearer if a right is to be meaningfully exercised.131 This Court
has held that in giving effect to the right to vote, positive obligations are imposed on
the state132 and citizens.133

[264] For instance, the regulatory framework that requires voters’ names and details to
be captured on a voters’ roll and a bar -coded identity document to identify the voter s
on election day gives effect to the right to vote.134 There is an inextricable and necessary
link between these positive steps that the g overnment takes and the realisation of the
right to vote. Those steps remain inchoate unless and until citizens take reasonable
measures to apply to have their names recorded on the voters’ roll and to apply for a
bar-coded identity document. 135 It would then not constitute an infringement of the

130 Id.
131 Richter above n 34 at para 55.
132 Positive obligations imposed on the state include: ( a) setting a date for elections; (b) securing the secrecy of
the ballot; (c) establishing an independent and impartial IEC to ensure free and fair elections and to take reasonable
steps to ensure eligibl e voters are registered; (d) providing the legal framework, infrastructure and resources
necessary for the holding of free and fair elections; (e) ensuring that national, provincial and municipal elections
are held in terms of an electoral system which mus t be prescribed by national legislation; (f) an electoral system
that must, in general, result in proportional representation; (g) elections for the National Assembly must be based
on the national common voters’ roll; and (h) 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.
Richter above n 34 at paras 53 -4; Minister of Home Affairs v National Institute for Crime Prevention and the
Re-Integration of Offenders (NICRO) [2004] ZACC 10; 2005 (3) SA 280 (CC); 2004 (5) BCLR 445 (CC) at
para 28; New National Party above n 16 at para 13; August above n 35 at para 16.
133 Positive obligations imposed on citizens include: (a) registering in good time; (b) travelling to a voting station
(which may be at a distance); (c) being in possession of a bar-coded identity document; and (d) potentially standing
in a long queue to vote. See Richter above n 34 at para 55; New National Party above n 16 at para 15.
134 New National Party above n 16 at para 26.
135 Richter above n 34 at para 55.
KOLLAPEN J
123
right if citizens were able to access the mechanisms to have their names recorded on the
voters’ roll or to apply for and receive a bar -coded identity document. 136 It is under
these circumstances that a conclusion of a limitation will only arise when the citizen is
able to show that, despite reasonable steps, they will be unable to exercise the right.137

[265] It is a different matter when the government’s positive action limits the right.138
This implicates the first leg of the Walters inquiry.139 When positive action limits rather
than gives effect to a right, the government must justify the limitation.140

[266] Government action may both give effect to a right and simultaneously limit it.
When government action primarily gives effect to a right , it is accompanied by an
individual’s duty to take reasonable measures to access the right.141 If, however, those
reasonable measures do not enable the right to be realised, a limitation may still result.142
This is the effect of the holding in New National Party.143 This Court did not purport
to change the law on limitation. The Court did not address the position on limitation
but rather located the decision in relation to a law that gives effect to the right to vote.144
The jurisprudence on the first stage, described above,145 was left intact.

[267] On the other hand, when the primary intention or effect of government action
limits a right, the right is limited as a matter of law.146 The test then moves to the second
stage – the justification. This distinction is both necessary and important as it

136 New National Party above n 16 at para 21.
137 Rautenbach “Political Rights” above n 128 at 1A67.4.
138 Walters above n 86 at para 27.
139 Id at para 26.
140 Moloto above n 74 at paras 69-71 and Garvas above n 23 at para 60.
141 New National Party above n 16 at para 23.
142 Rautenbach “Political Rights” above n 128 at 1A67.4.
143 Id.
144 New National Party above n 16 at para 26.
145 See [234] to [236] above.
146 Rautenbach “Limitations” Bill of Rights Compendium Service 37 (2018) at 1A31.
KOLLAPEN J
124
jurisprudentially retains the difference between a limitation of a right and the
justification of a limitation under section 36. It cannot be correct that the state attra cts
no duty to justify the limitation, merely because its stated intention was to give effect
to the right rather than to limit it.

[268] For the above reasons, I differ with the first judgment’s determination of whether
there is a limitation in this matter. I now turn to consider whether a limitation has been
established and start such a consideration with the first Walters inquiry: what a re the
boundaries of the rights in question?

The first stage: has there been a limitation?
The first inquiry: what are the boundaries of the rights?
[269] The instant matter implicates a cluster of three rights contained in
our Constitution: the section 18 right to freedom of association, the section 19(1) right
to freedom to make political choices, and the section 19(3) the right to stand for and
hold public office.147 In Ramakatsa,148 this Court held that “the text of a section in the
Bill of Rights must be read generously and purposively in order to give the right-holders
the full protection afforded by the guaranteed right”. 149 Further, this Court in
New Nation Movement provides particular guidance as to the interpretation of the rights
raised in this matter:

“The importance of these rights cannot be overstated in the South African context
where – for centuries – those rights were enjoyed only by the white minority.
Accordingly, the rights at stake here fall to be interpreted generously, rathe r than
restrictively. In August Sachs J held:
‘The achievement of the franchise has historically been important both
for the acquisition of the rights of full and effective citizenship by all
South Africans regardless of race, and for the accomplishment of an

147 Sections 18, 19(1) and 19(2) of the Constitution.
148 Ramakatsa v Magashule [2012] ZACC 31; 2013 JDR 2203 (CC); 2013 (2) BCLR 202 (CC).
149 Id at para 70.
KOLLAPEN J
125
all-embracing nationhood. The universalit y 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. In a country of great disparities
of wealth and power it d eclares that whoever we are, whether rich or
poor, exalted or disgraced, we all belong to the same democratic
South African nation; that our destinies are intertwined in a single
interactive polity. . . . Legislation dealing with the franchise must be
interpreted in favour of enfranchisement rather than
disenfranchisement.’”150

Section 19(3) – the right to stand for and hold public office
[270] Our Constitution explicitly guarantees the right to candidature. 151
The Constitution does so for two reasons: the legal right to candidacy is well established
internationally and in foreign jurisdictions, and the moral right to candidacy invokes far
more than legal instruments. It is grounded on values like equality, autonomy, and
participatory democracy. In his autobiogra phy, former President Nelson Mandela
recounts an early observation of traditional participatory democracy at the local level:
“all men were free to voice their opinions and were equal in their value as citizens” .152
It is this freedom to participate, and th e dignity found therein, which informs our right
to candidacy.

[271] While supported by international and foreign law, the right to candidacy and the
principles underpinning it , are neither new nor foreign to South Africa. The
Freedom Charter itself so proclai ms.153 Included within its first demand, The People
Shall Govern!, the South African Congress Alliance pressed that “[e]very man and

150 August above n 35 at para 74.
151 Section 19(3)(b) of the Constitution.
152 Mandela Long Walk to Freedom (Little, Brown and Company, 1994) at 28.
153 African National Congress “The Freedom Charter”, available at: https://www.anc1912.org.za/the-freedom-
charter-2/.
KOLLAPEN J
126
woman shall have the right . . . to stand as a candidate for all bodies which make
laws”.154

[272] More recently, this Court in My Vote Counts II155 said that the Constitution did
not prohibit adult citizens from standing as independent candidates, there was just no
legislation to facilitate such candidacy:

“This does not mean that the right is not available to be enjoyed by whoever might have
lost confidence in political parties. It does, in my view, remain open to be exercised
whenever so desired, regardless of whatever logistical constraints might exist.”156

Similarly, in New Nation Movement , this Court found that section 19(3) was
inextricably linked to the freedom of association.157

[273] And so, when independent candidates invoke their right to stand for public
office, law and principle dictate that such candidates should not be disadvantaged by
their choice not to associate with political parties. The choice to associate and not to
associate are both worthy of equal constitutional protection, which must manifest itself
in the rules and arrangements that govern elections.

[274] The language of section 19(3)(b) creates the right to stand for public office and
does so without limitation, except that it applies to adult citizens only. 158 Given the
importance of this right both in itself as well as in relation to the cluster of political
rights, its boundaries must be drawn generously. Both impli ed and express limitations
are inimical to a constitutional framework designed to give effect to, rather than
circumscribe, section 19(3)(b).

154 Id.
155 My Vote Counts NPC v Minister of Justice and Correctional Services [2018] ZACC 17; 2018 (5) SA 380 (CC);
2018 (8) BCLR 893 (CC).
156 Id at para 29.
157 New Nation Movement above n 3 at para 14.
158 Section 19(3)(b) of the Constitution.
KOLLAPEN J
127

[275] Accordingly, the ambit of the right is wide, in that it seeks to ensure an open and
democratic political system i n which all eligible candidates can participate in the
electoral system and all citizens are able to exercise their right to vote for their preferred
candidate.

Section 19(1) – freedom to make political choices
[276] The purpose of the freedom to make political choices is twofold. It is positive
in that it safeguards an individual’s political choice, and it is negative, in that it bars the
state from curbing such a choice .159 The Constitution elaborates on this right by
enumerating concomitant rights to form a political party, to participate in the activities
of, or recruit members for, a political party, and to campaign for a political party or
cause.160 This list in section 19(1) is not exhaustive .161 The right also protects “[a]
conscious choice not to form or join a political party”. 162 Its scope is expansive as it
contains the intersection of various rights and freedoms of a political character,
including expression, association, belief, opinion and conscience.163

159 Cheadle et al above n 90 at 14-3.
160 Section 19(1) of the Constitution states:
“Every citizen is free to make political choices, which includes the right—
(a) to form a political party;
(b) to participate in the activities of, or recruit members for, a political party; and
(c) to campaign for a political party or cause.”
161 New Nation Movement above n 3 at para 17.
162 Id.
163 In this regard, in Pilane v Pilane [2013] ZACC 3; 2013 JDR 0295 (CC); 2013 (4) BCLR 431 (CC) at para 69,
this Court stated as follows:
“It strikes me that the exercise of the right to freedom of expression can be enhanced by group
association. Similarly, associative rights can be heightened by the freer transmissibility of a
group’s identity and purpose, expressed through its name, emblems and labels. These rights are
interconnected and complementary. Political participation, actuated by the lawful exercise of
these rights, can and should assist in ensuring accountability in all forms of leadership and in
encouraging good governance.”
My Vote Counts II above n 157 at para 27 states the following:
“Choice is of its own a loaded concept. And much more is required of a choice -maker if the
choice to be made is political in character and affects important national interests. The gravity
of the choice is more pronounced in relation to the right of an adult citizen to participate or vote
KOLLAPEN J
128

