top of page

Proposed electoral model for South Africa

A response to the Constitutional Court judgement declaring the current electoral legislation unconstitutional

Copyright © 2021 Inclusive Society Institute 50 Long Street Cape Town, 8000 South Africa 235-515 NPO All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means without the permission in writing from the Inclusive Society Institute DISCLAIMER Views expressed in this report do not necessarily represent the views of the Inclusive Society Institute or its Board or Council members. Cover photograph:

Setting the scene

In the light of the recent Constitutional Court judgement declaring the current Electoral Act unconstitutional, the Inclusive Society Institute (ISI) embarked on a process to design a potential new electoral model for South Africa.

The institute appointed an expert panel, convened by Mr. Roelf Meyer, to undertake the work and mandated the panel to design an electoral model that will meaningfully give effect to the judgement, respect the boundaries set out in the Constitution, retain proportionality as a basis for representation in that it best promotes inclusivity, and which enhances representativity, accountability, and transparency.

The court has given the legislature 24 months to introduce new legislation that will enable independent candidates to stand for election in the national and provincial spheres of government. The institute has prioritised the development of proposals in this regard so as to, in an effort to enrich the public dialogue, timeously provide its findings to the political stakeholders and public policymakers, of potential solutions. This report will also form the basis of the institute’s own advocacy in the legislative process that lies ahead.

The institute wishes to acknowledge the work of the panel and thanks its members who generously gave of their time and expertise. The members were:

  • Mr. Roelf Meyer: In Transformation Initiative. Chief Government negotiator during the democratic transition in South Africa, former Minister of Constitutional Development, and currently a director of In Transformation Initiative.

  • Ms. Deyana Isaacs: University of Stellenbosch. Researcher and lecturer in Political Governance at the School of Public Leadership, University of Stellenbosch.

  • Prof. Dirk Kotze: University of South Africa. Professor in Political Sciences at the University of South Africa (UNISA), a Vice President of the International Political Science Association and National Secretary of the South African Association of Political Studies.

  • Prof. William Gumede: University of the Witwatersrand & Democracy Works. Professor at the University of the Witwatersrand School of Governance and Chairperson of Democracy Works Foundation.

  • Mr. Ebrahim Fakir: Auwal Socio-Economic Research Institute. Political commentator and Director of Programmes at Auwal Socio-Economic Research Institute. Former head of political parties and parliamentary programme at the Elecoral Institute of Southern Africa (EISA), researcher at the Institute for Democracy in South Africa (IDASA) and the Centre for Policy Studies (CPS). Visiting fellow at the Institute for Development Studies at the University of Sussex for 2006 and visiting Draper Hills Summer Fellow at Stanford University for 2011.

  • Ms. Dren Nupen: The Elexions Agency. Former Executive Director of the Electoral Institute of Southern Africa and former Regional Director for Africa at the Open Society Initiative. She brings extensive election experience to the dialogue.

  • Prof. Firoz Cachalia: University of the Witwatersrand. Professor at the Law School at the University of the Witwatersrand.

  • Professor emeritus Jørgen Elklit: Aarhus University, Denmark. Member of the South African 1994 IEC and of the 2002-2003 Electoral Task Team (aka the van Zyl Slabbert Commission). 1996-2009, a member of the board of directors of EISA. Secretary to the 2008 Independent Review Commission in Kenya. 2013 recipient of IFES’ Joe Baxter Award.

  • Prof. Cherrel Africa: (Flexi-member) University of the Western Cape. Formerly at the Institute for Democracy in South Africa, election analyst for the SABC and ENCA and currently associate professor and chair for political sciences, University of the Western Cape.

  • Prof. Rassie Malherbe: University of Cape Town. Currently offers an extra-curricular programme in the drafting of legislation at the University of Cape Town. Former professor of public law and Head of the Department of Public Law at the University of Johannesburg.

  • Mr. Grant Masterson: Programme Manager: Africa Peer Review Mechanism at the Electoral Institute of Southern Africa. He has

  • lectured at the University of Witwatersrand on international relations and the international political economy.

  • Mr. Daryl Swanepoel: Chief Executive Officer of the Inclusive Society Institute and former Member of Parliament.

Acronyms and abbreviations

CC Constitutional Court

ETT Electoral Task Team

FPTP First past the post

MMC Multi-member constituency

MMP Mixed-Member Proportional (system)

NCOP National Council of Provinces

IEC Independent Electoral Commission

PR Proportional representation


Setting the scene

Acronyms and abbreviations

Chapter 1: The Constitutional Court judgement

1.1 Unpacking the Constitutional Court judgement declaring the Electoral Act


1.2 Discussion on the judgement and preferences related to new models that need to be


1.3 Assessing and determining the degree of electoral reform that is required

1. 4 Conclusion

Chapter2: Interrogating options for a new South African electoral system

2.1 Introduction

2.2 Minimalist approach

2.3 Multiple modelling approaches regarding multi-member constituencies and PR balancing


2. 4 Hybrid parallel electoral model combining majority ‘first past the post’ with closed list

proportional representation

2.5 Additional points to be considered in developing a new electoral model

2.6 Conclusion

Chapter 3: A proposed new electoral system for South Africa

3.1 Introduction

3.2 Specifics of the proposed electoral system

3.3 Illustration of the outcome of proposed electoral model based on the 2019 national


3. 4 Ancillary matters to consider

3.5 Testing the new system against the guiding principles and criteria

3.6 Question and answer guide to understand the proposed system


Annexure A

1. Introduction

2. Findings of Van Zyl Slabbert Commission & Presentation on the Danish electoral model:

Presented by Prof Emeritus Jørgen Elklit, Aarhus University, Denmark

2.1 The Van Zyl Slabbert Commission

2.2 Danish Electoral Model

3. German mixed member proportional (MMP) system: Presented by Prof Michael

Krennerich, University of Erlangen-Nurnberg

4. The Irish system - Proportional representation by the single transferable vote: Presented

by Prof David Farrell, University College Dublin

5. Spanish electoral model: Presented by Prof Ignacio Lago, University Pompeu Fabra,


6. Turkish electoral model: Presented by Prof Ali Çarkoglu, KOÇ University, Istanbul, Turkey

Chapter 1: The Constitutional Court judgement

1.1 Unpacking the Constitutional Court judgement declaring the Electoral Act


On 11 June 2020, the Constitutional Court declared the following:

“4. It is declared that the Electoral Act 73 of 1998 is unconstitutional to the extent that it requires that adult citizens may be elected to the National Assembly and Provincial Legislatures only through their membership of political parties. 5. The declaration of unconstitutionality referred to in paragraph 4 is prospective with effect from the date of this order, but its operation is suspended for 24 months to afford Parliament an opportunity to remedy the defect giving rise to the unconstitutionality.”

