The Constitution is silent on details of the electoral system, only requiring that it “results, in general, in proportional representation”. At municipal level, this is still required by the Constitution, but with the addition of ward councillors.
The 2003 Van Zyl Slabbert Commission traversed some of the issues of electoral reform, including a constituency system at national and provincial levels, with proportional top-up, much as with municipal voting, but did not address independent candidates.
At municipal level, independent candidates can stand, but only in wards. That leads to a defect in proportionality: in an extreme case, independent candidates could be the biggest voting bloc and not win any wards. Because independent candidates do not have a proportional-representation vote, they are not represented in any top-up to correct the ward result. This violates the principle of proportionality required at all levels by the Constitution.
In addition to the Constitutional Court judgment being implemented at national and provincial levels, this defect should be remedied at municipal level too.
The proportional representation system we currently use is a closed list. The party decides on the order of the list and the voter gets to decide only which party to support. By contrast, an open list system is any design that gives voters some say on the candidates, as well as the party.
Within the current proportional representation design of our electoral system, addressing the defect found by the Constitutional Court would be difficult, because independent candidates would have to agree on a list.
Let’s start by trying to remedy this in the existing municipal system, then generalise to other levels.
If proportional representatives were based on candidates who received the most votes but not enough to win a ward, that would remove the need for independent candidates to agree on a list. A simple model would be to work out how many councillors the independent candidates should have collectively, and construct a proportional representation list of independents who did not win a ward. The order: from the highest vote for independents who did not win a ward, down.
In national and provincial elections, a similar principle could apply. If, for example, independent candidates around the country (with a similar principle applying to provinces) collectively scored 10% of the vote, equating to 40 seats, the 40 independents who scored the most votes would be elected. That could be done in the current pure proportional representation system. Almost: I will get to the caveat later.
The next question: how to qualify as a candidate. Because a candidate can only win one seat, a deposit far lower than that required for a party to register nationally would be fair. This would be refunded to candidates above a certain vote threshold. A new party also requires signatures of 500 voters to register for the first time and a similar requirement for an independent candidate could be considered. The same fee that applies for a new party to register could also be considered — currently R500.
All of that can be done as a minor tweak to the electoral system; the bar for independent candidates to nominate needs some thought, because setting it too low will result in a flood of candidates; setting it too high will exclude those without resources. To avoid a massive national ballot, an independent candidate should appear on the ballot only in a particular geographic area.
However, this could be a good moment to revisit the Van Zyl Slabbert report and reconsider the value of a constituency plus proportional system at all levels. The Van Zyl Slabbert report recommends a multi-member constituency system with a similar party list to the kind we currently have — closed lists, drawn up by each party.
However, if my idea for accommodating independent candidates is adopted, it is not compatible with a closed list, because the order of my proposed list depends on votes for candidates.
Given that our closed list system has proved to blur lines of accountability — working a party patronage network can get you a higher place on a list with no reference to public preference — we should also reconsider that.
Rather than make the minimal fix to the Electoral Act to accommodate independent candidates, this would be a great opportunity for real electoral reform in which voters are given much more say about who represents them.
Philip Machanick is an associate professor of computer science at Rhodes University