Section 18 – freedom of association
[277] The right to freedom of association protects against coercion in our private and
political lives .164 The right has dual content in that it allows for a person to freely
associate (positive element) as well as to decide not to associate at all (negative
element).165 Although not inherently political, the right to association has political
corollaries. For instance, it nurtures “vigorous exchanges and debates between
competing interest groups to give real choices to the electorate to confront the political
class”.166 Association is an expansive right directly related to the political rights in
question, as was held in New Nation Movement in respect of section 19(3).167

[278] In conclusion, against the backdrop of South Africa’s colonial-apartheid history,
the inclusion of these rights in our Bill of Rights holds significant symbolic value which
supports the inherent importance of this cluster of rights. 168 Their inclusion, both
individually and collectively, is crucial for the functioning of our constitutional

in the elections for ‘any legislative body ’. This is because of the centrality of elections in the
functioning, preservation and effectiveness of our constitutional democracy.”
See also Cheadle et al above n 90 at 14-3; Brickhill and Babiuch above n 128 at 45-30; Rautenbach above n 128
at 1A67.1.
164 New Nation Movement above n 3 at para 27 captures Haysom’s reasoning for the protection of freedom of
association:
“These are a wide range of reasons why freedom of association is so highly prized, vigorously
protected and widely acclaimed as a cornerstone of a democratic society. These reasons belong
to one or other of two perspectives: a perspective which emphasise s the need to associate in
order to realise fully one’s humanity – to interact, combine, make common purpose and enjoy
life with other persons sharing one’s cultural, personal, political or economic interests. The
second perspective emphasises the necessity to a functioning democracy of such a freedom, for
a proper and coherent expression and interplay of collective interests. Both perspectives are,
however, grounded on the same understanding that a person alone is an atomised, powerless,
lonely being wit hout a foundation for developing an identity or the capacity to influence or
change his or her physical environment or social world.”
Haysom “Significance of the Freedom of Association” in Cheadle et al (eds) in South African Constitutional Law:
The Bill of Rights Service 27 (2019) at 13.1.
165 New Nation Movement above n 3 at para 22.
166 Haysom “Significance of the Freedom of Association” above n 166 at 13.1.
167 New Nation Movement above n 3 at para 14.
168 Id at para 143.
KOLLAPEN J
129
democracy. South Africa was founded on the values of universal adult suffrage; a
national common voters’ roll; regular elections, and a multi-party system of democratic
government.169 The Constitution does more than enumerate these rights; it establishes
various institutions to give effect to them, such as the IEC.170

[279] The content of these rights must be interpreted to promote the value system of
an open and democratic society based on human dignity, equality , and freedom. As to
their scope, the interrelatedness of the rights is indicative of their wide breadth. It has
been found that the right to stand for public office is linked to the right to make political
choices and freedom of association.171

[280] Further, international and regional law widely recognises the interrela tionship
between the vote, suffrage, association, and candidacy.172 Most notably, the Universal
Declaration of Human Rights links direct participation to Article 21’s right to take part
in the government. 173 While the Drafting Committee originally included t he right to
candidacy in the Declaration, South Africa’s apartheid government was one of two
nations to object to enumerating candidacy as a human right. 174 What may be

169 Section 1(d) of the Constitution and Cheadle et al above n 90 at 14-1.
170 Section 181(1)(f) of the Constitution.
171 In New Nation Movement above n 3 at para 14, this Court stated as follows:
“It seems to me that in the context of this matter the freed om of association challenge is
inextricably linked to what the content of the section 19(3)(b) right really is. That is so because
the applicants’ plea is not only about adult citizens not being coerced to be members of political
parties. It is about not being so coerced so that they may exercise the section 19(3)(b) right.
And they can exercise that right in this fashion only if it is guaranteed by section 19(3)(b). ”
172 Article 25 of the International Covenant on Civil and Political Rights, 1994 (enshrining the rights “to vote and
be elected”); Article 23 of the American Convention on Human Rights , 1969; Article 13 of the African Charter
on Human and People’s Rights , 27 June 1981 (proclaiming that “every citizen shall have the right to participate
freely in the government of his country”); Protocol 1, Article 3 of the European Convention on Human Rights,
1953 (interpreted to include the right to stand for election in 1987 in the decision in Mathie-Mohin v Belgium
(1988) 10 EHRR 1 at para 51).
173 Article 21 of the Universal Declaration of Human Rights, 1948.
174 Comments from Governments on the Draft International Declaration on Human Rights, Draft International
Covenant on Human Rights and the Question of Implementation, UN doc E/CN.4/82/ADD.4 (1948) at 23,
available at: https://documents-dds
ny.un.org/doc/UNDOC/GEN/GL9/001/46/PDF/GL900146.pdf?OpenElement.
KOLLAPEN J
130
fundamental in other democracies 175 has acquired a heightened significance for
post-apartheid South Africa.

[281] Having considered the boundaries of the rights in question, I now turn to the
second inquiry, a determination of whether the signature requirement crosses these
boundaries and constitutes a limitation.

The second inquiry: does the signature requirement cross the boundaries of the
rights?
[282] Whether or not the signature requirement crosses the boundaries of the rights in
question requires determining the nature of the infringement.176 An impugned law will
cross a right’s boundaries wher e it intentionally seeks to restrict protected activity or
unintentionally limits the right where it is overbroad or has deterring effects. 177 As
established above, this can occur even when the impugned law seeks to give effect to
the exercise of the right.

[283] Does the impugned provision constitute a limitation? The impugned signature
requirement creates a threshold requirement that any person seeking to stand for public
office must meet. It is a procedural precondition and a material limitation on who may
exercise the right to stand for election and of the kind described in CWIU and Moloto.
It also entails a precondition of substance in its demand that a potential candidate
actively seek out and secure s the written support of a designated number of voters for
their candidacy. The inability to do so would bar them from standing for public office
as contemplated in section 19(3)(b) of the Constitution.


175 The constitution of Canada explicitly protects the right to candidacy. Canadian courts have interpreted this
provision to be “one of the most fundamental in the Charter, going as it does to the very heart of a free and
democratic society”. Harvey v New Brunswick (Attorney General) 1996 CanLII 163 (SCC); [1996] 2 SCR 876
at 905.
176 Cheadle et al above n 90 at 30-3 to 30-4.
177 Id; Woolman and Botha above n 30 at 34-5 fn 2.
KOLLAPEN J
131
[284] The rights in this matter, and the right to stand for public office, in particular, do
require a degree of regulation to be given effect to. According to the respondent s,
independent candidates could take reasonable steps to meet the signature requirement.
If the impugned provision were one that simply ga ve effect to the rights in this matter,
then this criticism would be valid, and the infringement challenge would be dismissed.
The impugned provision does more than give effect to the right, however. It limits the
right. As such, Moloto, Garvas and Mlungwana advise that the justification analysis –
not the reasonable steps test – must follow. In that event, the consideration of
reasonable steps simply does not arise at the limitation stage but may well do so at the
justification stage.

[285] Though the imposition of a signature requirement gives effect to, facilitates or
regulates the right to stand for public office, the signature requirement set out in the
impugned provision creates a legal barrier or a pre -condition for its exercise. This is
very different from a requirement to having a bar -coded identity document or taking
steps to apply to have one’s name on a voters’ roll. These measures are formal in nature,
the citizen has the right to have his or her name on the voters’ roll and to h ave a
bar-coded identity document issued and can take legal action to assert such a right if
need be. But more than that, such a requirement may well be necessary to facilitate the
vote. These requirements, unlike the signature requirement in the present case, do not
have as their object to cut down the number of candidates, let alone to do so in order to
make the administration of elections easier.

[286] During the legislative process, the overt purpose of adopting the signature
requirement was to curtail the number of independent candidates that could stand for
office. In other words, the object was that, with this signature requirement, there should
be fewer independent candidates than there would have been without the
signature requirement.

[287] In addition, it is relevant to consider what the respondents say about the
impugned provision in their papers before this Court:
KOLLAPEN J
132
(a) The Minister submitted that the signature requirement was “a sifting
mechanism” that sought “to discourage independent candidates and
political parties who have no plausible hope of obtaining a seat from
contesting the elections” and to prevent political parties and independent
candidates from contesting an election “frivolously” , thereby keeping
ballot papers at “manageable length”.
(b) The IEC explains the two-fold purpose of the signature requirement as follows:
“12.1 First, it ensures that candidates have a serious intention of contesting
elections and do not participate frivolously , and have some prospect
of doing so successfully; and
12.2 Second, it enables the Commission to run free and fair elections more
efficiently. Fairness to voters is a critical consideration in this regard.
Too many candidates on a ballot can compromise voters’ ability to
vote according to their preferences, as it increases the risk of confusion
between candidates and causes spoilt ballots. Limiting the number of
candidates to a manageable number is also necessary to ensure that
voting and vote counting can be done, and the election results declared,
within the prescribed period.” (Emphasis added.)
(c) Parliament states that “the purpose of the signature requirement is to
minimise the prospect of frivolous entries into the election race” and
describes it as a “pre -emptive requirement that operates before the
elections”.

[288] The signature requirement was intended to exclude potential candidates from
standing for public office. This purpose is manifest from the requirement in its own
terms. The rationale offered by the state respondents and the IEC puts this beyo nd
doubt. The introduction of the signature requirement was to limit the right to stand for
public office to those who were able, in the respondents’ view, to demonstrate that they
had a serious chance of winning. While a signature requirement in itself may not be a
limitation, what this case calls for us to do is consider whether a signature requirement
set at 15% of the regional quota and requiring 10 000 to 14 000 signatures constitutes a
KOLLAPEN J
133
limitation. At 15% of the regional quota, the signature requirement places a substantial
burden on independent candidates who wish to contest elections.

[289] The first judgment says that the applicant has not stated why the 15% signature
requirement constitutes a barrier nor has the applicant proven that the requirement goes
beyond mere regulation . This is incorrect. Further, the first judgment faults the
applicant’s May 2023 application for not arguing that candidates had insufficient time
to collect the requisite signatures.

