PAID CONTENT

Vote management: South Africa’s first electoral reform outcome is best reviewed
The Constitutional Court heard arguments for a challenge to the Electoral Amendment Act, case CCT 144/23. The Court’s website gives access to court documents. The Ruling [67] construes that the dispute concerns the number of compensatory seats reserved for political parties. Two matters are potentially destructive to the Applicant’s case: foreseeability of overhang [72] under the proposed 350/50 split, and acknowledgement [66] that the 200/200 split serves a legitimate governmental purpose of ensuring proportionality in results. On the requirement of achieving proportionality [68] the split chosen by Parliament passes constitutional muster; and while other proposals may achieve the same, wide remit [55] is given to Parliament to enact the voting system.
As second and third respondents, Parliament in its affidavit [6] mentions competing considerations that require trade-offs [70.11] and balancing [PA8-6]. Differences [12] between independents and political parties render parity not possible [13], it is contended, before framing the national elections structure [14] as the ‘only two-tier way’ to do it [17.2; 67.4], drawing on the local government structure [56.2.1] and bolstered by what the Applicant accepts [15] ‘as it must’. Independent candidate exclusion is based on practicalities [57.2(c)] risk [57.2(d)] and opinion [57.2(e)] and contrasts [PA8-21] with parties that are deemed ‘prime bodies’ [56.2.4] for correcting proportionality [57.2(a)], filling compensatory seats that exist [70.3(a)] purportedly not to elect representatives directly [68; 70.4(b)], a view resisted [46.1(b)] to no avail. Impetus is gathered [46.3], quota discrepancies pacified [50.4], double counting concerns muffled [50.8; PA7-37], and Constitutional blessing deduced [56.2.3] for the national ballot structure. A: It is the provincial ballot structure enacted by Parliament [45.2] that stands tall: incontestable on the proportionality requirement (with one safeguard ideally added) and singlehandedly negating the contention [17.1; 70.3(c)]; PA8-14] that the inclusion of independents may distort proportionality. The ‘Lekota Bill’ (B34-2020) also may have demanded no ‘balancing’ between independent inclusion and proportionality.
Parliament [10] was assisted by the Independent Electoral Commission (IEC) and fifth respondent, who offers to share knowledge, also on proportionality, which is claimed in their affidavit [IEC_20] but distinguished from absolute/perfect versions [IEC_22] viewed as impossible due to rounding [IEC_28.3]. Researchers were engaged who conducted preliminary analyses (not available) and further analyses attached as a report [PSM4] that state caveats: votes for independents require excess vote measures; excluding independents measures inter-party proportionality.
B: Four brief points: (i) The first [3.1] finding of the report [PSM4] is expected to follow from any correct application of valid quotas. (ii) Inter-party proportionality is a select subset of (and resolved by) the broader concept of ‘share-of-vote’ or proportionality in the widest sense as directly sourced from choices made by the electorate. (iii) Analyses of predisposition towards parties with higher vote shares may need to control for the share of seats (up to say 15%) left unallocated in the quota’s wake, as a proxy for sensitivity to kurtosis or numbers/ratios of no-seat allocations. (iv) Importantly, no attack on its proportionality can cure the enacted system’s perceived problems, as the sources thereof exist elsewhere more rudimentary.
The Bill was challengeable. Parliament [50.1] recognised it (and acted nimbly). A ‘reset’ occurred [PA7-20] proposed by the IEC and embraced and developed by Council, who, in their Jul ‘22 Opinion [PA7-39] suggests two ballots and proposes [PA7-40] how voters should vote. The Local Elections framework is construed as a balance struck between ‘independent participation’ (on one hand [13]) and (on the other) ‘proportional party representation’ [PA7-43] and/or ‘ensuring proportionality’ [PA7-48] – apparently presented as competing considerations [17.2] as if mutually exclusive; views that are repeated in a Nov ‘22 Opinion where it is mentioned [PA8-13] that the Structures Act has never been challenged. Parliament [56.2.2] refers to the New Nation Judgement and provides an extract [PA8-7] of paras 79-80 thereof – from which interpretations follow that may be open to challenge.