Upon assessment, the judgement does not provide extensive guidance but is focused on the shortcomings or deficiency of the Electoral Act with regards to independent candidates. Moreover, the judgement also does not provide guidance as to how the act should or can be amended. The Electoral Act is subject to the Constitution, and therefore the task of the panel must be accomplished without breaching the constitutional boundaries relating to electoral systems.

The following provisions in the Constitution can be considered boundary posts regarding the amendments that can be made to the Electoral Act:

Section 1(d): Multi-Party democracy

Section 1 in the Consitution is the value clause that speaks of multi-party democracy. The Constitutional Court was requested to consider this clause in respect of independent candidacy in elections. Its view on the matter, however, is that independent candidates

do not stand in the way of or impede on the functioning of multi-party democracies. This is considered a boundary post, as the amendments that give effect to the judgement may not undermine this value clause. It is also of importance since this value clause is in the

Constitution to safeguard South Africa from becoming a one-party state.

Section 19(3)(a): Right to vote

This section is normally seen as comprising four aspects:

  • General - For all who qualify according to the basic requirements to vote.

  • Equal - To ensure that some votes do not weigh more or less than others, which has an implication for the type of electoral system that is established and how, in this case, independent candidates can be accommodated.

  • Direct - Every vote has a direct influence on the outcome.

  • Secret - Every voter votes voluntarily, without coercion, intimidation or duress and in secret.

Therefore, equality of the votes forms another boundary post that affects the amendment. An electoral system cannot be established where some votes outweigh others. Independent candidates cannot be accommodated in such a way that the votes brought out for them weigh more or less than for others.

Section 19(2):

Right to Free, Fair and Regular Elections is of importance as it is the basic political right contained in the Bill of Rights. Three aspects are contained in this section:

  • Regular elections - Legislature has a fixed term, enforcing that there will be elections at regular intervals.

  • Free - To participate without interference or coercion and to ensure that candidates and political parties must be able to participate.

  • Fair - Equal opportunity for all parties or candidates to contest the elections.

These aspects will have a bearing on the type of amendments made to the electoral system.

Sections 46 and 105: Electoral system that results, in general, in proportional representation

Section 105 refers to the provinces.

Proportional Representation (PR) is not defined in the Constitution. However, it may include: Proportional Representation of political parties and Proportional Representation reflecting voter preference.

There are many PR systems that give effect to both PR for political parties and voter preference such as list systems, preferential systems, and combined systems. The current South African electoral system is based on party list proportional representation, which

means that parties are represented in proportion to their electoral support. The preferential system entails that the voter can indicate preference for a party and candidates on the ballot. The ballot can list party representatives in order of preference and can also move over party lines in this system. The combined systems incorporate PR with constituencies.

Therefore, section 46 and 105 of the Constitution serves as another boundary post within which amendments to the system must be made. The system chosen must result, in general, in proportional representation. Additionally, within the Constitution and, consequently,

because of the court judgement, Parliament assumes the responsibly to choose a particular electoral system. The particularities of the system are left to Parliament and it has some leeway on how to give effect to the judgement.

Section 19(3)(b): Right to stand for and if elected to hold public office

This is the provision that the Constitutional Court judgement focused on, and forms the main grounds for the Electoral Act being declared unconstitutional. Within the current electoral system, the exercising of the section 19(3)(b) right can only be channelled through political parties. Therefore, the act does not provide for candidates to participate in elections as individuals or independent candidates. This is the crux of the judgement and what the majority of the courts’ ruling dealt with.

Furthermore, Justice Mandlanga added another aspect, which can also be considered a boundary post, namely: section 18, Freedom of Association. Justice Mandlanga’s argument was that this also includes the freedom NOT to associate. Under the current Electoral Act,

candidates are “forced” to join a political party in order to be a candidate and exercise their right under 19(3)(b), which negates the right NOT to associate while simultaneously seeking to stand for representative political office. Therefore, the Electoral Act violates this right.

Furthermore, the court also dealt with a number of other provisions that were put to the court, including:

  • Sections 46(1)(a) AND 105(1)(a): The electoral system is prescribed by national legislation. The national legislature (Parliament) is subject to the Constitution, thus whatever the national legislature decides, its decisions are still subject to the Constitution.

  • Sections 47(3)(c) AND 106(3)(c): Loss of membership of legislatures. The Constitutional Court said that this only applies to political parties and does not affect the issue of independent candidates.

  • Sections 57(2), 70(2), 116(2), 178(1)(h), 193(5) AND 236: Participation by parties in legislatures.

  • 57(2): Deals with the rules and orders of the National Assembly providing for minority parties to participate in the proceedings.

– Section 70(2): Rules and orders of National Council of Provinces providing for


– Section 116(2): Rules and orders of provincial legislatures providing for


– Section 178 (1)(h): Participation of opposition or minority parties in the process

for the appointment of members of the Judicial Service Commission.

– Section 193(5): Deals with chapter 9 institutions where minority parties will also


– Section 236: Legislation to be made for the funding of political parties.

All these provisions are about political parties’ participation in legislatures. The court found that these provisions strengthen multi-party democracy and do not prevent making provision for independent candidates to participate. These are boundaries which dictate that whatever amendments are to be made; they cannot undermine the participation of parties in the legislature.