[290] In order to illustrate the extent of the burden, the applicant calculated the number
of signatures from each voting district independent candidates would need to collect in
order to obtain the requisite number of signatures in the respective regions. They are
as follows:
(a) In the Eastern Cape, an independent candidate would have to obtain the
equivalent of three signatures from voters in each of 4 869 voting districts.
(b) In the Free State, an independent candidate would have to obtain the
equivalent of eight signatures from voters in each of 1 582 voting districts.
(c) In Gauteng, an independent candidate would have to obtain the equivalent
of five signatures from voters in each of 2 799 voting districts.
(d) In KwaZulu-Natal, an independent candidate would have to obtain the
equivalent of three signatures from voters in each of 4 972 voting districts.
(e) In Limpopo, an independent candidate would have to obtain the
equivalent of four signatures from voters in each of 3 223 voting districts.
(f) In Mpumalanga, an independent candidate would have to obtain the
equivalent of seven signatures from voters in each of 1 813 voting
districts.
(g) In North West, an independent candidate would have to obtain the
equivalent of seven signatures from voters in each of 1 737 voting
districts.
KOLLAPEN J
134
(h) In the Northern Cape, an independent candidate would have to obtain the
equivalent of 15 signatures from voters in each of 732 voting districts.
(i) In the Western Cape, an independent candidate would have to obtain the
equivalent of nine signatures from voters in each of 1 572 voting districts.

[291] The applicant further argues that the time, resources , and energy that
independent candidates would have to invest in order to meet the 15% signature
requirement to secure the requisite signatures would be immense. It points out that , at
15% of the regional quota, this is between 1 000 and 1 400 percent increase from the
previous 1 000 signatures required from registration to contest elections. I am inclined
to agree with the applicant on this score. While the applicant need not obtain a signature
in all voting districts described above, this illustration reveals the extent of the burden.
The purpose of the signature requirement was to prevent frivolous contestation. The
effect of it is that at 15% o f the regional quota, the signature requirement places a
significant burden on independent candidates to contest elections.

[292] Accordingly, and again leaving aside the merit of the requirement, the
signature requirement limits the right to stand for office, to vote, to make political
choices and to freely associate, by limiting these rights’ protected activity and making
inroads into the boundar ies of the se rights. It cannot, therefore, be said that the
impugned legislation merely gives effect to the se rights. On the contrary, it
intentionally limits them, alternatively, has the effect of limiting them. This in turn
triggers the justification analysis.

[293] This is where I depart from the first judgment. It is incorrect to apply the
New National Party standard whenever government action allegedly limits political
rights. The standard is restricted in scope to situations where impugned law or conduct
gives effect to the right. New National Party, Moloto, Garvas and Mlungwana confirm
as much. None of these cases says that where a provision facially constitutes a
limitation then no limitation exists, and that the applicant must meet a further
KOLLAPEN J
135
evidentiary burden. The signature requirement limits candidacy both in intention and
effect.

[294] Finally, on this aspect, I take the view that the New National Party standard and
the doctrine of limitation coexist in separate though related constitutional spaces. If a
right is found to be limited using the criteria in Walters, the bearer of the right cannot
then be expected, in addition, to demonstrate that by taking reasonable steps they will
not be able to exercise the right. It is for these reasons that I conclude that the rights
have been limited and that the New National Party standard does not apply.

[295] The first judgment concludes that the 15% signature requirement does not
adversely affect an independent candidate as it does not require them to do anything
unconnected with their personal goals or ambitions. 178 This is not the test which must
be applied. The test is whether the impugned law crosses the boundaries of what the
right seeks to protect. This is the test that the applicants have met. In any event, the
logistics involved in obtaining individual signatures from thousands of peo ple appear
to be a task of some magnitude if regard is had to how the first judgment suggests that
this could be achieved.179

[296] Although, for these reasons, the applicant did not bear a burden of proving that
its members could not by taking reasonable steps meet the signature requirement, a brief
observation on the reasonable steps standard is apposite. The applicant launched thes e
proceedings in June this year. The date of the next national elections had not been
announced and is yet to be announced, but it will be in the period between May to
August 2024. At the time the application was launched, therefore, the election date was
unknown and lay nearly a year or more ahead. Furthermore, no regulations had as yet
been promulgated to prescribe the form in which signatures were to be obtained, and

178 First judgment at [171].
179 First judgment at [72].
KOLLAPEN J
136
this is still the case. 180 It is not apparent, in the circumstances, how potential
independent candidates could have been expected to start the process of trying to gather
signatures in order to gauge the achievability of the signature requirement. Voter
registration weekends had not been held, voters at that early stage might be apathetic
and the formal requirements for collecting signatures were not yet in place.

[297] Section 36 of the Constitution permits the state to limit a right, provided of
course, that it is reasonable and justifiable in an open and democratic society based on
human rig hts, dignity and freedom. This singular and unequivocal constitutional
standard applies to all rights and therefore no separate limitation regime can exist in
respect of any cluster of rights, be it political rights or voting rights as the first judgment
suggests. Simply put when the state limits a right it carries with it the duty to justify
the limitation.

[298] There is nothing in New National Party that signals a departure from this
standard nor the creation of a different standard – the Constitution simply does not
permit it and we must be careful of reading into New National Party what is not there.
The test in New National Party can therefore only fit into the scheme of the Constitution
if it fits into the space New National Party says it was designed for – instances where
an enactment regulates a right.

[299] Where an enactment goes beyond this and limits a right then section 36, Garvas,
Mlungwana and the doctrine of justification all tell us the same thing – the duty of the
state to justify the limitation is triggered. To suggest that the existence of the limitation
can be ignored and the rights -bearer be burdened in showing that the right cannot be
achieved by taking the reasonable steps the enactment provides for, does an injustice to
section 36, offends the principle of accountability and fundamentally unsettles the
approach to limitation that the Constitution demands. Its implications could be
far-reaching and ominous.

180 Draft regulations were published on 28 July 2023 for comment by 27 August 2023: GN 1934 of 2023,
GG 49051, 28 July 2023.
KOLLAPEN J
137

[300] In the presence of a limitation, the second stage of the analysis probes its
justification.

The second stage: is there a justification?
[301] Section 36(1) of the Constitution provides that a limitation will be regarded as
reasonable and justifiable in an open and democratic society based on human dignity,
equality and freedom when it is effected, in terms of a law of general application, after
taking all relevant factors into account, including—
(a) the nature of the right;
(b) the importance of the purpose of the limitation;
(c) the nature and extent of the limitation;
(d) the relation between the limitation and its purpose; and
(e) less restrictive means to achieve the purpose.

[302] A section 36 justification analysis is a balancing exercise, described by this Court
in Manamela181 as follows:

“[T]he Court must engage in a balancing exercise and arrive at a global judgment on
proportionality . . . . As a general rule, the more serious the impact of the measure on
the right, the more persuasive or compelling the justification must be. Ultimately, the
question is one of degree to be assessed in the concrete legislative and social setting of
the measure, paying due regard to the means which are realistically available in our
country at this stage, but without losing sight of the ultimate values to be protected.”182

[303] In the main, the state respondents argue that the signature requirement serves
two legitimate government purposes. First, arising from the historical trend of increased

181 S v Manamela (Director-General of Justice Intervening) [2000] ZACC 5; 2000 (3) SA 1 (CC); 2000 (5)
BCLR 491 (CC).
182 Id at para 32.
KOLLAPEN J
138
participation in elections, the signature requirement ensures that those running are
bona fide candidates with serious intentions and avoids frivolous participation. Second,
the signature requirement: avoids confusing, unwieldy ballots; decreases spoilt ballots;
lowers the cost implications ; and avoids delays in voting, voting counting and the
declaration of results. The state submits that the threshold set is reasonable as it is less
than one -fifth of the total number of votes and is fair as it applies t o
independent candidates and unregistered political parties equally. Further, it is
submitted that 1 000 signatures would not constitute a less restrictive means to achieve
the same government purpose as the requirement for 1 000 signatures was a registration
requirement and not a contestation requirement.

The nature of the rights
[304] The rights in question form a cluster of political rights. They are indispensable
to both a functioning democracy and the dignity of those who bear such rights. I have
already dealt in some detail with their importance earlier in this judgment and no more
need be said on the issue, save to state that—

“[a] right that is of particular importance to the C onstitution’s ambition to create an
open and democratic society based on human dignity, freedom and equality will carry
a great deal of weight in the exercise of balancing rights against justifications for their
infringement.”183

The importance of the purpose of the limitation
[305] The limitation is purportedly designed to prevent frivolous contestation and the
logistical complications that arise as a result. Perhaps these purposes are legitimate.
But they are also brand new. The limitation is based largely on the assumption that the
introduction of independent candidates, without a signature threshold for contestation,
will increase the number of frivolous and unmeritorious candidates. There is simply no
evidence to suggest this is likely to happen.

183 Currie and De Waal above n 90 at 178.
KOLLAPEN J
139

[306] In fact, the evidence from past election patterns points in the opposite direction.
In its explanatory affidavit, the IEC provides an overview spanning six elections for the
National Assembly over the period 1994 to 2019. From this, it appears that the number
of political parties contesting elections increased by more than 250% from 19 in 1994
to 48 in 2019. The overwhelming majority of them – some 70.8% – were unsuccessful
in the 2019 election. Of the 34 unsuccessful parties, 22 (65% of them) obtained fewer
votes than the lowest threshold now in place for the 2024 election. In addition, there
are some 322 registered political parties all of whom would have, but for the amendment
of the Electoral Act, been eligible to contest the elections , subject to meeting only a
registration requirement of 1 000 signatures.

[307] Thus, prior to the decision of this Court in New Nation Movement and with no
contestation threshold in place, save for the registration requirement of 1 000 signatures,
the large and increasing number of registered political parties did not create a risk to the
electoral system. Previously, a relatively large number of parties have unsuccessfully
contested elections and secured relatively low voter support. However, none of this
prompted concern about a low threshol d o r the need to consider a higher
signature requirement for registration. Evidently, there was no need to do so .
Therefore, one can safely assume that , but for the decision of this Court in
New Nation Movement, the 2024 election would have been conducted without any
signature contestation threshold for political parties, save for the requirement of 1 000
signatures to be registered.

