Why South Africa’s electoral reforms are neither inclusive nor fair

South African President Cyril Ramaphosa/CREDIT: EPA-EFE/Yeshiel Panchia
Facebook Twitter Email

In June, 2020 the Constitutional Court in South Africa ruled in favour of the applicants, the New Nation Movement NPC and others, 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”. Parliament was given 24 months to remedy this defect.

Many have emphasised the importance of the 2024 election. The ANC may be in danger of losing an absolute majority in Parliament, albeit that none of the larger opposition parties shows substantial growth. There is also the prospect that Gauteng at least will be ruled by a coalition government. In a time of upheaval, both within the ANC, and in the broader society, any change to the electoral system could play an important role in shaping the future trajectory of the country. The dynamics of inter- and intra-party power relationships are particularly touched by the prospect of independent candidates, and by the possibility of introducing some form of constituency system.

Elections are the lifeblood of any democracy, and in a new democracy, the electoral system is of particular significance in shaping the success and sustainability of democratic outcomes. It is perhaps a little-known fact that the purely proportional system employed in the 1999 election was not intended to be the final electoral arrangement for the new South Africa. A major review of our electoral system was undertaken by the Electoral Task Team chaired by Frederik van Zyl Slabbert in 2003, which recommended a multi-member constituency system. This was echoed in the High Level Panel on the Assessment of Key Legislation (page 524), chaired by former President Kgalema Motlanthe in 2017.  Despite these reviews, it has taken a court ruling to bring Parliament to the point of amending our electoral system. However, even this process has proved to be and problematic.

In March, 2021, the Minister of Home Affairs convened a Ministerial Advisory Committee (MAC), chaired by former Minister of Constitutional Affairs, Valli Moosa, to, “help develop policy options on the electoral system that address the defects of the Electoral Act 1998”. The Report was submitted in June, 2021. This report presented two options, being a “minimalist” system where independent candidates are placed alongside parties for seat allocation calculations, and a hybrid system balancing 200 single-member constituencies with proportional representation.

Beyond the MAC, three substantive proposals have been advanced, each describing a variation on the principle of multi-member constituencies:

  1. Electoral Laws Second Amendment Bill, 2020 (MGP Lekota)
  2. Helen Suzman Foundation
  3. Inclusive Society Institute

The MAC presented two options as they were unable to agree on a recommendation to advance. Those in favour of the minimalist option were, Pansy Tlakula, a former Chairperson of the Independent Electoral Commission (IEC), Michael Sutcliffe, a former Chairperson of the Municipal Demarcation Board, and Norman du Plessis, a former Chief Operating Officer of the IEC (who was responsible for setting up our electoral systems during the 1990s).

It is difficult to imagine a more qualified trio than Tlakula, Sutcliffe, and du Plessis, and yet this makes their proposal all the more astonishing. Proportional representation, by definition, is an electoral system for political parties, and thus inherently excludes the direct participation of individuals. While hybrid systems are common, with separate individual and proportional elections held in conjunction, there is no jurisdiction in the world where individuals are treated as parties in proportional calculations, except for the narrow circumstances of multi-member constituencies that typically have between three and seven members.

In November, 2021, a draft Bill commissioned by the Minister of Home Affairs was published, based essentially on the “minimalist” option set out by the MAC. On January 10, this Bill was presented to Parliament for the first time. Following a period of public consultation, the Home Affairs Committee must now consider submissions, and prepare a final draft of the Bill to put to Parliament. At the time of writing, no meeting of the Home Affairs Committee is scheduled.

What the Bill proposes is that independent candidates for the National Assembly contest 200 seats, distributed among the nine provinces (appearing on the ballot only in their home province), with political parties contesting the 200 compensatory seats separately. For provincial legislatures, independent candidates contest elections directly alongside parties. In each case, obviously, regardless of the number of votes secured by an independent candidate, only a single seat can be awarded.

The fundamental problem of the system set out in the Bill is that there is inherently going to be a significant variation between the share of seats gained by independents collectively, and their collective share of votes (precisely because a successful candidate can achieve voter support equivalent to multiple seats, but can only be awarded a single seat). When seats are allocated to parties, all seats won by independents, and all votes cast for independents, are subtracted from the totals before seat allocation calculations are carried out. If there is a substantial variation between collective seat share and vote share for independents, subtracting these automatically means that there is a corresponding variation for parties, where seat share noticeably exceeds vote share. This disproportionality becomes a real problem in respect of the requirement within sections 46(1)(d) and 105(1)(d) of the Constitution, where it is specified that we must have an electoral system that, “results, in general, in proportional representation”.

Any electoral system where significant disproportionality exists as an automatic consequence of the design of the system must inevitably come into question on constitutional grounds.

A second very significant problem pointed out by some is that only 200 seats are available to independent candidates, whereas the full complement of votes cast is used in the calculations. This is done is avoid the difficulty of having every single independent candidate on the national ballot by restricting candidates to their home provinces. With half of the seats being allocated in relation to all of the votes, the obvious numerical consequence is that independent candidates for the National Assembly require, in general, twice as many votes to secure a seat compared to the numbers of votes that parties require per seat. With provincial variations, this means that independent candidates would require between 80,000 and 96,000 votes to secure a seat in the National Assembly (using the figures from 2019 as a guide), whereas political parties required about 44,000 votes per seat.