C: The New Nation Judgement, halfway through para 80, reads: ‘the clearest possible statement that dispels the notion that proportional representation is consonant only with representation through political parties’; wording and context that may not bode well for concepts such as ‘prime bodies for proportionality’ or ‘legal nexus’ [IEC_39.1; PA8-21], or ‘inter-party proportionality’. The first part of para 79 (not in the extract), reads: proportionality may ‘find application in a combination of party representation and independents’. Nonetheless, Respondents seem intent on compartmentalising proportionality, possibly in support of the ‘special purpose’ ballot (which might be fine) from which contestants are excluded (which is troublesome): Prevent contestants from appearing on a ballot and no vote exists for those barred from it when the ‘share of vote’ is targeted for proportionality, seemingly sidestepping one constitutional safeguard. The Judgement contains wording whose importance may still evolve – towards its end para 78 reads: ‘whoever the participants may be’. And the enacted system is known to (selectively) decide who the contestants for the second (compensatory) ballot will be.
With two tiers in one ballot [IEC_11.2] there may not have been compelling need for two ballots [PA7-38], but two are now enacted [IEC_13.3]. Amongst the advantages offered is equalised quota values for contestants [70.2], which circumvents the trouble [PA7-19] of a possibly discriminatory result [IEC_13.4], now avoided by the introduction of the second (constrained) ballot. Unequal vote weight across ballots is now possible; the Ruling [106] records it being defended (on grounds of differences, including a differing purpose). Vote splitting is now possible too, as described in the Ruling [124] as an unavoidable consequence of introducing the second ballot.
D: Many of us (not all of us) are already gifted the option of vote splitting from a second vote [50.10.1]. Still, the principal gift of Parliament’s [38.7] ‘minimalist option’ (approved by Cabinet but never enacted), could be that it was brazen or divulging of matters to look out for.
D1: Some potential issues are inherited or became institutionalised as time passed. ‘National’ lists appear exempt from having to ensure fair geographic/regional representation nationally, which is the objective for having regional seats and regional lists [PA7-31]. The fixing of regional seat numbers in advance may give the nod to an increased risk of overhang (beyond rounding): it is not necessary; it pits regions against one another; it rewards low voter turnout with a low quota and penalises high turnout with a high quota; and it causes the risk of overhang to depend on deviance between actual and estimated voter turnout ratios, which is recognised [70.5] but not controlled for. One universal pair of upper and lower limits imposed on regional quotas is sufficient to mitigate excessive (overhang threatening) distortion of proportionality. Combined with holding off prematurely finalising the regional seat allocations [IEC_40.4], these limits may be applied to first scale back excessive distortion should it occur.
D2: If distorted, the (minimum) number of compensatory seats needed to restore proportionality, is not known in advance. A stated purpose for the 200/200 split is to lower overhang risk, which it is better at than its 350/50 rival split. Risk aversion involves balancing: the more successful the (type-A) risk of overhang is averted (other things being equal; ceteris paribus), the more of another (type-B) risk is incurred, in this case, the risk that proportionality is restorable within the scope of 200 seats reserved for it [67.7], leaving excess/residual seats that threaten to falsify the stated objective. Parliament [32.4] is aware of past deviance below 5% and might have gauged how generous seat reservation at 50% may be – all 200 will breach purpose if the regional ballot is made proportional (e.g. by tight limits). Parliament [50.11] also deems the matter a policy decision and may concur that, to the extent compensatory seats are too many for their purpose, they may contribute to the caution heeded [PA8-16] in determining the split, and do not serve the ‘different purpose’: they may need their own stated objective. Any such residual seats should be contested for by all, more so for the 200/200 split compared to the 350/50 split, meaning that all eligible contestants are to appear on the second (compensatory) ballot paper – a duplicate of the first, only marked ‘second’ to be distinguishable, or alternatively, a single ballot paper with two columns where votes may be cast distinctively.