[308] Although the requirement of 1 000 signatures for registration as a political party
is, in form, a requirement for registration, rather than contestation, it has always been a
de facto contestation requirement. In terms of section 26(a) of the Act, only registered
political parties may contest elections. There would seem to be little purpose for a
political party to register except so as to have the option of contesting elections. The
requirement for registration must thus have been set by the IEC on the basis that it was
KOLLAPEN J
140
an appropriate threshold for allowing a political party to participate fully in political
activity, including elections.

[309] The Electoral Commission Act (ECA) makes this apparent.184 The very fact that
the registration of political parties is regulated by the ECA and overseen by the IEC
shows the close link between registration as a political party and election contestation.
In terms of section 15(1) of that Act, an application for registration as a political party
has to be made to the chief electoral officer, and must be made either for the entire
Republic, a particular province, or particular district or metropolitan municipality. The
proviso to section 15(1) states that a registered party’s right to contest an election is
confined to its geographic registration. In terms of section 17(1) of the ECA, the IEC
may cancel a party’s registration if satisfied, among other things, that the party has no
intention to participate in an election; or is not currently represented in a legislature and
has not participated in an election since the date of its registration or since the date it
was last represented in a legislature.

[310] This being so, the introduction and increases in the threshold requirements for
registration as a political party are revealing. In 1994, there was no signature
requirement. In 2000, regulations were introduced that required a party to have 50 voter
signatures for registration. 185 In 2008, this figure was increased to 500 for national
participation and 50 for municipal participation. 186 As recently as August 2021, the
requirement was increased to 1 000 signatures for national participation, 500 for
provincial participation, and 300 for municipal participation.187 In August 2021, when
the most recent threshold regulations were passed, the IEC knew everything about the
trends from 1994 to 2019, on which it and the oth er respondents now rely to justify a
contestation requirement of 15% of the relevant quota. In 2021, it only increased the

184 51 of 1996.
185 Regulations for the Registration of Political Parties, GN R712 GG 21386, 13 July 2000. Regulation 3 required
the party’s deed of foundation to be signed by at least 50 qualified voters.
186 Regulations for the Registration of Political Parties, GN R1204 GG 31452, 22 September 2008.
187 Proc R35 GG 45060 of 27 August 2021. The signature requirements are contained in regulation 3(1)(a).
KOLLAPEN J
141
registration requirement from 500 to 1 000 signatures, without any suggestion of a
separate contestation requirement. In 2023, just two years later, the
signature requirement has been introduced with much said about the need for a
contestation threshold.

[311] Parliament says that there was historically no need for a signature requirement
until New Nation Movement since “[political] parties have generally elected not to incur
the expense of contesting an election unless they have at least some possibility of
gaining a seat in the National Parliamen t of the Provincial Legislature ”. Parliament’s
characterisation of the politics and the economics of contestation suggests that frivolous
or meritless contestation has never been a problem with political parties. Political
parties, Parliament tells us, make sensible political and economic decisions around
whether and when to contest an election. This begs the question: why should the same
not be expected of independent candidates ? This question goes unheeded and
unanswered.

[312] The extent of the problem is un known. However, past data suggest s that the
absence of signature contestation thresholds did not result in frivolous or meritless
contestation nor were the respondents concerned as much. This leads me to the
conclusion that there is a relatively low impo rtance for the purpose of the
signature requirement.

The nature and extent of the limitation
[313] Based on the 2019 elections, the signature requirement would require that
independent candidates obtain 10 000 to 14 000 signatures from voters. 188 There are
three considerations that I wish to make in determining whether this is a serious
infringement. The first consideration is the arbitrariness of the signature requirement

188 The number of signatures required at 15% of the regional quota for the 2019 elections, as per the IEC’s
July 2022 presentation, range from: 10 271 (Northern Cape); 10 652 (North West); 11 329 (Limpopo); 1 1 340
(Free State); 11 657 (Eastern Cape); 11 925 (Mpumalanga); 13 045 (Kwa-Zulu Natal); 13 201 (Western Cape);
to 13 890 (Gauteng).
KOLLAPEN J
142
where it was based on incorrect quantifications. The second consideration is the extent
to which the requirement is out of line with foreign democracies. And the third
consideration is the assumptions regarding the prospects of winning a seat. I deal with
them in turn.

[314] The signature requirement was calculated by Parliament as a percentage o f the
votes per regional quota for the National Assembly in the 2019 election, which was
about 44 000 votes – the wrong vote quota. In actuality, the quotas were to be calculated
based on the regional quota for a seat in the National Assembly. The quota was
determined by reference to the last elections which were in 2019. They were as follows:
Region Regional Quota
Eastern Cape 77 713
Free State 75 602
Gauteng 92 601
KwaZulu-Natal 86 967
Limpopo 75 529
Mpumalanga 79 499
North West 71 016
Northern Cape 68 474
Western Cape 88 008
Weighted average 83 511

[315] The Amendment Act imposes a signature requirement of 15% of the quota for
the region in the preceding election, as per the figures in the table above. The quota in
this formula is in all instances substantially higher than 44 000 votes, which was the
figure used by Parliament in its deliberations and debates in setting the signature
threshold. Had the correct figures been utilised – and it is clear that they are the only
figures that could properly have been used, as that is what section 31B of the Act
provides for – a totally different picture would have emerged in respect of the raw
KOLLAPEN J
143
numbers required. Those numbers are illustrated below, using a rounded weighted
average of 83 500 as the regional quota:
Percent Raw number using
44 000 quota
Raw number using 83 500
quota
50% 22 000 41 750
30% 13 200 25 050
20% 8 800 16 700
15% 6 600 12 525

[316] If the percentage set at 15% is calculated based on a 44 000 quota, then 6 600
signatures are required. But, if it is calculated based on an 83 500 quota, then
12 525 signatures are required. Evidently, the differences in the signature requirements
are substantial – almost 100% higher using the correct quota. This discrepancy exists
at every poi nt the percent is set at. This is critically important and underpins the
importance of the raw numbers that would in the end be required to secure standing for
election. If the use of the 44 000 quota proved to be incorrect, then the raw numbers
change considerably upward as the table demonstrates.

[317] Simply to further illustrate the point, Counsel used the incorrect 44 000 quota in
its advice on three of the four percentage levels, as follows:
(a) When the threshold was set at 50%, Counsel calculated this to r equire
22 000 signatures. On the correct quota, it would be 41 750. Counsel’s
advice was that the substantially lower requirement of 22 000 signatures
would be an unjustifiable limitation, and no more than 20% should be
prescribed.
(b) When the threshold was set at 30%, Counsel calculated this would require
14 667 signatures (though it should have been 13 200). On the correct
quota, it would be 25 050 signatures. Counsel’s advice was that the
substantially lower requirement of 14 667 signatures would be an
unjustifiable limitation.
KOLLAPEN J
144
(c) When the threshold was set at 20%, Counsel calculated this would require
8 800 signatures. On the correct quota, it would be 16 700 signatures.
Counsel’s advice was that the substantially lower requirement of 8 800
signatures was eight times the requirement for political parties to be
registered with the IEC, and that the contestation requirement should thus
be extended to political parties not yet represented in Parliament.

[318] Ultimately, when Parliament resolved to adopt the 15 % requirement, it did so
after it had received submissions from the IEC and an opinion from Counsel addressing
the use of the incorrect quota. None of the documents, opinions or memoranda put up
by Parliament reference the correct quota in raw numbers or the correct base quota –
the regional quota. The only inference that can be drawn is that Parliament believed,
mistakenly so, that it was setting a threshold of 15% of 44 000 – which translates to a
signature requirement of 6 600 – while, in reality , it s et a signature requirement of
between 10 271 and 13 890. In many regions, this is close to or equates to 100% more
than the 6 600 contemplated. It is neither clear nor explained why Parliament used the
wrong figure. It is also not clear whether Parliament’s final resolution was on the basis
that 15% would require 6 600 signatures (on the incorrect quota) or 12 525 signatures
(on the correct quota).

[319] This is quite remarkable. The error was clear and previously identified in the
IEC’s February 2023 response to the submissions . The relevant section was even
entitled “Wrong Quota”. The correct quota was provided to the Portfolio Committee
by the IEC in July 2022 and again in February 2023.

[320] Moreover, the incorrect quota was pointed out by the applicant in its application
to this Court. In response, Parliament did not adequately address the error in
assumption. It simply said that it has no knowledge of the “erroneous assumptions”. It
was only during the hearing of the matter that Counsel for Parliament conceded that the
incorrect quota was used . U nder those circumstances , lawmakers would have
KOLLAPEN J
145
deliberated and voted on a signature requirement based on figures that were materially
incorrect.

[321] The response from Parliament is inaccurate, inadequate and fails to address the
core of the problem. The quota was almost 100% higher when using the correct quota.
This should not be a matter of speculation as the correct quota had been provided to the
Portfolio Committee in July 2022. It is not clear why the c orrect figure was not used
(between 68 474 to 92 601). In addition, it was the raw numbers that ultimately
mattered – the percentage was merely the calculating tool. There can be little doubt that
Parliament cared about the raw numbers. At each instance , Counsel’s concerns
prompted Parliament to seek a reduction in the applicable percentage. The problem was
that it consistently used the incorrect base quota . This had the effect of fundamentally
tainting the accuracy and the reasonableness of the process.

[322] In the absence of any clear motivation as to why the requirement was set at the
level it was, and given its movement from 50% to 15%, past data is again instructive.
From a 1 000 signature threshold, it is now fixed at a threshold between 10 000 and
14 000 signatures.189 Independent candidates must secure these signatures in order to
even stand for election. Parliament accepted Counsel’s view that a figure of 14 667
(in fact, 33% of the 44 000 quota, and not 30% as Counsel said) could arguably be an
unjustifiable limitation of the section 19 right to contest elections as
independent candidates. Even a threshold of 8 800 (20% of the 44 000 quota) prompted
similar concerns that, at eight times more than the 1 000 signatures political parties
required to register, the contestation requirement might be arbitrary and unfair.

[323] To overcome this concern, the registration requirement was extended to
unrepresented political parties, but this did not address the problem identified by
Parliament. It pertinently recognised that a political party would have a greater ability
to motivate voters to assist it in registering to contest elections than an independent

189 See fn 190.
KOLLAPEN J
146
would. And so simply extending an admittedly arbitrary and unfair
signature requirement to unrepres ented political parties so as to create a veneer of an
equal and non-discriminatory requirement was no solution at all. Expanding the reach
of an arbitrary and unfair requirement to others does not vitiate its capriciousness. For
these reasons, it must f ollow that the final threshold of between 10 000 and 14 000
signatures constitutes an unjustified limitation and an unfair and arbitrary one.