  • Section 157(2)(a): The municipal election system can either be a pure PR system according to the list system, where only parties are being presented on the ballot or a PR system combined with wards. This section does not affect the national and provincial spheres. Thus, this only affects the municipal level and does not prevent the court from declaring the Electoral Act unconstitutional.

  • Section 6, Items 6(3)(a) and 11(1)(a): Nominations by Political Parties. This only refers to the first elections after the 1996 Constitution came into effect and is thus not applicable anymore and has no current influence.

In conclusion, all the above listed sections are boundary posts that the Constitution imposes on the Electoral Act and the amendments brought to the act. These can be viewed as the most important ones within which the amendment of the act must take place.

Below is a diagramme representing the constitutional boundary posts.

1.2 Discussion on the judgement and preferences related to new models that need to be


“Proportional Representation (PR), in general” was an important point of discussion. The panel was of the view that “in general” was an important qualification that needed to be taken into account when considering how to give effect to the court ruling as it pertains to the inclusion of independent candidates. It was adamant that PR, as stipulated in the Constitution, must continue to remain a prominent feature of the new electoral system that will provide for the participation of independent candidates.

Absolute proportionality under any system does not exist. It is for this reason, the panel believes, that the term “in general” has been used. More often than not, political parties fall short of the votes to secure the next member to represent them in the legislature. The surplus is then allocated elsewhere and, accordingly, the end result is not 100% proportional. That said, the panel noted that, currently, South Africa has the most proportional electoral system in the world, due to it not having an electoral threshold for a party to be represented in the legislature and due to the size of the National Assembly.

Proportional representation means that parties get a certain number of seats in the legislature in accordance with the percentage of votes that they received in an election. So, for example, if a party gets 15% of all the votes in the country, then it gets close to 15% of the seats in Parliament. In South Africa, there are 400 seats in the national parliament, so for every 0.25% of the vote a party gets, in principle, one seat.

Consequently, when independent candidates are incorporated into the new electoral system, the requirement pertaining to PR, should, as far as practically possible, be the same for them as for the political parties. If an independent candidate gets a sufficient number of votes to qualify for a seat in the National Assembly or Provincial Legislature, the candidate will be regarded as a “type of party” for purposes of distributing seats. The independent candidate will therefore require an equal number of votes as would a party for gaining a seat in the legislature. The challenge for independent candidates will be to mobilise the minimum number of votes needed to gain a seat in the legislature. Campaign rules must therefore also make provision for independent candidates to share in the public resources made available for purposes of gaining exposure.

The key point of PR is that the composition of the legislature must reflect, in a proportional manner, the preferences of the voters, as expressed in the number of votes cast in a particular election.

Regardless of the electoral system proposed by the panel, there were several issues that needed to be considered. South Africa’s current electoral model, at the national level, results in the constitution of a National Assembly with 400 members. The National Assembly is one of the two houses of Parliament, the other being the National Council of Provinces (NCOP). When designing the new electoral system to give effect to the inclusion of independent candidates, the following will need to be considered:

Some of the necessary considerations when implementing a PR electoral system, where the size of Parliament is fixed and incorporating independent candidates includes the following:

  • The size of the legislature is fixed.

  • Should individual independent candidates or lists of independent candidates be allowed in both houses?

  • Should the system allow voters to cast their votes for one (or more) individual candidates (so-called personal or preferential votes) or shall it only be possible to vote for the list as such?

  • What will the ballot paper access requirements be pertaining to:

– Parties elected in previous parliaments and still represented in Parliament? For

example, direct access, a deposit, or a certain number of seconding voters.

– New parties: Payment of deposit or a certain number of seconding voters?

– Independent candidates (or lists of independent candidates, if allowed):

Payment of a deposit or a certain number of seconding voters?

  • Will electoral thresholds be different or the same for parties and individual independent candidates, for example, based on a certain percentage of the vote, an absolute number of votes or some other criteria?

Consideration to these key decisions were given by the panel, the outcome of which will unfold later in the report.

And in addition to the procedures to elect representatives, the introduction of independent candidates may very well have consequential implications for a number of other pieces of legislation. For example, the argument has been made that if individual candidates are to be accommodated in the PR system, then apart from the Electoral Act having to change, the Political Party Funding Act has to be amended to also require independent candidates to disclose the funding and support they receive.

So too, the national and provincial legislatures may, as a consequence of the judgement, have to re-think how they function now that independent candidates may take up seats. The rules and procedures of the legislature may require amendments, as may the constitutional provisions providing for minority party participation in the business of the legislatures, which will have to be expanded to include independent candidates. For example, section 57.2 of the Constitution, which provides for the rules and procedures of the National Assembly to be made with due regard to the participation of minority parties. It will have to be expanded to include independent members. Similarly, section 61, dealing with the allocation of delegates to the National Council of Provinces, may have to provide for a formula broader than just parties in order to accommodate independent candidates.

To illustrate: the current rules of the National Assembly under section 57(2)(b) of the Constitution, indicate that the representation of political parties in the structures of the legislature, for example committees, must also accommodate minority parties. This is

proving difficult in that it has not been possible to accommodate every political party represented in the legislature on every one of its committees. This will prove more difficult with the introduction of independent candidates. Moreover, the text on party representation

as it now stands (minority parties), may in future not be adequate to provide for the participation of independent candidates.

In discussing the desired preferences and guiding principles that a new electoral model would need to aspire to, the panel articulated a number of core essentials that they sensed needed to be included in the new system. It should provide a greater ability for voters to directly elect candidates, and to exercise an influence over how party candidates are elected. The current system is lacking in this regard.

Likewise, the new electoral system should, in their view, continue to encourage diversity. Demographic, ethnic and religious diversity and inclusivity should be encouraged through promoting a diversity of candidates.

Other features, the panel believed necessary, were that the system should be relatively simple to administer and easy for the voter to understand. And there had to be a meaningful balance between accountability, responsiveness and representativity.

