How does this ‘PR top-up’ system of voting work?

5 August 2016

At the time of writing this, the Nelson Mandela Bay metro results are hanging in the balance, but most of the ward vote has already been declared. Of the 60 wards in the metro, 34 have been called for the ANC, 21 for the DA and 5 still are undecided.

Despite the ANC’s strong lead in the ward seats, the DA has almost half of the total votes while the ANC has around 40 percent of the vote. How does this system work?

Proportional voting versus first-past-the-post voting

Our municipal electoral system has two kinds of votes. The first one, the proportional representation (PR) system is quite easy to understand. Under the PR system, parties are awarded seats in proportion to their shares of the total vote.

So, if there is a municipality of 50 people, and there are five PR seats, a party will need ten votes to win one seat (i.e. 50 total votes divided by five seats = ten votes per seat). If the ANC won 30 votes and the DA won 20 votes in this municipality then the ANC would receive three seats and the DA would receive two seats.

We use the PR system for national and provincial seat allocations. This system ensures that smaller parties are represented in the National Assembly and the National Council of Provinces – a party needs to win just a quarter of one percent of the vote to get a seat in the National Assembly.

Under a first-past-the-post (FPTP) or constituency-based system, there is a winner-takes-all approach. The party or person with the most votes in an area will win the seat in that area, and nobody else will be represented.

Assume that our municipality of 50 people had just five ward seats instead of five PR seats, with ten people in each ward. If a particular ward of ten people has six voting ANC and four voting DA, the ward would be won by the ANC and the DA would receive nothing.

The advantage of the FPTP / constituency system is that you vote for people who live in your area and who you can interact with directly. Under a PR system you don’t have that direct link. However, the obvious drawback of the FPTP system is that it gives an advantage to bigger parties and hurts small parties.

It can get even worse – if the vote is split between three or more parties in an ward, the winner of that ward might get significantly less than 50 percent of the vote. This has actually happened in some KZN wards in 2011 where the ANC, IFP and NFP split the vote three ways and the winner got less than 40 percent of the vote in some cases.

In systems where there is only a FPTP system, the weaknesses can be quite glaring. In the 2005 general elections in the UK, the Labour Party won just 36 percent of all votes but the party took 57 percent of the seats. In contrast the Liberal Democrats took 23 percent of the vote but won just 10 percent of the seats. The FPTP system ends up penalising smaller parties.

Our municipal system combines both types of votes in an effort to get the best of both systems and to minimise the weaknesses in both. It’s referred to colloquially as a ‘PR-top-up system’ and it works like this:

In any muncipality there are both ward and PR seats – half of the total seats are ward and half are PR. Every voter votes twice, once on a ward ballot and once on a PR ballot. The ward seats are determined according to the normal FPTP system and the PR seats are determined according to the total vote count (ward plus PR).

Confused? Let’s look at an example.

Illustrated example (with stickpeople)

Imagine a municipality with 50 people, where 30 support the ANC and 20 support the DA:

Our imaginary voting population in our imaginary municipality

Our imaginary voting population in our imaginary municipality

There are five wards in our imaginary municipality, and each ward contains ten voters, in the same proportion as the overall municipality – in other words every ward has six ANC voters and four DA voters:

A typical ward in our imaginary municipality

A typical ward in our imaginary municipality









Each of our voters votes twice, and all of them vote for their party twice. There is no splitting of the vote and no spoiling of ballot papers. There will be 100 votes cast, 60 for the ANC and 40 for the DA.

The ANC will win all five wards, picking up six ward votes in every ward to the DA’s four votes. If our system only had ward votes then that would be the end of the story – the ANC would have 100 percent of the wards with only 60 percent of the votes. The DA would have 0 percent of the wards although it won 40 percent of the vote.

There are still five PR seats in our system. If these five seats were determined just by the PR votes (50 votes out of the 100 cast) then they would be allocated in proportion to the voting patterns. In that case the ANC would get 60 percent of the PR seats and the DA would get 40 percent: the ANC would win three of the five PR seats and the DA would win two seats.

Overall, the ANC would have eight of the ten seats (five ward, three PR) and the DA would win two PR seats. But our system works like a giant PR system, where both the ward and the PR votes can count towards the allocation of PR seats.

In our example there are 100 votes cast (50 ward and 50 PR). There are ten seats (five ward, five PR). Each party should receive a total number of seats that is in proportion to its total number of votes.

The DA received 40 votes out of 100, or 40 percent of the total. It should receive 40 percent of the seats, or four out of ten. It therefore receives four PR seats, not just two.

The ANC received 60 votes out of 100, or 60 percent. It should receive six of the ten seats. It’s already won five ward seats, so it’s only entitled to only one PR seat.

In simple terms, the municipal system is like a simple PR system, similar to our national and provincial voting systems. But it also includes provision for FPTP / constituency-based seats so that people have representation right down at their ward level. The system corrects for the weaknesses of a FPTP system so that large parties don’t get power at the expense of small parties.

In reality, most wards in South Africa are either strongly ANC or DA (i.e. more than 70 percent of the vote). Many people may have felt that their vote wouldn’t make much difference in a ‘safe’ ward and didn’t turn out to vote.

In these 2016 elections, the DA campaigned hard to get its supporters to vote, emphasising that every vote would count towards the party’s success. What seems to be happening in Nelson Mandela Bay metro (and some other metros) is that the safe ANC wards were won by the party but with a low turnout of voters, while the DA wards were won with high turnouts.

Nelson Mandela Bay explains why the DA can be ahead in municipalities where the ANC has more wards. It also helps to explain why every vote counts, even in a ward where one party has an overwhelming majority, and why a low turnout of ANC voters has hurt the party so badly in these elections.


3 thoughts on “How does this ‘PR top-up’ system of voting work?

  1. Rivaj

    Nice one Paul, appreciate the effort to write this up. It was as clear as mud and now, while not crystal clear, is at least a little less muddy!

  2. Andrew Fraser

    This is such a clever system, we should have a similar system for National Elections. There should be some constituency responsibilities for MPs.

  3. Toba Mofokeng

    Thanks Paul, it was so difficult for me to explain to my team as to how seats are allocated in PR system. You have make my next session easy.. Hope to learn much from your blog


Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>