[324] To this extent, the absence of any evidence of reasonable steps by the applicant
to demonstrate that the attainment of the requirement is onerous or beyond their reach,
cannot assist the respondents in the justification analysis. The onus is on the
respondents to demonstrate that the justification is reasonable and necessary in an open
and democratic society. The above strongly suggests that they too saw the requirement
as being unjustified and possibly arbitrary and unfair.

[325] A second consideration under this factor is how the signature requirement
compares to other jurisdictions.

[326] In search for a comparison, the IEC relies on signature requirements in Denmark.
The IEC may be right to pick Denmark as a useful comparator. Currently, there are 16
parties represented in the Danish Parliament compared to 14 in our
National Assembly.190 As is soon to be the case here, Denmark also allows
independents to run alongside party candidates. In Denmark, voters elect
parliamentarians by choosing either independent candidates or candidates presented on
a party list. Voters are also free to cast their vote for a party without indicating specific
candidate preferences. One hundred and thirty-five parliamentarians are chosen in such
a manner. The remaining 40 seats are distributed to parties based on their percent age
of the national vote as compensatory seats, to achieve proportionality.


190 The Danish Parliament Political Parties , available at: https://www.thedanishparliament.dk/political-parties;
and Parliament of the Republic of South Africa National Assembly , available at:
https://www.parliament.gov.za/national-assembly.
KOLLAPEN J
147
[327] The IEC notes that Denmark – with an electoral system akin to South Africa’s –
required 20 194 signatures in its 2022 parliamentary elections. This is only partially
correct. While Denmark did indeed require 20 194 signatures, it only did so for political
parties.191 Independent candidates, on the other hand, “ha[d] to submit between 150
and 200 recommendations in writing from voters resident in the nomination district”.192

[328] By way of comparison, the Danish exampl e – properly construed – has a
signature requirement far lower than that endorsed by the IEC. The same is true
elsewhere.193 In our region, South Africa is an anomaly. Candidates for the Assembly
in Zimbabwe must collect only five of their compatriots’ si gnatures. In Lesotho, two
signatures are required. Zambia asks for 15. A candidate in Botswana must obtain nine
signatures. Namibia – where the figure is highest , apart from South Africa – requires
300 signatures. These figures stand in stark contrast to the 10 000 to 14 000 now needed
for independent candidates in South Africa. The table below demonstrates that many
of the world’s democracies that permit independent candidates to participate in
elections, present similarly low signature requirements for independent candidates:
No. State Structure of
Parliament Electoral System
Signature
or
nomination
requirement
1. Albania194 Unicameral Proportional 300

191 Ministry of the Interior and Health and The Danish Parliament The Parliamentary Electoral System in Denmark
(2011) at 19, available at: https://elections.im.dk/media/15737/parliamentary-system-dk.pdf.
192 Id.
193 The first judgment at [178] claims my consideration of foreign jurisdictions “violates their right to a fair public
hearing guaranteed by section 34 of the Constitution ”. This is not so. The matter of comparative jurisdictions
was traversed on the papers. It addresses precisely the legal issue before us: signature requirements for
independent candidates. Rather than representing an infringement on the parties’ right to a fair hearing under
section 34, my consideration of foreign law directly aligns with section 39(1)(c)’s encouragement to “consider
foreign law” when interpreting the Bill of Rights. Section 1(1) of the Law of Evidence Amendment Act 65 of
1988 provides in relevant part: “Any court may take judicial notice of the law of a foreign state and of indigenous
law in so far as such law can be ascertained readily and with sufficient certainty. Our courts do this often, and
this matter is no different.
194 Inter-Parliamentary Union Albania Kuvendi (Parliament) , Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2001_B.htm.
KOLLAPEN J
148
2. Australia195 Bicameral (House and
Senate)
House: majority, direct
preferential vote;
Senate: proportional
single transferable vote
House: 6,
Senate: 50
3. Botswana196 Unicameral Simple majority direct
election
9
4. Canada197 Bicameral (House of
Commons and
Senate)
House of Commons:
single-member plurality
system, first past the
post
50-100
5. Denmark198 Unicameral Proportional with
compensatory seats
150-200
6. Finland199 Unicameral Proportional 100
7. Germany200 Bicameral (Lower
House and Upper
House)
Both proportional and
first past the post
200
8. Greece201 Unicameral Proportional with some
majority vote
constituencies
12
9. India202 Bicameral (House of
the People and
Council of States)
House of the People:
majority vote
1-10
10. Italy203 Bicameral (Chamber
of Deputies and
Senate)
Chamber of Deputies:
mixed proportional and
first past the post
500-1 000

195 Inter-Parliamentary Union Australia (Senate), Electoral System , available at: http://archive.ipu.org/parline-
e/reports/2016_B.htm.
196 Inter-Parliamentary Union Botswana (National Assembly) , Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2041_B.htm.
197 Inter-Parliamentary Union Canada (House of Commons) , Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2055_B.htm.
198 Inter-Parliamentary Union Denmark (Folketinget), Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2087_B.htm.
199 Inter-Parliamentary Union Finland (Eduskunta Riksdagen), Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2111_B.htm.
200 Inter-Parliamentary Union Germany (Deutscher Bundestag), Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2121_B.htm.
201 Inter-Parliamentary Union Greece (Vouli Ton Ellinon ), Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2125_B.htm.
202 Inter-Parliamentary Union India (Lok Sabha ), Electoral System , available at: http://archive.ipu.org/parline-
e/reports/2145_B.htm.
203 Inter-Parliamentary Union Italy (Camera dei Deputati ), Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2157_B.htm; and ACE Project: the Electoral Knowledge Project Parties
and Candidates, available at: https://aceproject.org/main/english/pc/pcc05a.htm.
KOLLAPEN J
149
11. Kenya204 Bicameral (National
Assembly and Senate)
National Assembly:
majority first past the
post; Senate: first-past-
the post
Senate:
2 000
12. Lesotho205 Bicameral (National
Assembly and Senate)
National Assembly:
mixed-member
proportional
representation
2
13. Mauritius206 Unicameral Majority vote 6
14. Namibia207 Bicameral (National
Council and National
Assembly)
National Assembly:
proportional party-list
system
300
15. Zambia208 Unicameral First past the post 15
16. United Kingdom209 Bicameral (House of
Commons and House
of Lords)
House of Commons:
Single-member
plurality system, first
past the post
10
17. Zimbabwe210 Bicameral (National
Assembly and Senate)
National Assembly:
majority first past the
post
5

[329] Ultimately, Parliament chose a number out of kilter with most foreign
democracies. Perhaps this difference is best explained by the respondents’ peculiar
reason for setting the signature bar so high. Contrary to the respondents’ assertions, the
purpose of a signature requirement should not be to prejudge someone’s viability as a
candidate. Instead, as summarised by the Venice Commission, “the right to stand for
election may be subject to the support . . . of a certain number of electors to exclude

204 Inter-Parliamentary Union Kenya (Senate), Electoral System , available at: http://archive.ipu.org/parline-
e/reports/2168_B.htm.
205 Inter-Parliamentary Union Lesotho (National Assembly ), Electoral System , a vailable at:
http://archive.ipu.org/parline-e/reports/2181_B.htm.
206 Inter-Parliamentary Union Mauritius (National Assembly), Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2209_B.htm.
207 Inter-Parliamentary Union Namibia (National Assembly) , Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2225_B.htm.
208 Inter-Parliamentary Union Zambia (National Assembly) , Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2359_B.htm.
209 Inter-Parliamentary Union United Kingdom (House of Commons) , Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2335_B.htm.
210 Inter-Parliamentary Union Zimbabwe (National Assembly), Electoral System , available at:
http://archive.ipu.org/parline-e/reports/2361_B.htm.
KOLLAPEN J
150
frivolous candidates”. 211 The Commission continues that even when screening for
frivolity, signature requirements are only “acceptable if the number is sufficiently
small”.212 When compared to other nations, South Africa has introduced a requirement
both untethered from its conventional purpose and far from sufficiently small.

[330] As a final consideration, in its arbitrary setting of the signature requirement at
15% of the regional quota, the lawmaker has also made an unjustifiable assumption
about the link between th e signature requirement and the prospects of winning a seat.
In accordance with a typical South African election timetable, an independent candidate
will need to register for an election about two months before polling day. This means
that the collection of 10 000 to 14 000 signatures will have to be done several months
before voting day. At that stage, voters will not be as caught up in electioneering as
they will be later. The independent candidate effectively has to run two campaigns.
First, a preliminary campaign to garner 10 000 to 14 000 signatures; and then a second
campaign to win a seat, once their candidacy has been accepted by the IEC.

[331] An independent candidate with limited resources is faced with a quandary as to
whether to devote them to the preliminary campaign or the real thing. The fact that, in
such circumstances, the candidate may only muster one or two thousand signatures at
the preliminary stage does not mean that he or she will not do very much better once
voters become interested in issues and the candidate goes on an all-out campaign. And
their prospects might be beneficially affected by missteps by political parties in the
run-up to polling day.

[332] By confusing the purpose of signature requirements, the respondents also
disregard elections’ holistic value. Elections are not solely a matter of arithmetic – who
wins a seat and who loses. Instead, elections have value wholly separate from the

211 Venice Commission Report on thresholds and other features of electoral systems which bar parties from access
to parliament (March 2010) at 3
available at:https://www.venice.coe.int/webforms/documents/default.aspx?pdffile=CDL-AD(2010)007-e.
212 Id.
KOLLAPEN J
151
outcome of the vote. In Canada and the United States , this sentiment has been
reinforced in matters challenging restrictive ballot measures. Similarly, to the
Canadian Supreme Court,213 the United States Supreme Court expressed the dual role
of elections: “an election campaign is a means of disseminating ideas as well as
attaining political office”.214

[333] In light of these considerations, I am of the view that the signature is an arbitrary
and extensive limitation.