To achieve this, it will require bold and comprehensive reforms that are in the public interest, more so than in the interest of political parties or personalities. Therefore, the panel established the following principles that they considered necessary to guide the design

of a new electoral system:

  • Representation should be translated into legislative seats and needs to represent the expressed will of the voters. Representation can take the form of geographical representation, where the voters in each region, town, city, province or electoral district choose the candidates who will represent them, and to whom the candidates should ultimately be accountable. The national legislature should be a mirror of the nation and its ideological diversity. To achieve this, some form of link with a particular constituency needs to be established.

  • Transparency should exist in both the process of electing the representatives as well as in the overall electoral system itself.

  • Inclusiveness: The electoral system should enable as many voters as possible to participate in the voting process. Inclusion will be promoted through a system that is easy to understand and easily accessible to all voters.

  • Accountability to the electorate is considered fundamental and needed to be improved. In the current system accountability to the voter was traded for party compliance.

1.3 Assessing and determining the degree of electoral reform that is required

In order to determine the degree of electoral reform that is desired, and the options that can be explored in considering aforementioned constitutional boundaries and guiding principles, the panel analysed the strengths and weaknesses of the current electoral system.

Strengths included the emphasis on promoting multi-party politics through the inclusion of as many political parties as possible. This provided wide-ranging diversity in terms of political ideology, demographics, ethnicity, religious beliefs, amongst others. This was important given the country’s history and need for healing, reconciliation, and nation-building.

A further strength of the current system is that it provides the possibility for minority parties to secure seats, which is important to promote inclusion. Moreover, the system is relatively easy to administer and for voters to understand. And it is fair and just in that every vote carries equal power.

One weakness, some argue, is that voters are represented by parties and not individuals. This reduces the ability of the voter to hold individual representatives accountable. Moreover, since the current electoral system operates through a closed party list system, voters are unable to express their personal preferences regarding the candidates standing for election. Neither can they effectively influence these appointments, nor recall ineffective, or rogue representatives. Whilst the current system does allow voters to raise their complaints and concerns with party structures, to date this has proved to rarely influence action against non-performers.

Another weakness is that the voters do not have a direct influence over who will become their representatives in the legislatures. In fact, neither do they have a say as to the election of the president, as this is carried out by the National Assembly. The panel decided, however, that it would not explore the presidential aspects of electoral reform, since its focus was on giving effect to the Constitutional Court ruling.

That said, since the voter votes for a party and not an individual, the reality is that the electorate has no influence over which individuals will be appointed to the legislature. It is the domain of the party leadership and party organs.

It is the aforementioned, many argue, that has led to the ruling party appointing candidates in legislatures that rarely question the executive, lest they be removed or sanctioned by the party. As a consequence, the legislature risks becoming a lame-duck, which leads to a trust deficit and loss of public credibility. This compels extra-parliamentary politics, forcing, in a sense, civil society and the courts to become involved in what should rightfully be the domain of the legislature.

These challenges have resulted in large numbers of voters becoming disillusioned, with many citizens feeling politically disenfranchised, choosing rather not to participate in formal politics, but to engage in extra-parliamentary activism. This is particularly worrisome amongst

the youth who have become disinterested in party politics and participation in the formal multi-party electoral processes.

In addition to the strengths and weaknesses analysis, the panel also considered some of the practical considerations that should guide their thinking regarding a new electoral system.

Considering that the designing of a new electoral system takes time to devise, the 24-month time limit stipulated by the court suggests that immediate electoral reform should focus mainly on giving effect to the requirement of accommodating independent candidates in the election processes. Wider and more far-reaching reform would require more time and should therefore be considered only in the longer term.

Another consideration must be cost. Far-reaching reform could potentially be complicated to administer, and difficult for the voter to understand. This will require greater budgetary resources to implement and to carry out voter education. Recognising that by limiting

the extent of reform, problems with the overall system would remain, the panel nevertheless felt it prudent to limit the current reform to creating room for independent candidates to participate. However, where consequential improvements could be made as a result of

system changes needed to include independent candidates, they should be taken on board.

Thus, of the three reform options identified by the panel, the first two could be considered.

  • Option 1: Make minimal technical adjustments to the current system only to allow for independent candidates to be added as candidates. This option would not address the weaknesses in the current system.

  • Option 2: Make targeted design changes to the proportional representation system which will give effect to the court ruling and go some way in dealing with the weaknesses of the current system.

  • Option 3: An entire overhaul of the electoral system.

The panel was of the view that a minimalistic change allowing only for independent candidates to be added to the ballot papers under the current system, would not be practical. Imagine adding say 100 independent candidates to the current 48 parties. That would mean an unwieldly ballot paper containing 148 options to vote for.

Accordingly, the panel opted for Option 2.

1.4 Conclusion

In designing a new electoral system to give effect to the Constitutional Court judgement requiring the Electoral Amendment Act to be amended to allow for independent candidates to stand for election at national and provincial level, the panel concluded that the new system should:

  • Give effect to the Constitutional Court judgement

  • Fit within the current boundaries prescribed by the Constitution

  • Be guided by the principles of representativity, transparency, inclusiveness and accountability

  • Be relatively easy to administer, and easily understood by the electorate.

The panel’s deliberations to this point provided them with the necessary background and information to enable them to proceed to the next step – that is the design of the system.

But before they did so, they received several presentations on different electoral models from various jurisdictions, including Germany, Spain, Denmark, Ireland and Turkey. All of these models combine PR with the right of independent candidates to stand for election. They also received a presentation on the electoral models contained in the 2003 Van Zyl Slabbert Commission on Electoral Reform Report. A summary of these presentations are contained in Annexure A of this report.

Chapter 2 that follows, interrogates options for a new electoral system for South Africa, whilst Chapter 3 provides the panel’s proposed model.

Chapter 2: Interrogating options for a new South African electoral system

2.1 Introduction

During the next phase of the expert panel’s deliberation, they started to conceptualise a unique South African electoral system that may give effect to the court judgement. Various panellists presented their initial thinking with regard to potential electoral models for South Africa. The merits and specifications of each of the proposed models were interrogated and discussed.

This chapter offers a summary of the various electoral models presented by the panellists, as well as additional issues identified during the dialogue.