D3: There may be no need to calculate residual seat numbers: ample seats are available to correct for inter-party proportionality and achieve share-of-vote or ‘in general’ proportionality. A conceptual (roughly back-to-front) replica of the enacted system may be crafted with vote splitting utilised as the objective for (modest) reforms requiring two all-inclusive ballots for national elections. Arguments for excluding independents fall away essentially by rearranging the core modules or sequences for translating votes into seats. (a) First up, the replica resolves proportionality: a universal Droop quota, based on an aggregate all-inclusive vote across two ballots for 400 seats – similar to the one described in the Ruling [107] – is used to reserve (minimum) total seats per political party, which also secures inter-party proportionality, with both ballots properly inclusive of independents; some parties may later be allocated one more seat, but not more than one (as that may deviate from the objective). (b) Secondly, with varying regional quotas endured for historic reasons, the replica resolves overhang risk: universal upper and lower limits (above and below the universal quota) will be applied to regional quotas (cleverly, so as not to interfere with the targeted 200/200 split) should (and when) any party’s (provisional) aggregate regional seat allocations happen to exceed its reserved total. (c) Thirdly, it replicates the enacted system’s regional seat allocations up to 200 for parties and independents alike. (d) Finally, the replica simply fills the deficit seats (total reserved seats less regional allocations) for each party from their (second ballot) national lists, leaving some seats unallocated. (e) Note that independents are not excluded as there is no ‘correcting ballot’ necessary for parties, and (along with parties) independents may be allocated seats as is done for the regional ballot, of course restricted to a single seat in the legislature. (f) To achieve a close replica the limits controlling for overhang risk need to be dropped (possibly to embolden the risk of overhang) and the second ballot needs to disallow independents (possibly to reveal bias, deemed justifiable).
D4: The enacted system’s share of vote is assessed from the aggregate of two votes cast for consecutive ballots – the second serves to add votes to the aggregate, as does the first: Save for potentially differing quotas, the translation of votes into seats is the same. The second (compensatory) ballot does not correct for proportionality, which is the domain of the universal (Droop) quota as applied to the aggregate vote cast (whether of one, two, or many ballots) – but it does prevent the translation of votes into seats for independents and deny (at least some) voters free choice and the option of either voting for their next favourite second, or voting for their (singular) favourites twice.
D5: Voters should at least be free to choose which ballot paper they wish to use for their second vote, potentially subtracting from one and adding to the other ballot paper pile. Similarly, with the secret vote principle in mind, all-inclusive ballot papers that distinguishes between ballots may be used, with votes cast for independents on the second (compensatory) ballot allowed to be transferred to the first pile. Both ballots may then proceed as enacted. This may be a pragmatic solution for the 2024 elections as it requires minimal adjustments: It will be necessary to safeguard equal vote weight across ballots, requiring an outcome-based first/second ballot paper split, post the transfer of votes, appropriately rounded, to replace the 200/200 split, and to automatically address residual (too many) compensatory seat issues: a smaller integer number of seats is expected to remain reserved for political parties.
D6: The enacted compensatory ballot translates votes into seats. Every voter should have two votes [IEC_13.4] regardless of whom they support. Voters may be aware thereof and prefer to vote for their favourite(s), as they are choosing representatives. Leaving candidates off the list creates a dilemma for the voter who, as a result, may forfeit the second vote, which carries no veto as the process continues absent of any information withheld. The process is driven by inter-party information that does become available as votes are cast. Evidence of voters making do with what they have [IEC_13.5] might be left at that. Properly including all contestants on the ballot does not create an untenable situation for the system, as the process may continue absent information not deemed helpful in this instance. Whether inter-party proportionality holds with independent candidates included on the ballot, was not insisted on being verified [PSM4], regrettably, as it is suggested [PA10-3.7]. The Ruling [113] records that rights may not necessarily be violated with categories of persons being treated differently. In what might be a twist of fate, the compensatory ballot may well correct for proportionality for parties, as designed, but it does not necessarily have to do the same for independent candidates nor exclude them from the ballot. For both categories, however, votes cast must (equally) translate into seats allocated, requiring all-inclusive ballot papers. Reserving seats for political parties may violate independents’ equal right to hold office, and/or may be pitted against the New Nation Judgement.