The relation between the limitation and its purpose
[334] Setting the requirement was no easy task, and this judgment does not suggest
otherwise. The respondents demonstrated the difficulty in striking a balance between
limiting elections to serious con tenders and setting a threshold that is not arbitrary,
unfair, or unconstitutional. To this end, it was resolved by Parliament that the
requirement would be set as a percentage of votes per seat obtained in the previous
election. While a percentage repre sented a useful and necessary tool to properly
manage the quantification of the signature requirement, the raw numbers are what
ultimately matter. T hose raw numbers that would determine whether the signature
requirement could be said to be unfair, arbitra ry or unconstitutional since the y tell us
what conditions are required to be met by a candidate seeking to access the right to
stand for public office.

[335] The initial proposal was set at 50%. There is nothing on the papers that provides
any insight as to how 50% was identified as the initial threshold. In time, the threshold
was moved to 30%, then to 20% and finally to 15%, largely out of concerns that it was
too high. I n all of this time , neither the Minister nor Parliament said what informed
their approa ch to the setting of the requir ement. In the 1 February 2023 opinion,
Counsel advised Parliament that if a candidate could not obtain the support of 15% of

213 See Figueroa above n 79.
214 Illinois State Board above n 80 at 186.
KOLLAPEN J
152
the voters for contestation, they were unlikely to obtain enough votes to win a seat.
Parliament made the same argument in its submissions before this Court. This,
however, is no explanation at all for setting the mark at 15%.

[336] As I have said, the process, the timing, and the formal requirements to get support
for a nomination are differe nt from that of getting electoral support. In addition, it
appears that 50% was an entirely random requirement to start with and its subsequent
reduction was equally random. There appears to have been little rega rd, except in
passing, for the time, effor t, costs, and logistics, which would be involved in securing
between 10 000 and 14 000 signatures in support of contestation. That is precisely why
it was changed as often and as substantially as it was – simply because there was no
way to test it except for what may have been an instinctive sense of what would
constitute too high a figure in raw numbers.

[337] There is an inexplicable gap, both in the reasoning and the process, leading to
the signature requirement. While we know it is intended to limit frivolous participation
and avoid logistical frustrations to ensure free and fair elections, we do not know why
it was set at 50% and then finally reduced to 15%. Nor do we know how the requirement
was linked to the objective it sought to achieve.

[338] I have set o ut the nature of the limitation, how the limitation evolved, and
concerns about the raw numbers used. There is a discrepancy that arises with regard to:
(a) the numbers that were used in determining whether the figure arrived at would be
acceptable; and (b) the correct numbers that should have been used.

[339] While the proposed signature requirement was to be calculated at a percentage
of the regional quota in the previous election, approximately 83 500 votes, the
compensatory quota of approximately 44 000 vot es was used. In real terms, this
emerges from opinions attached to Parliament’s answering affidavit: according to
Parliament, the raw numbers ranged from 22 000 (50%), 14 667 (30%), 8 800 (20%),
and to 6 600 (15%). This is on the basis that the base figu re to be used was 44 000
KOLLAPEN J
153
votes. By using the incorrect and low base of 44 000, the raw figures were considerably
lower than what they actually turned out to be. The raw numbers mattered and ,
ultimately, the raw numbers relied on throughout were incorrect . It is this error that
tainted the process of how the signature requirement was set. The relationship between
it and the purpose it sought to achieve was fundamentally distorted.

Less restrictive means to achieve the purpose
[340] It is noteworthy that the only signature requirement that was in place at the time
of our most recent national elections, those of 2019, was that of 500 signatures for
national registration as a political party and that this was increased to only 1 000 as
recently as August 2021. T his increase in August 2021 occurred when the IEC was
fully aware of our election trends over the period 1994 to 2019. During that period,
modest signature requirements served the electoral system well and without exception.
To this end, no consideration was given to setting a requirement fixed at 1 000
signatures, or one in proximity to that.

[341] The respondents identify the risk of frivolous contestation and logistical
difficulties in broad terms but present little evidence to support it. Even if such risks do
exist, the extent of the risks remain unknown. And it must also be borne in mind that
the signature requirement is not the only mechanism for discouraging frivolous
contestation. An independent candidate will also be required to lodge a monetary
deposit. The draft regulations promulgated for comment in July 2023 set the deposit at
R20 000. For all these reasons there was no proper evaluation of less restrictive means
to achieve the purpose.

Conclusion on justification
[342] For all these reasons, I am not satisfied that the state respondents have shown
that the limitation is reasonable and justifiable in an open and democratic society. The
applicant’s rights challenge must be upheld, entitling it to the relief it seeks on one
aspect of its challenge.
KOLLAPEN J
154

[343] New Nation Movement has led to a historic moment in the journey of our young
constitutional democracy. For the first time, independent candidates will stand for seats
in the national and provincial legislature. Parliament meets this historic moment with
a first -of-its-kind signature requirement for independent candidates. We must be
cautious and guard against such a requirement becoming a barrier to contestation. In
effect, the signature requirement renders som ewhat hollow the promise that
New Nation Movement heralded in unlocking and giving section 19 of the Constitution
its full and proper effect.

[344] There is a final issue I wish to address. Parliament created
signature requirements for both independent candida tes and unrepresented political
parties. It foisted identical restrictions onto unrepresented political parties via
section 27(2)(cB) as it did onto independent candidates via section 31B(3). Owing to
the similarity between the two sections, the amicus submitted that this Court’s judgment
will have an impact on both sections and both types of candidates. However, the
contours of this Court’s consideration are sketched by the applicant, not the amicus. A
challenge to section 27(2)(cB) was not properly before us in this matter. The general
rule is that courts should only consider the issues properly before them.215 As such,
section 27(2)(cB) falls beyond the scope of our review.

[345] While there are instances when a court may raise questions of law mero motu,
those questions may only be asked when doing so “involves no unfairness to the
parties”.216 This is not such an instance. The respondents deserve a proper opportunity
to ventilate section 27(2)(cB). For instance, although the provisions read identically,
the plausible justifications for their enactment may differ. Parliament presses a
legitimate concern: the imbalance in capacity between independents and unrepresented

215 Molusi above n 14 at paras 27-8.
216 Id; Maphango v Aengus Lifestyle Properties (Pty) Ltd [2012] ZACC 2; 2012 (3) SA 531 (CC); 2012 (5) BCLR
449 (CC) at paras 109-14; CUSA v Tao Ying Metal Industries [2008] ZACC 15; 2009 (2) SA 204 (CC); 2009 (1)
BCLR 1 (CC) at para 68; and Barkhuizen v Napier [2007] ZACC 5; 2007 (5) SA 323 (CC); 2007 (7)
BCLR 691 (CC) at para 39.
KOLLAPEN J
155
political parties may warrant disparate treatment. True or not, this Court would benefit
from argument on this point and others like it. To hold otherwise would be “unfair”.

[346] Furthermore, that the applicant has failed to challenge section 27(2)(cB)
alongside section 31B(3) should not impact the remedy I fashion. The applicant has
successfully challenged section 31B(3) as an unjustified limitation to the rights
contained in sections 19 and 18. They are entitled to the relief they seek.

Remedy
[347] The proper remedy for an unjustified infringement of a right is a declarati on of
constitutional invalidity. The applicant submits that the order of invalidity should be
suspended for a period of 24 months for Parliament to attend to the defects and that,
during the period of suspension, a limited striking -out of the 15% requirem ent should
be combined with a reading -in of 1 000 signatures as a contestation requirement.
I agree.

[348] Given the policy -laden nature of such a matter, Parliament – not this Court –
should decide the ultimate number for the signature requirement. The desig n features
of our electoral system fall squarely within Parliament’s remit. As
New Nation Movement II217 reminds us:

“Parliament’s description of the research and actions und ertaken to amend the
Electoral Act indicate just how policy -laden the legislative choices that Parliament
must make are. It is a process that requires not just the parties that are before us to
provide submissions, but also to allow other interested parties and the public to have
their say.”

[349] While the Court would not generally interve ne in such a policy decision,
practical concerns warrant an interim determination before the next general election

217 Speaker of the National Assembly v New Nation Movement NPC [2022] ZACC 24; 2022 JDR 1766 ;
2022 (9) BCLR 1165 (CC) at para 76.
KOLLAPEN J
156
between May and August 2024. There is insufficient time for Parliament to address the
matter. This is the sole reason we go beyond adjudica ting the requirement’s
constitutionality. In these circumstances, the Court must make an interim determination
until Parliament is able to set a constitutionally compliant signature requirement.

[350] In the interim, the only plausible figure is 1 000 signatur es. First, no other
alternative figure exists. Setting aside the impugned 15% requirement, the sole extant
signature requirement is the 1 000 signatures required for registration as a political
party. Second, while labelled a “registration requirement”, this figure is essentially a
threshold for allowing parties to contest elections. In other words, it is a de facto
contestation requirement. If not for New Nation Movement, the
1 000 signature requirement would have applied to unregistered political par ties
wishing to register and contest in the next election. I have referred earlier to the fact
that a low contestation requirement has not had a prejudicial effect on the management
of past elections. Under these circumstances, I would read in a
1 000 signature requirement for contestation by independent candidates.

[351] This reading-in will only apply for the period of 24 months while the order of
invalidity is suspended to afford Parliament an opportunity to properly consider the
signature requirement and r emedy the defect. The consideration of any other figure
would require this Court to embark on an exercise in speculation.

[352] This consideration must be left to Parliament to assess through its deliberative
and consultative processes during the suspension pe riod. In the event that Parliament
does not effect the required amendments to the section within the two -year period, or
any extended period that this Court may grant, then section 31B(3), as read -in, will
continue to endure either permanently or until su ch time as Parliament effects
amendments to the section.

KOLLAPEN J
157
Costs
[353] While the applicant has achieved success in its challenge to the constitutionality
of the signature requirement, the state respondents have successfully withstood the
challenge to the recalcul ation provisions in the Act. The cost order that must follow
must take cognisance of this as it must the principle in Biowatch.218 On that basis the
applicants are entitled to those costs incurred in relation to the signature challenge but
not those in relation to the recalculation challenge. They would ordinarily be liable for
those costs of the respondents incurred in successfully resisting the recalculation
challenge, but Biowatch immunises them from such an adverse costs order.