2.2 Minimalist approach

A preliminary minimalist approach suggests that the current PR closed list system should be maintained, while independent candidates would be accommodated either individually or in a different form. Therefore, this model should remain within the current boundaries of the Constitution, not requiring any amendments.

It therefore satisfies the ruling of the Constitutional Court which requires provisions to be made within the electoral system, on both national and provincial level, to accommodate independent candidates. This model therefore calls for the most basic amendments to only make provision for independent candidates, without changing the electoral system in any extensive manner. The electoral system can in time be further improved to mitigate against the limitations and weaknesses of the current system, including issues such as a lack of accountability, representativity, weak parliamentary oversight and the effects and consequences of the closed list system. The PR closed list system will be used to accommodate the independent candidates, simply by amending the Electoral Act to include

a provision for the participation of independent candidates. These candidates would, in a sense, be considered in the same manner as smaller political parties, with the same rules and processes applying. Consequently, the question of how to maintain overall proportionality,

would be resolved.

It should be noted that there are, however, still a few issues relevant to this proposed model that will require attention. Firstly, due to the fractious nature of current South African politics, the possibility exists that an increased number of smaller parties and independent candidates resolve to contest elections. In order to manage this, a reasonable threshold for access to the ballot paper would have to be considered, and, in this regard, a 1% threshold of registered voters is the suggested approach.

Parties and independent candidates would therefore have to demonstrate that they have this level of support by providing a required number of verified signatures from voters. The required number of seconders will depend on whether parties and independent candidates

are running in provincial or national elections, respectively. This, however, would place an additional administrative burden on the Electoral Commission of South Africa (IEC), who would be responsible for verifying seconding signatures and ensuring that there is no abuse of

the system. An assessment will have to be made as to the capacity of the IEC to undertake this task and whether they would be able to adapt their systems accordingly. Concomitant additional resources may have to be allocated by Treasury. On the downside, the introduction

of a threshold requirement may be interpreted as a way of undermining the constitutional value of inclusivity.

Secondly, the minimalist approach may be considered a more viable and implementable system over shorter a period. The danger of this approach is, however, that it could allow for too many parties and independent candidates and it may become too difficult and unwieldly a system to manage and control. With this in mind, a few minimal requirements should be set in order to limit the number of candidates on the ballot paper. Reaching agreement on these requirements could be complicated and time-consuming.

On the upside is that the model suggests that it could be relatively easy to enforce gender parity by requiring parties to effect a quota within their lists. The same does however not hold true for independent candidates, as each candidate will only represent him- or herself.

2.3 Multiple modelling approaches regarding multi-member constituencies and PR balancing


During the panel’s discussions, the multi-member constituency and PR compensatory/balancing list appeared to be a preferred model suggested by most individual panellists.

The Van Zyl Slabbert Commission on Electoral Reform (the Electoral Task Team [ETT]) reviewed the PR closed list system after the

1999 national and provincial elections. The report was published in January 2003 and proposed a number of electoral reforms and electoral models. This model is based on the ETT majority recommendations for a preferred electoral system for South Africa, with some adjustments made.

This model proposes that an electoral system should guard against excessive fragmentation as it could lead to difficulty in the composition of government and the functioning of Parliament. This model promotes accountability to the electorate as it pertains to the representation in national and provincial legislatures. Oversight will be improved, since MP’s carry constituency mandates. And this model will not require significant constitutional amendments. In designing the model, consideration was also given to, amongst others, political party and independent candidate’s threshold, registration qualification, distortion of proportionality, and prevention of candidates being elected without real support.

The electoral model proposes that the National Assembly consist of 300 representatives elected from multi-member constituencies (MMCs). A further 100 compensatory seats at national level will be used to ensure overall proportionality. The constituency or PR split

can be adjusted slightly, nevertheless the overall number of seats should not exceed 400. The constituencies will need to coincide with metropolitan and district municipal borders. Small districts can be combined with neighbouring districts to ensure workable units. And metropolitan areas could be subdivided along sub-council or regional lines for similar purpose. The aim would be to have approximately 70 MMCs with three to seven seats in each MMC.

To ensure that parties or independent candidates with significant numbers of votes have a chance of winning a seat, each MMC must have at least three seats, but not more than seven. The number of seats will be based on the quota of registered voters in the particular MMC. The quota will be calculated according to the number of registered voters in the country, divided by 300 (the number of MMC seats nationally). As it stands, the number of current registered voters is 26 756 649, which divided by 300 seats, suggests the current quota

to be 89 189. Therefore, an MMC will have approximately 270 000 to 624 000 voters, that is an MP to voter ratio of around 90 000.

Under this MMC system, the voter votes for either a candidate, through an open list, or the party as such. Consequently, voters only vote in the multi-member constituency. The political parties’ overall share of all 400 seats will be calculated based on the total sum of votes received for all parties’ candidates across all MMCs. Proportionality will be restored through the use of the PR lists.

Both independent candidates and candidates representing political parties can stand in constituencies. The IEC will be responsible for the preparation of ballot papers, which will be based on the list of names of candidates received for each MMC. To fill the political party’s compensatory or national seats, a national list with up to 100 candidate names is required. Political parties may also nominate more than one candidate per constituency, provided that the number of candidates does not exceed the number of seats in that constituency plus one. This will allow for sufficient reserves, should vacancies need to be filled from the list.

With regard to the allocation of seats under this system, it is suggested that as an alternative to the quota system for allocation of the MMC seats, the Sainte-Laguë divisor system be considered. Under this divisor system, the sum of votes for each party and the votes for independent candidates is divided first by 1, then 3, and finally 5 in a 3-seat MMC. The three highest quotients (in declining order) are allocated a seat each. In a 7-seat MMC, the divisors are again 1, 3, and 5 – and then 7, 9, 11, and 13. The Sainte-Laguë system is generally considered a very fair system, securing a high level of mathematical fairness as good quota systems also do. But the Sainte-Laguë system is less dependent on the comparison of decimals in determining the eventual allocation of seats.