D7: The New Nation Judgement [62] summarises its discourse as a serious matter implicating no fewer than three fundamental rights: to freedom of association (Section 18;) freedom of conscience (Section 15(1)); and dignity (Section 10). It may be that the case (CCT 110/19) could be ‘copied’ (by substituting its Section 19(3)(b) right for the Section 19(3)(a) right to vote) to challenge the exclusive second (compensatory) ballot [PA7-39.2] that may deny voters free political choice by refusing some voters the option to vote for participants-of-choice. The right to vote freely, including for your favourite, including when the vote translates into a seat, may not be itemised in the Constitution, but the Judgement [17] proclaims that it is not necessary. The matter still needs to be argued, which is a slippery road, but, drawing on the Judgement’s [18] wording, the assertion sought might read: once an adult citizen is forced to exercise the Section 19(3)(a) right to vote through a constrained ballot list, that may divest her or him of the choice guaranteed by Section 19(1) to choose freely – and that (too) cannot be.
D8: Parliament’s [12.2; 46.3(c)] concern for voters seems to revolve around independent candidates occupying one seat if elected, which would apply to any type of candidate, and continues to apply post a candidate receiving both votes from a single voter, freely cast, or a first vote from one voter and a second from another. With all contestants included on every ballot list, the entire electorate may vote freely, resulting in sound voting information, to be targeted to have at hand for the 2024 elections. Whether information yielded (or votes cast) may then be discarded on grounds of not being ‘helpful’ to a special purpose ballot, remains to be decided. It is clear that such discarding of votes achieves a similar end result to the prevention thereof from entering the system. And if preventing entry, however well intended [IEC_40.6], infringes on a right (like dignity), so too might the discarding of votes.
D9: The body with wide remit administers differing vote weights amongst regions in the first tier and corrects for it in the second tier with a universal quota applied across regions and ballots. Like it is required for regional party candidates [PA7-32], independent candidates should have been restricted to contest only the region where they are registered to vote. It remains the way to go (also for contesting the compensatory ballot). Respondents went after a bigger bite allowing independents to contest across regions like parties; bearing the familiar tier-one dilemma [PA8-40.7] at which they throw up their hands [PA8-40.5], deflect blame [PA8-40.8] and sympathise with victims [PA8-40.4]; instead of following through on their knowledge of tier-two correction by means of a universal quota. Apparently, the same process corrects for their predicament too: Respondents may have to swallow having independent candidates partaking in the consecutive processes of a two-tier voting system. What stays different for independents post the correction processes, is their restriction to a single seat, either regional or compensatory, for which the regional ballot may be geared towards but the compensatory ballot requires specific adjustments.
D10: The constraint imposed on political parties not to field their candidates appears related to ‘intermediation’ or securing a vote at grouping level. Any political party who so chooses should be allowed to field candidates, perhaps listed according to the party ranking, for any ballot, including the compensatory ballot, to compete with independent candidates – restricted to the ballot paper covering their registered voting district, as all fielded candidates should be. The matter may have been revived by arguments about a false dichotomy [PA7-41] seemingly made in support of the exclusion of independents from the second (compensatory) ballot. If addressed constructively, then the voting system may expand to recognise the formation of groupings like coalitions by allowing apparentement between willing contestants who retain their names/brands on the ballot list but found grounds for competing together and require proportionality to apply at the broader grouping and at inter member level.
There were problems with the proposed system prior to the ‘reset’ and there are problems with the voting system as enacted, perhaps already recognised in it being readied for a short life expectancy [PA8-24], that may materialise or not, but that suggests that problems are perpetuating into the future.
One adult citizen may be wishing to vote freely in the 2024 national elections, worried about vote management; another may be worried about rumours of interference within the casting/counting process; others for perceiving Parliament to show insincere concern for the electorate; more still for seemingly short-sighted or inadequate legislation being enacted; some perhaps about entirely different matters. In South Africa, at least for voting, the standard must be beyond reproach: free political choice through all-inclusive ballot papers for each vote to be cast. May we all have preferred choices on the ballots for 2024 and clarity on what our country needs for the next five years.
Thomas R Labuschagne, 2024/01/29