[354] That being the case and broadly quantifying the share of the two issues in the
overall costs of the application, I would say that 50% represents a fair allocation of
those costs to the signature requirement and the other 50% to the recalculation
challenge. The Atkins report was used wholly in support of the challenge on
recalculation and to the extent that that challenge was not successful the applicant
should not be entitled to recover the costs associated with that report. This is in no way
a reflection of the competence or otherwise of Mr Atkins but rather a common sense
and practical determination as to how costs are to be awarded given the conclusions
reached by this Court. I intend making an order that the respondents pay 50% of the
applicant’s costs of the application, which costs are to exclude those associated with the
report of Mr Atkins. Given the truncated timeframes and the complexity of the matter,
the costs of two Counsel are warranted.

Order
[355] I make the following order:
1. The applicant is granted direct access to this Court.
2. The recalculation relief sought in prayers 4, 6.2 and 6.3 of the applicant’s
notice of motion is refused.

218 Biowatch above n 76.
KOLLAPEN J
158
3. Section 31B(3)(a)(i) and (ii) of the Electoral Act 73 of 1998 ( Electoral Act)
as inserted by the Electoral Amendment Act 1 of 2023 is declared invalid and
inconsistent with the Constitution, to the extent that it unjustifiably limits the
rights to freedom of association, freedom to make political choices and to
stand for public office.
4. The declaration of invalidity ref erred to in paragraph 3 is suspended for
24 months from the date of this order to afford Parliament an opportunity to
remedy the constitutional defects giving rise to the constitutional invalidity.
5. From the date of the order of this Court and during the period of suspension,
section 31B(3)(a)(i) and (ii) of the Electoral Act will read as follows, the
underlined words being read into the section with the words in strike-out text
being severed:
“(3) The following must be attached to a nomination when it is submitted:
(a) A completed prescribed form confirming that the independent
candidate has submitted, in the prescribed manner, the names,
identity numbers and signatures of voters whose names
appear—
(i) in the case of an elect ion of the
National Assembly in respect of regional
seats, on the national segment of the voters’
roll and who support his or her candidature ,
totalling 1 000 signatures for each region in
which the candidate intends to contest an
election;
(aa) totalling 15 percent of the quota for
that region in the preceding election,
if intending to contest only one
region; or
(bb) totalling 15 percent of the highest of
the regional quotas in the preceding
election, if intending to contest more
than one region, provided that where
15 percent of the highest of the
KOLLAPEN J / THERON J
159
quotas is not achieved, that the
independent candidate may only
contest the region or regions as
determined by the next highest quota;
(ii) in the case of an election of a provincial
legislature, on the segment of the voters’ roll
for the province and who support his or her
candidature, totalling 1 000 signatures least
15 percent of the quota of that province in the
preceding election, which the independent
candidate intends to contest,
provided that an independent candidate who was elected to
either the National Assembly or a provinc ial legislature as an
independent candidate in the preceding election shall be
exempt from this requirement.
6. In the event that Parliament does not remedy the constitutional deficiency in
section 31B(3)(a)(i) and (ii) within the period provided for in paragraph 4 of
this order, or any extended period grante d by this Court, then
section 31B(3)(a)(i) and (ii) will be deemed to read as set out in paragraph 5
above.
7. The second, fourth and fifth respondents, jointly and severally, are to pay
fifty percent (50%) o f the applicants’ costs , which costs shall include the
costs of two Counsel. The applicant is not entitled to recover any costs
associated with the report of Mr Atkins.



THERON J (Rogers J concurring):


Introduction
[356] This is an application for direct access to this Court by One Movement
South Africa NPC (OSA or the applicant ). The applicant challenges the
constitutionality of the Electoral Act (Act) as amended by the
THERON J
160
Electoral Amendment Act (Amendment Act) on two grounds. First, the applicant
contends that the Amendment Act has introduced a signature requirement which acts as
an impermissible and arbitrary barrier for independent candidates to contest elections.
Second, the applicant contends that the recalculation method introduced by the
Amendment Act is unconstitutional in that it results in an unfair advantage for political
parties with a larger share of the vote (large political parties) and a disadvantage for
political parties with a smaller shar e of the vote (small political parties) as well as for
independent candidates.

[357] I have read and considered the judgments of my brothers, the Chief Justice
(first judgment), and Justice Kollapen (second judgment). I agree with the
first judgment insofar as it holds that direct access should be granted on the
signature challenge. I agree with the second judgment to the extent that it finds that the
signature requirement is unconstitutional in terms of section 31B(3)(a)(i) and (ii) of the
amended Act. I therefore concur in the order of the second judgment insofar as it relates
to the signature requirement . And, given the conclusion I reach on the recalculation
challenge, I also agree with the second judgment’s order on costs.

[358] The purpose of this judgment i s to briefly express my concerns about the
constitutionality of the recalculation method. The facts have been adequately set out by
the first two judgments and I do not propose to repeat them except where necessary.

[359] A vacancy can occur in the National As sembly or provincial legislature for the
following reasons:
(a) when an independent candidate gains sufficient votes to be awarded
two or more seats and thus has to forfeit the seats in excess of the one
he or she is able to take up;
(b) when a party forfeits seat s by virtue of supplying fewer candidate
names than the number of seats won in a particular election;
THERON J
161
(c) when an independent candidate is elected to both the National
Assembly and provincial legislature and thus has to forfeit a seat in
one or other of the legislative bodies;
(d) when an independent candidate gains election in more than one region
for the National Assembly and thus has to forfeit the regional seats in
excess of the one he or she is able to take up;
(e) when an independent candidate vacates a seat thro ugh death or
resignation; and
(f) where a party represented in a legislature is dissolved or ceases to
exist.

[360] When a vacancy arises in respect of a candidate from a political party, the party
concerned will fill the seat by nominating a person from the party according to item 22
of Schedule 1A of the Act.219 This Court in New Nation Movement found that the Act
was unconstitutional to the extent that it prevented adult citizens from standing for
election to the National Assembly and provincial legislatures as
independent candidates.220 An independent candidate can only ever hold one seat, and
they cannot nominate a representative to fill a vacancy if they cease for any reason to
hold a seat. Therefore, a recalculation is necessary in order to re -allocate the seat or
seats that were initially allocated to an independent candidate.

219 Item 22 of the amended Schedule 1A, which is a repeat of item 23 of the unamended Schedule 1A, says the
following:
“(1) In the event of a vacancy in a legislature of a seat allocated to a party, the party which
the vacating member represented must fill the vacancy by nominating a person—
(a) whose name appears on the list of candidates from which that party’s
members were originally nominated; and
(b) who is the next qualified and available person on the list.
(2) A nomination to fill a vacancy must be submitted to the Speaker of the legislature in
writing.
(3) If a party represented in a legislature dissolves or ceases to exist and the members in
question vacate their seats in consequence of section 47(3)(c) or section 106(3)(c) of
the Constitution, the seats in question must be allocated to the remaining parties with
the changes requir ed by the context as if such seats were vacated seats in terms of
item 23 and item 24, as the case may be.”
220 New Nation Movement above n 3 at para 120.
THERON J
162

[361] Items 5, 7, 11, 12, 23 and 24 of the amended Schedule 1A of the Act deal with
recalculation in the event of a vacancy. The applicant contends that these items are
unconstitutional to the extent that they disproportionately favour large political parties.
Both the applicant and the respondents agree that the recalculation method benefits
large political parties and thus skews proportional representation. However, the
respondents assert that the recalculation method still complies with the constitutional
injunction that the electoral system must result “in general, in proportional
representation” as required by sections 46(1)(d) and 105(1)(d) of the Constitution.

The recalculation method
[362] Parliament is empowered to determine how elections are to be run. It is not for
the courts to determine if there are better ways of conducting elections or if the methods
chosen by Parliament are reasonable.221 Courts must be mindful of the fact that
elections are a highly technical, policy -laden and polycentric matter. A court must be
careful not to attribute to itself superior wisdom in relation to a responsibility entrusted
to Parliament.222

[363] However, this does not mean that this power is unfettered. Pa rliament’s
decisions are subject to judicial scrutiny: they must be rational and cannot limit any
fundamental rights enshrined in Chapter 2 of the Constitution, 223 unless the limitation
can be justified in terms of section 36 of the Constitution.

[364] As the second judgment makes clear, the regulation of the exercise of a right
which goes beyond regulation, and which has a limiting effect, will amount to a

221 Id.
222 Bato Star Fishing (Pty) Limited v Minister of Environmental Affairs and Tourism [2004] ZACC 15;
2004 (4) SA 490 (CC); 2004 (7) BCLR 687 (CC) at para 48.
223 New National Party above n 16 at paras 19-20.
THERON J
163
limitation.224 When a measure limits a right, a court is required, in terms of section 36
of the Constitution, to conduct a justification analysis in order to determine whether the
limitation is reasonable and justifiable. It is not the case that any law that purports to
regulate the exercise of a right is shielded from constitutional scrutiny .225 To ho ld
otherwise would be untenable.

[365] The recalculation method is required to strike a balance between allowing
independent candidates to contest elections and ensuring that the electoral system
results, in general, in proportional representation. 226 The respondents contend that the
“options to address the concerns raised in respect of vacancies were limited” and that
the fact that the system favoured large political parties was “simply a logical
mathematical consequence”. They also contend that the Constitutio n, as confirmed by
this Court in New Nation Movement, does not require the electoral system to result in
perfect proportionality.

[366] The applicant argues that Parliament should adopt a recalculation method that
awards the vacant seat to the candidate who has the highest number of remainder votes
in terms of the initial allocation phase (highest remainder method). According to the
applicant, this is far more representative and democratic. The respondents claim that
the highest remainder method also distorts proportionality.

Does the recalculation method infringe section 19 of the Constitution?
[367] It is common cause between the parties that the result of the recalculation is that
votes cast for large political parties count more than votes cast for small political parties
or independent candidates.


224 Second Judgment at [230].
225 Id at [242].
226 Sections 46(1)(d) and 105(1)(d) of the Constitution.
THERON J
164
[368] The figures provided by Mr Atkins demonstrate this. For illustration , I have
inserted two graphs from the Atkins report. The lighter bars represent the vote gain
after recalculation. The graphs, which reflect a com parison for parties that obtained
seats in the 2019 provincial elections in the Eastern Cape and Free State, clearly
illustrate that large political parties have a larger vote gain than small political parties
in the recalculation. The graphs for the othe r provinces exhibited a similar pattern.
Essentially, this means that large political parties are advantaged in each recalculation.