In the MMC, a “natural” threshold is at play. In a 7-seat MMC, a party or a candidate with more than 7.15% of the vote is guaranteed a seat. In the 3-seat constituencies the natural threshold guaranteeing a seat is 16.67% (see explanation in Chapter 3). The actual thresholds might be lower than the percentages prearranged, depending on the actual vote distribution among parties and independent candidates, nonetheless the values mentioned will guarantee a seat. If an independent candidate obtains more votes than required for election, the surplus of votes is, however, discarded.

With regard to independent candidates accessing the ballot under an MMC system, candidates will be required to demonstrate that they have voter support in the chosen MMCs. This they do by submitting signatures and voter registration numbers of seconding individuals that are registered in the specific MMC. A threshold of 1% can be considered as a reasonable number. For example: in a 3-seat MMC the number of signatures might amount to between 2 500 and 3 000 signatures and in a 7-seat MMC the number will be between 5 000 and 6 000 signatures. It should be mentioned though that it may be much easier for independent candidates to qualify to participate in the election, than to actually win a seat.

Political parties elected in the previous election and which are still represented in the National Assembly, are automatically eligible to participate in the election. These parties may submit lists of names in a few or all MMCs and have their share of the 100 compensatory seats. Other (new) political parties may become eligible by submitting names and voter registration numbers of a substantial number of voters. Parties who are not able to collect substantial voter support (and thus demonstrate that they actually have support on the ground), should not be allowed to participate. The verified number of supporting voters should be equivalent to the valid votes in the previous election divided by the number of seats (400). That can be said to be the price of one seat, which is currently 43 591 signatures. Voters can only second one party (or one independent candidate) prior to each election. The IEC will have to verify the number of voters’ support. A monetary fee should not be levied to qualify for electoral participation.

The use of only two ballot sheets is proposed. The first ballot sheet will be allocated for the constituency (MMC) list for the National Assembly, with the names of the political parties fielding candidates in the MMC as well as the independent candidates marked as such. The second ballot sheet will consist of the constituency (MMC) list of names for the Provincial Legislature.

Consequently, if this model is implemented, the following needs to be considered:

  • All advantages and disadvantages of an open or closed list need to be measured. The question to be answered is whether the country is ready for an open list system. The possibility to continue with a closed list system for national and provincial elections and only move to an open list system at a later stage should also be considered.

  • Should candidates be allowed to stand for election both in an MMC and be on the proportional list?

  • What needs to be done if vacancies cannot be filled from the political party’s candidates list?

  • Should the quota system or Sainte-Laguë divisor system be used?

  • The number of seconding signatures required for both political parties and independent candidates must be determined.

  • If floor crossing is allowed, should it only be from a political party to independent candidate? And if so, what impact will it have on the composition of the National Assembly in terms of party proportionality? If yes, this may break the stranglehold that political parties have over the MPs, but it should eliminate opportunism. Similarly, an independent candidate may come to the conclusion that his or her views can best be promoted through joining one of the parties, but, again, this may promote opportunism.

2. 4 Hybrid parallel electoral model combining majority ‘first past the post’ with closed list

proportional representation

This proposal is made bearing in mind the need to overhaul South Africa’s election system to improve representativity, oversight, accountability and responsiveness in the governance architecture.

The proposal recognises that each electoral system has distinct advantages and disadvantages. For instance, pure proportional representation systems account for every single vote in determining the outcome of an election, but do not embed the potential

for closer constituency responsiveness and accountability. In addition, pure proportional systems often provide inordinate power to political party elites to determine policy.

Pure First-Past-the-Post systems (winner takes all), on the other hand, bear the potential for greater accountability to constituencies, allow ordinary members of political parties and back-bench legislators greater influence in policy by virtue of the constituencies they command, but are only rarely a fair reflection of the choices made by an electorate. In this system winners of an election by very slim margins take all the power, with all the other votes for candidates who might lose by small margins, being discarded.

Mixed systems may minimise disadvantages and maximise advantages, but depending on the mix used, can potentially create such overwhelming systemic complexities that they are rendered indecipherable to citizens. In addition, they create serious complications in the management and administration of elections as well as the tabulation of results. This can cast a pall of doubt on the credibility of electoral processes and consequently serve to delegitimise the electoral outcomes, as oversight over and transparency of the calculation and counting of outcomes are rendered ever more complicated. Simplicity of the electoral system should not be under-estimated as a great virtue.

A hybridised parallel system envisages a simple First-Past-the-Post system requiring a 50% +1 majority to have won the seat for 350 or 400 seats in a 600-member Assembly. Alongside it, would be 250 or 200, closed party list PR seats. This would combine the benefits of Proportional Representation with single-member constituency representation.

This proposal is made, bearing in mind the principles of:

  • Accountability, responsiveness and openness – Act 108 of 1996, 1(d)

  • Human dignity, equality and freedom – Act 108 of 1996, 7(1)

  • Effective Choices – Act 108 of 1996, 19(1), 19(1)(c), 19(3)(b)

  • Maximising Representation. Preferably at its most basic and decentralised level – Act 108 of 1996, 42(3)

  • Responsibility and Responsiveness

  • Inclusivity, diversity and representativity

  • Proportionality

  • Oversight and Accountability (in executive) – legislative relations, between party representatives, party organs, leaders and members, and finally between elected public representatives and the voting public/communities/constituencies.

  • Simplicity and Transparency for voters, parties, candidates, and election management and administration.

The model would necessitate that section 46(1) of the Constitution be amended to read “National Assembly consists of no fewer than 525 and no more than 600 women and men…” and there be an amendment to section 46(1)(d) to read: “reflects an element of proportional representation in its result”.

It also envisages a unicameral Parliament with a single National Assembly and no second chamber. The NCOP would then disappear and all constitutional references to it, including in the legislative process, would be amended accordingly. If it is to be retained, its size would need to be reduced dramatically to at least half.

The model envisages retaining the Provincial Legislatures in their current form, with reduced size and with the oversight powers maintained. Their legislative powers would largely remain intact, except to expand their scope with respect to the legislative process.