What this means is that even successful votes for independent candidates (before considering unsuccessful votes, or “excess” votes) count half of what votes for political parties count. From this perspective of individual voters, the constitutionality of the Bill is brought into question on grounds of proportionality. A further consequence of this arrangement is that half of all successful votes for independent candidates are effectively discarded, along with many votes for candidates who obtain more votes than parties’ per-seat requirement, but do not gain a seat. This “discarding” strengthens the numerical effect of independents collectively gaining far fewer seats than their share of the votes, which has the above-mentioned effect of giving parties a greater share of seats than their vote-share would indicate.

In the calculations for seat allocation of parties, the disproportionality of the independents’ seats and votes that are removed causes the effective quota for parties to be somewhat reduced. Clearly, the party with the most seats in any context benefits the most from any reduction in quota, as the seats allocated to them “cost” fewer votes, meaning that they gain more seats. Where “extra” seats are allocated to parties in this manner, the largest parties gain the most. However, the gain goes beyond the relative proportionality among the parties, as a marginal reduction of quota would make almost no difference for the smallest parties, while still making a distinct difference for the largest parties.

At an absolute minimum, for the National Assembly, at least half of all votes cast for independents will effectively result in seats being granted to the largest parties, above and beyond their respective share of the votes. Numerical simulations carried out to model the effects of the Bill indicate that the effective rate of discarding votes is likely to be around 75% of all votes cast for independents for the National Assembly.

There are further provisions in the Bill where novel arrangements (particular details of how seat allocation calculations are carried out that have not previously been used in the South African context) place higher burdens on independent candidates, and effectively transfer some seats from smaller parties to the largest party. These serve to amplify the disproportionality of the final outcomes. Changing these would soften the effect, but would not alter the fundamental numerical distortions described above.

Using the 2019 election results as a base, and assuming overall support for independent candidates as ranging between 5% and 15% of total votes cast, provisional modelling has shown that the African National Congress would gain a majority of seats in the National Assembly with as little as 44% of the votes cast. A similar effect would occur, to a lesser extent, in the provincial legislatures, where the largest party could generally gain an outright majority of seats with 46% of the vote.

Any electoral system that merely blends individuals with proportional representation is bound to suffer from the difficulties and contradictions described herein. Only a hybrid system, where separate constituency and proportional elections are combined, can satisfy the particular requirements in the Constitution of, “in general, proportional representation”.

A multi-member constituency system does share some of the difficulties of the current Bill, albeit to a lesser extent, if a single ballot is used for both constituency and proportional seat allocations. If separate ballots for constituencies and for proportional representation are used, then the proportionality problems fall away, providing that the number of constituency seats is not too high, which makes the system prone to “overhang” (where the excess of constituency seats gained by a dominant party is greater than the balancing effect of the proportional-only seats). In particular, a single-member constituency system would satisfy the constitutional proportionality requirements, as the second ballot would ensure the overall proportionality, and there would be no “discarding” of ballots.

Even though unsuccessful constituency votes are discarded, voters are fully aware that they are voting for a candidate in their district, and that the overall proportionality of the National Assembly or provincial legislature is determined solely by their PR ballot.

It is telling that Valli Moosa has articulated concerns over the Bill publicly, saying recently, “Without any political analysis and without any mathematical analysis, the risk of unintended consequences are very high.”

This leaves us with the question of what happens from now. As yet, we have no indication of the extent to which public submissions may impact on the thinking of the Home Affairs Committee. Up to this point, none of the political parties in Parliament have raised concerns over the constitutionality of the Bill. Of the public submissions, very few individuals or groups have raised any of these concerns (the writer has consulted with two such groups, and a formal written analysis of the concerns was included with one of the submissions).

If the Committee presses ahead with the Bill in substantially the same form, then it will go to the National Assembly and then to the National Council of Provinces to be passed. After this, it would have to be signed by the President. Assuming that this goes ahead, it would then be up to civil society to challenge the constitutionality of the Bill in court. If such a challenge is successful, and the Bill is referred back to Parliament, then it will be nearly impossible for the 2024 National and Provincial elections to be held in terms of a new electoral system that accommodates independent candidates. Given that this was the clear intent of the court in the June, 2020 ruling, failure to do so would be a most unfortunate outcome.

Over all of these years, Parliament has failed to grasp the nettle of electoral reform, and even when compelled by the court, it is most unfortunate that this has still not happened. The convenient, “minimalist”, option has such severe unintended consequences as to render it unviable. This leaves us no further forward, and there is an urgent need for leadership and direction to move South Africa through to a new, inclusive, and fair system. It may just be that civil society and the courts have a crucial role to play to take this long-running saga to a successful conclusion.

Michael Atkins (@atkinsmike1) is an IT consultant, and independent election data analyst, focusing on the accuracy and transparency of electoral processes. He presented a proposal to the MAC in 2021, and submitted a formal analysis of the Bill to the Home Affairs Committee.

Join in the debate... let us know what you think!

Discover more from Democracy in Africa

Subscribe now to keep reading and get access to the full archive.

Continue reading