[369] The applicant argues that the gains of each party in the recalculation should be
in proportion to their initial vote share.227

[370] I now consider whether the recalculation method limits the rights of voters,
independent candidates and small political parties under section 19 of the Constitution.

[371] In my view, the recalculation method limits the norm that voters should have an
equal say as to who will represent them in Parliament. The principle of one person one
vote is foundational to our representative and participatory democracy. It is an inherent
component of the right to free and fair elections. This principle was espoused in a report
drafted by the Electoral Task Team, appointed by Cabinet to draft legislation for the
2004 elections, as follows:

227 “Vote share” is the proportion of the total votes cast that is received by a particular party.

THERON J
165

“Taking its cue from the Constitution, the majority felt not only that every eligible voter should
have the opportunity to vote but that, as far as possible, all votes should be of equal value. This
was the understanding of proportionality ‘in general’, where every vote has some relevance in
the composition and membership of the national and provincial legislatures. Fairness also lies
in the closeness of the relationship between votes cast and the composition of the body
elected.”228 (Emphasis added.)

[372] In Doctors for Life ,229 Sachs J stated that it is i mportant in a democracy that
equality of the vote is complemented with equality of voice. Sachs J references
Dworkin in “Equality, Democracy and Constitution : We the People in Court ” where
Dworkin argues that:

“In a genuine democracy, the people govern not statistically but communally . . . .
When we insist that a genuine democracy must treat everyone with equal concern, we
take a decisive step towards a deeper form of collective action in which ‘we the people’
is understood to comprise not a majority but everyone acting communally.”230

[373] The importance of the equality of the vote, given South Africa’s historical
dispensation cannot be overstated. The vote was used in apartheid to silence and denude
people of their dignity. Every person should feel that they are “part of the body politic
with the full civic dignity that goes with citizenship in a constitutional democracy”.231
I clarify my apprehension with reference to the formula itself.

[374] Parties are awarded seats by way of quota seats and remainder seats. If the quota
is reduced, which generally occurs in recalculation, then either the number of quota
seats is increased or the number of remainder votes is increased. The quota reduction
is due to independent candidates no longer being included in the recalculation.

228 Electoral Task Team Report of the Electoral Task Team (January 2003), available at https://pmg.org.za/policy-
document/346/.
229 Doctors for Life International v Speaker of the National Assembly [2006] ZACC 11; 2006 (6) SA 416 (CC);
2006 12 BCLR 1399 (CC) at fn 10.
230 Dworkin “Equality, Democracy and Constitution: We the People in Court” (1990) 28 Alberta Law Review 324.
231 Doctors for Life above n 233 at paras 19-20.
THERON J
166
Sometimes, this may result in small political parties or independent candidates gaining
a seat. However, the Atkins report demonstrates that the recalculation generally results
in large political parties gaining seats. It also means that large political parties’ vote
gain after recalculation is not in proportion to their initial vote share. This presents a
difficulty not only because it creates a bias for large political part ies, but also because
it fails to actualise the intention of the voter.

[375] When participating in the national elections, there are a range of available
options to a voter: they can choose not to participate at all; they can spoil their ballot or
they can vote for a candidate or party. If they do chose to vote, that vote has a twofold
purpose. First, they cast a vote towards a particular party or individual. Second, they
cast a vote to effect the overall election results. For example, one can choose to not
vote for the dominant political party to ensure that the seats awarded in Parliament are
distributed more evenly. This is a strategic decision that a voter can make to strengthen
our multiparty democracy. It also is doubtful that when a voter votes for an independent
candidate they are cognisant of the fact that their vote will favour large political parties
in the event of recalculation.

[376] Insofar as small political parties and independent candidates are concerned, they
do not benefit equally from the re calculation. In general, large political parties have a
larger vote gain than small political parties. The result of the recalculation is that there
is a disparity between the vote gain after recalculation and the vote share. This disparity
must inevitably raise a constitutional concern. In my prima facie view, the result of the
recalculation method is that small political parties and independent candidates’
section 19 rights are infringed. This is because favourable treatment would be a breach
of the right of small political parties and independent candidates’ right to stand for
public office in free and fair elections. This Court has alluded to this, albeit in a different
context, where it said:

“It was wisely not suggested that political parties wo uld be treated more favourably
than independent or unaffiliated candidates. That would be a clear and unjustifiable
THERON J
167
breach of the constitutional right of such candidates to stand for public offi ce in free
and fair elections.”232

[377] I am also concerned that th is limitation is not justifiable. The second judgment
neatly highlighted the importance of the rights under section 19 of the Constitution. It
is therefore not necessary to reiterate this. I am alive to the difficult task Parliament
was faced with and a lso the fact that perfect proportionality in an electoral system is
impossible or near impossible to achieve. However, the respondents failed to justify
this limitation.

[378] Generally, the legislature or the party relying on the legislation must justify the
limitation.233 This Court in Moise v Greater Germiston Transitional Local Council put
the matter thus:

“It is also no longer doubted that, once a limitation has been found to exist, the burden of
justification under section 36(1) rests on the party asserti ng that the limitation is saved by the
application of the provisions of the section. . . . The obligation includes not only the submission
of legal argument but placing before court the requisite factual material and policy
considerations. Therefore, although the burden of justification under section 36 is no ordinary
onus, failure by government to submit such data and argument may in appropriate cases tip the
scales against it and result in the invalidation of the challenged enactment.”234

[379] The respondents have not justified why Parlaiment chose this recalculation
method over the highest remainder method. On the evidence, Parliament seemingly
failed to consider other methods or the implications of the chosen method. The
respondents failed t o provide an adequate explanation as to why the votes cast for

232 Kham v Electoral Commission [2015] ZACC 37; 2016 (2) SA 338 (CC); 2016 (2) BCLR 157 (CC) at para 76.
233 S v Makwanyane above n 88 at para 102.
234 Moise v Greater Germiston Transitional Local Council: Minister of Justice and Constitutional Development
Intervening (Women’s Legal Centre as Amicus Curiae) [2001] ZACC 21; 2001 (4) SA 491 (CC); 2001 (8) BCLR
765 (CC) at para 18.
THERON J
168
independent candidates should not continue to influence the outcome in the
recalculation.

[380] As the second judgment points out, elections serve a variety of different
purposes.235 They are about m ore than merely securing office – they are also about
expression. A vote for an independent candidate is not necessarily only a vote for that
candidate. It can also be a vote which rejects the other parties or candidates. These
votes should therefore potentially influence the outcome in recalculation.

Does the recalculation method infringe upon sections 46(1)(d) and 105(1)(d) of
the Constitution?
[381] Section 46(1)(d) of the Constitution states that the National Assembly is made
up of members who are elected in terms of an electoral system that “results, in general,
in proportional representation”. In terms of section 105(1)(d) of the Constitution, the
provincial legislature is constituted of members elected in an electoral system that
“results, in general, in proportional representation”. It is clear that what is required is
not “perfect” proportionality. These sections recognise that perfect proportionality is
hard or near impossible to achieve , particularly when independent candidates are
introduced, and so Parliament is afforded some leeway in designing the electoral
system. Therefore, some deviation from perfect proportionality is not per se
problematic. However, it is the fact that the recalculation method favours large political
parties that is of concern.

[382] Further, i t is not sufficient for Parliament to deviate from proportionality, as
provided for in sections 46(1)(d) and 105(1)(d) of the Constitution, without providing
sufficient reason for this. The respondents appear to justify the deviation on the basis
that it is “a logical mathematical consequence” that the system favours large political
parties. I find this unconvincing. It bears emphasis that the respondents have failed to

235 Second Judgment at [213].
THERON J
169
produce any evidence to demonstrate why the favouring of large political parties is
inevitable.

[383] However, the Atkins report shows that this bias often only results in one or two
seats being allocated to large political parties. It is not possible, for a number of reasons,
to determine whether this is trivial or strays too far from the requirements of
sections 46(1)(d) and 105(1)(d) of the Constitution . Before we can rule the current
recalculation method unconstitutional, we must be sure that there is an alternative
method that would not have its own pitfalls. The applicant’s proposal for remedy has
its own difficulties. While the matters raised by Atkins are legitimate concerns, it is not
possible, in my view, to find that the current recalculation method is unconstitutional.
The urgent circumstances in whi ch the case was litigated has not allowed this to be
sufficiently explored. During the hearing of this matter, the applicant’s Counsel made
no oral submissions on the point, as a result of which the respondents only touched on
it in passing. The Court did not have the benefit of a proper engagement on the issue in
oral argument. It is thus not in the interests of justice to decide this challenge.

Conclusion
[384] It is clear that the first judgment centralises the rights enshrined in section 19 of
the Constitution. This is a red herring. Although the matter evidently triggers those
rights, it rests on accountability, a founding value of our Constitution. More precisely,
it is about Parliament’s duty to justify its decisions when those decisions limit rights
contained in the Constitution. In the same manner that our political rights have found
new meaning since 1994, so too has Parliament’s duty to justify its decisions. As
Etienne Mureinik aptly put it—

“If the new Constitution is a bridge away from a culture of authority, it is clear what it
must be a bridge to. It must lead to a culture of justification – a culture in which every
exercise of power is expected to be justified; in which the leadership given by
THERON J
170
government rests on the cogency of the case offered in defence of its decisions, not the
fear inspired by the force at its command.”236

I am not convinced that Parliament has adequately adhered to these prescripts.

[385] However, for the reasons I have stated, I do not consider that we can do justice
to the recalculation challenge by way of the present urgent application for direct access.
Accordingly, if I had commanded the majority, I would have declined to grant urgent
direct access in respect of the recalculation challenge. This does not affect my
agreement with the second judgment’s costs order, since both on my approach and on
the second judgment’s approach the applicant would not have achieved success on the
recalculation challenge.


236 Mureinik “A Bridge to Where? Introducing the Interim Bill of Rights.” (1994) 10 South African Journal of
Human Rights at 32.


For the applicant:



For the second respondent:

AA Gabriel SC, MZ Suleman,
S Mdletshe and CJ Moodley instructed
by Thompson Wilks Incorporated.

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

For the third respondent:



For the fourth and fifth respondents:

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


K Pillay SC, LJ Zikalala and G Solik
instructed by State Attorney,
Cape Town.

For the amicus curiae:





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