In orientation, the model seeks to harness the benefits of proportional representation without making it the primary feature of the system, and privileges would increase responsiveness, accountability and a closer fit between constituencies and elected representatives, while simultaneously sublimating the overweening and inordinate influence, extra ordinary power of political party elites and party bosses. While parties will no doubt remain hugely influential, as will party elites, this mode does provide greater leverage – even if only in theory – to voters in communities and constituencies.

In this hybrid parallel system proposed, of the 600 MPs, 400 would be elected from 400 single-member constituencies in a majoritarian First-Past-the-Post (FPTP) system, with a 50%+1 majority.

Constituencies would correspond to the current, approximately 234 municipalities and their municipal boundaries (excluded are the 44 district municipalities that don’t exist as constituencies but as coordinating structures of two or three individual municipalities).

These current municipal boundaries would then be reduced further, to enable the creation and delimitation of constituencies of a reasonable, meaningful and manageable size for effective representation. This delimitation and demarcation exercise would best be left to the Demarcation Board, using the Demarcation Act’s current provisions to determine and delimit constituency sizes according to a balance between land mass, population density and concentration of services. This can be augmented if necessary. Of course, gerrymandering of boundaries may be an issue, but the oversight and transparency safeguards in the current act and in electoral administration with regard to inspection, and objections and remediation of delimitations are sufficiently robust.

This approach is premised on the idea that multi-member constituencies may still be too large and might fail to adequately address the element of responsiveness and to some extent accountability, since the constituencies would not only remain too large, but responsibility may be shirked, from among the multiple members in the multi-member constituencies.

The remaining 200 seats in the National Assembly would be elected from a Proportional List, with an appropriate number of PR seats to be allocated for election in each province, based on population size, as is the current practice. This would be implemented to establish

a modest degree of proportionality, by electing 200 MPs from the existing PR list without it resulting “in general”, in a proportional outcome for the whole election result.

For greater proportionality in the overall outcome, the system could be changed for a more balanced distribution, such as 350 MPs from constituencies and 250 from a PR list, to have greater proportional representation, bearing in mind that this then increases constituency sizes and therefore may affect responsiveness adversely.

It is also proposed that the model be replicated at Provincial Legislatures. It is recommended that the size of each Provincial Legislature be reduced by between 20% and 25%. That means each Provincial Legislature will be faced with a reduction of between three seats at the smallest and 16 at the largest, and it is estimated that on a combined 20% reduction across Provincial Legislature sizes, there will be at the least 80 seats reduced, available for absorption nationally.

The total seats to be elected in the Provincial Legislatures will be on the basis of 2/3rd of seats elected directly in constituencies, and 1/3rd off PR lists. The size of constituencies in the provinces combine two national constituencies into one, for Provincial Representation purposes in the Provincial Legislature.

Consequently, consideration would have to be given with respect to costs.

First, with no NCOP or second chamber, or at least drastically reduced second chamber, those costs would be allocated and absorbed in additional NA seats. Second, the Provincial Legislatures will be reduced by at least 80 seats nationally (based on a 20% reduction), which will also be absorbed nationally. Third, there is a definite need to moderate the remuneration, benefits and perks accruing to public representatives at national and provincial levels. Adjustments here should not be hard to deal with and should not prove unpopular

with the public.

In this system, each voter would be given four ballots:

  • One for the National Constituency Representative

  • One for the National PR Representative

  • One for the Provincial Constituency Representative

  • One for the Provincial PR Representative

While this may prove administratively and logistically burdensome for Election Administration, it is not insurmountable with proper planning, supervision and management. Its greatest advantage is that it retains the elements of simplicity and trust in the system for voters and in the process of voter registration, voting, counting, tabulation, audit and verification processes.

This model consequently seeks to address the issue of inclusivity, by allowing independent candidates and to be elected from a mixed system, while addressing issues of accountability and responsiveness through maintaining the majority of political parties.

2.5 Additional points to be considered in developing a new electoral model

As mentioned previously, additional points for consideration were raised through individual contributions by members of the panel during the discussions on the initial proposals for electoral reform. The following are worth recording:

Some panellists emphasised that caution should be taken not to create new electoral challenges through amendments to the current system.

  • Any new electoral model should be simple to administer and the principles of inclusion should be the central aim.

  • It was also stressed that threshold changes should be considered attentively.

  • Political parties play a crucial role in fostering accountability and accessibility for voters, therefore mechanisms to enhance internal democracy and the representative composition of parties should be addressed.

  • These mechanisms should include opening the list process and requiring gender representation.

  • Furthermore, improved access to the system could be established by introducing automatic registration when identity documents are applied for.

Reference was given to a report by Steven Friedman entitled: “The system’s not to blame? Electoral systems, power and accountability”. It deliberates the subject of the deficit of accountability in South Africa’s electoral systems, which has been a significant point of discussion for the panel. The issue of accountability through geographical representation has been demonstrated at municipal level, where the electorate votes for a specific ward candidate. It has, in Friedman’s view, not improved accountability. It is therefore questionable to assume that accountability will automatically prove more efficient at the provincial or national level when it fails at municipal level.

Nevertheless, the inadequate accountability of elected representatives is a widespread concern. Many attribute this to the current electoral system. But electoral reform on its own may not be sufficient to address the problem as it is embedded in other root causes as well. The suggestion that accountability could be enhanced purely by increasing the voters’ capacity to choose and rank candidates is thus debatable. To the contrary, it may also serve to weaken political parties and legislatures, as well as to contribute to the fragmenting of the political system, which may open possibilities for more political corruption. This view was also held by Rosenbluth and Shapiro (2018) in their book: “Responsible Parties: Saving democracy from itself”.

Consequently, whilst provision could be made for electing individual party candidates (from an open list), the extent to which they are accountable to the electorate versus their nominating parties will need to be considered. This will include the extent to which parties may sanction their members and the right for elected representatives to change party affiliation. However, this discussion is not new in South Africa.

