While the Democratic Alliance (DA) controls only one of the eight metros and the African National Congress (ANC) the other seven, data on voting trends suggest that in next year’s local elections the ANC will be vulnerable in at least two other metros too.
Voting results from the 2000, 2006 and 2011 local elections, as well as municipal disaggregation from the provincial elections last year, suggest the proportional representation electoral system will be pivotal in the outcome of next year’s poll.
Last year’s provincial results reflect voters’ registered residence, but next year’s will be influenced by registration, turnout and changes in allegiance — including independent candidates.
Support for the ANC in Nelson Mandela Bay dropped by three percentage points from 2011 to last year, taking the party below 50% of votes cast, compounding a downward trend from 2000-2014, during which there has been a 17 percentage point drop in ANC support.
In data terms, some variable has to change to arrest this sustained downward trend. The ANC is well aware that a game changer is needed, presumably prompting its decision to parachute in Danny Jordaan as new mayor of Nelson Mandela Bay to replace octogenarian Ben Fihla.
The choice, before Fifagate, looked inspired; Jordaan has deep personal roots in the metro and, until recently, he had an untarnished 2010 World Cup delivery record as head of the South African Football Association.
The association’s dealings with Fifa officials have now proved to be a liability as the scandal continues to unfold.
In Tshwane, support for the ANC dropped from 55% to 49% from 2011 to last year, contributing to a decline of almost seven percentage points from 2000 to last year. Bedevilling ANC support in Tshwane is overt disapproval by alliance partners of mayor Kgosientso Ramokgopa.
The ANC is also vulnerable in other Gauteng metros, all of which have seen a drop in support for the party.
In Johannesburg it captured just over 52% in provincial support last year, and 55% in Ekurhuleni. Majority support remains likely next year in these two metros but cannot be assumed, with Eskom credit control in Soweto and unresolved billing issues in Johannesburg major hurdles, along with the national crisis of load shedding.
To reverse last year’s losses, the ANC is putting its faith in a more united provincial leadership and revisions to reviled e-tolls.
Given the close margins in some metros, it has become increasingly important to understand how the proportional representation (PR) system works.
Speculation that it will favour the ANC next year is based on the erroneous assumption that there are two separate, parallel systems for counting local votes that will favour parties that dominate ward outcomes. The South African local government system of seat allocation is calculated in a way that ensures that total seats (ward and PR) per party reflect support for that party.
Once ward votes are counted and ward seats allocated, the allocation of total (ward and PR) as well as PR seats in a local municipality is done through two transparent calculations, with the order of these calculations of critical importance. The total number of seats per party is worked out first, by adding all PR and ward votes cast for a party in a particular municipality, and dividing it by the quota, a figure representing the proportion of votes needed for a seat, less independent votes and seats.
A party’s total (ward and PR) seats in a local council is the same as its proportion of total ward and PR votes cast in the council or, put another way, the percentage of the popular vote.
Once the total number of council seats per party is determined in the first calculation, the number of PR seats is determined in a second calculation by subtracting the number of ward seats for a party from the total seats for that party.
PR seat allocation ensures that no matter how many seats a party gets at ward level, its total seat allocation is proportionately the same as its total support (or the total number of votes cast in a municipality).
Hotly contested Western Cape Bitou provides a simple example of how the PR system works.
In 2011, marginally more votes were cast (in total) for the DA than the ANC (16,502 versus 15,780), even though the ANC won five out of the 13 ward seats while the DA won only two.
As both parties were entitled to six seats (dividing votes by a quota derived from the total number of votes), the ANC’s seats comprise five ward and one PR allocation, while the DA’s six seats are divided between two ward councillors and four PR ward councillors.
The Congress of the People (Cope) won one PR seat in Bitou and it is in alliance with this Cope councillor that the DA took control of the council.
The Bitou example illustrates how the percentage of total (ward and PR) votes cast for the ANC in contested metros will be reflected in the proportion of total seats allocated to the ANC, regardless of how well it does at ward level.
Each party will be allocated PR seats, which, together with ward seats, will reflect proportionately the total number of votes cast for it. If the ANC gets 45% of all votes cast, it will get 45% of the total seats — the seat calculation process ensures it.
If a smaller party, such as the Economic Freedom Fighters (EFF), wins no ward seats in a council with a total of 100 seats (ward and PR), but gets 15% of total votes cast, it will get 15 PR seats – ensuring that its proportion of total seats reflects its popularity across the municipality.
SA’s local government electoral system is thus truly representative, where percentage support for a party is reflected proportionately in that party’s council seats.
The system favours no party, big or small. It means that no vote is lost; that voters who support smaller parties are assured of representation on their municipal council.
The PR system also means that voter abstention can undermine the outcome for the party for which voters would have previously voted, even if the ward in which they reside continues to be carried by party loyalists. A significant danger to the ANC in some areas is the threat of stayaways and apathy by disenchanted supporters in 2016.
More importantly, the PR system means parties that are not the majority party in a ward or block of wards — defined by a living area such as a township or a municipality — can still “win on the margins”.
In Nelson Mandela Bay or Tshwane, where the ANC’s majority is fragile, the opposition does not have to beat the ANC in the township wards in which the ANC dominates decisively.
It has only to claw out a small percentage of votes from these areas and maintain majorities in traditional opposition strongholds to push the ANC below 50% across the municipality, thus opening these metros to the very real possibility of a non-ANC coalition government and, in so doing, winning on the margins.
• Allan is Municipal IQ’s MD, Heese its economist