A constant theme throughout the panel’s deliberations has been the desirability to allow all political parties, including smaller parties, the possibility of having representatives elected to the national and provincial legislature. This despite the concern that a growing number of parties, and now including independent candidates, contesting the elections can present a logistical challenge for the IEC. It may also be frustrating for the voters. Nevertheless, in the panel’s mind, smaller parties, in the South African context, cannot be excluded, as they have offered a political home to diverse segments of the electorate that do not feel accommodated within the bigger political parties. To address the divided society, South Africans should be offered a choice to vote for these smaller parties. In light of the aforementioned, it is recommended that the call for the introduction of a reasonable threshold be considered, that balances the aforementioned with the need to eliminate opportunistic parties and individuals that do not enjoy real support on the ground.

There was also a view that democracy works best with strong, disciplined and programmatic political parties, who are able to offer large groups of voters clear choices and incentives to coalesce. Political parties are essential vehicles for collective action and democratic accountability. Thus, the system should guard against over- fragmentation which undermines this. That said, mechanisms should be introduced to promote greater levels of democracy within the parties, as representatives should be empowered to exercise their conscience and be responsive to the electorate.

For this, transparency is key. And the closed list system inhibits transparency. Therefore, the candidate list should, ideally, be opened. Should the closed list system be adopted, however, the names of candidates must be well publicised so that the electorate know who will be representing them. And it empowers them to exercise sanction over non-performing or disreputable candidates. This is especially important in the MMCs and quite practical given the small number of candidates being put forward in each MMC. It is also suggested that political parties reveal their presidential and provincial premier candidates to the voters before the election, as opposed to announcing their decisions after the fact.

Additional proposals emanating from the discussions included the following:

  • The Van Zyl Slabbert Commission on Electoral Reform (the Electoral Task Team [ETT]) report should be used as a baseline for developing a new electoral model.

  • Specific affirmative proposals should be made to ensure gender representation, as the introduction of multi-member constituencies could potentially have an adverse effect on gender representation.

  • The current closed list system does not necessarily have to be immediately abolished as political parties could in the interim choose to introduce a system of primaries for their own candidates. Once the electorate has become accustomed to the new system, further reform could allow for the opening up of the lists.

2.6 Conclusion

This chapter represents the collection of ideas for reforming the electoral system in South Africa, as introduced by the panellists. Whilst the main objective is to design a system that gives effect to the Constitutional Court’s ruling to allow independent candidates to stand for election at national and provincial level, they recognised that such a step will require a new model. A minor tweaking of the current system will in all probability be too unwieldly; and may not give effect to the court’s underlying intention for the system to provide a reasonable and fair chance for independent candidates to be elected.

In developing the new model, panellists considered it prudent to also use the opportunity to address some of the flaws in the current system, insufficient accountability, representativity, and transparency being key, amongst others.

In the next chapter, the panellists’ combined thinking culminates in a proposed new electoral model for South Africa.

Chapter 3: A proposed new electoral system for South Africa

3.1. Introduction

This chapter contains the culmination of the panel’s thinking as informed by the discourse between the members of its expert panel and the presentations of the international contributors. The criteria set for the system’s design included that it:

  • accommodates independent candidates

  • adheres to the constitutional prescript requiring the composition of the legislature to reflect, in general, proportional representation

  • requires no or minimalistic amendments to the constitution

  • be simple for the Independent Electoral Commission to implement and for the voters to understand

  • promotes gender parity within the legislature

  • promotes demographic and geographic inclusiveness

  • promotes representativity

  • promotes, to a certain extent, accountability.

In essence, the panel needed to decide between three broad approaches:

  • A winner take all, pure constituency approach

  • A simplistic proportional representation model

  • A hybrid model that accommodated constituencies together with a compensatory proportional list allowing for overall proportionality to be established

The pure constituency approach was rejected out of hand as its design would not make it possible to meet the constitutional prescript of the outcome to reflect, in general, proportionality. It would also not accommodate sufficient diversity within the legislature. The simplistic proportional representation model – a single proportional list at the national level and nine proportional lists at the provincial level – would be impractical to administer. Imagine the length of a ballot paper should, say 100 independent candidates wish to stand in addition to the 48 existing political parties. It would also not necessarily advance geographic representation and will do little to improve voter representativity and accountability to the voters.

Thus, the model being proposed in this report is a 400-seat National Assembly of which 75% (300) of the seats are allocated to multi-member constituencies (MMC) comprised of three to seven members per MMC. The constituencies will be supplemented by a proportional list of 25% (100 seats), which will be used to ensure overall proportionality, in general, in terms of the total number of votes cast for parties in the election.

The key tenets of the system are set out hereunder. For purposes of explanation, only the election to compose the National Assembly is described, although in reality nine provincial legislatures will be simultaneously composed. The elections for each of the provincial legislatures will, however, follow the same methodology.

Furthermore, the panel did not consider detailed structuring of the NCOP, suffice to agree that it should be retained and, in the main, retain its existing construct. It should be borne in mind that delegations in the NCOP represent provinces, not individual constituencies.

Accordingly, independent candidates will not be able to be elected to a permanent seat in the NCOP, as these representatives are appointed by parties based on their relative strength in each of the provinces. Rules can, however, be provided for within the provincial legislatures to allow for some flexibility for independent members to be nominated as one of the special delegates deployed by the legislature in terms of Section 60 of the Constitution to deal with matters before the NCOP affecting the province.

3.2 Specifics of the proposed electoral system

3.2.1 Structure

There would be two components to the establishment of the legislature. The first would be representatives elected via MMCs to the legislature (with 66 MMCs of three to seven members each). The second component would comprise representatives elected via a compensatory PR list (which ensures that the political parties will be represented proportional to their share of the overall national vote).

3.2.2 Number of seats

There will be 400 seats, with 300 seats allocated to 66 MMCs and 100 compensatory seats, which will be used to ensure overall proportionality based on votes cast for each of the parties.

It is necessary to have at least three members per MMC to promote diversity within each of the MMCs. Too large a number would be counterproductive in terms of promoting geographic representativity, or for bringing the representatives closer to the electorate. Accountability is strengthened when the representatives are closer to the electorate. It should be borne in mind, however, that given the closed list system being proposed, the electorate will to a lesser extent be able to h