The South African Electoral System

This is the first in a series of Briefs dealing with elections in South Africa. This Brief unpacks some of the main components of the South African General Elections.

Background

South Africa’s electoral system, as we now know it, was carefully considered. The challenge faced by our leaders at the Convention for a Democratic South Africa was to select an appropriate electoral system for a highly divided and unequal society in the process of a delicate transition.[1] The configuration of an electoral system is of fundamental importance to the nature of a country’s politics and, very importantly, the match between the preferences of citizens, the preferences of elected officials, and government's policy direction. South Africa’s ‘proportional representation’ system was selected for its inclusiveness, its simplicity, and its tendency to encourage coalition government.[2]

Electoral systems can be compared along three broad dimensions: Its ballot structure (how citizens cast their vote, and what they vote for); its district structure (how many districts there are, and the number of seats per district); and the electoral formula (how votes are converted into seats).[3] For South Africa’s National elections, citizens cast a vote for a single party of their choice; the country is divided into 10 large multi-member district regions: 9 corresponding to the 9 provinces (with a total magnitude of 200 seats, ranging from 5 to 48 seats in each region [4]), and 1 national district for the country as a whole (with a magnitude of 200 seats). We employ a proportional representation system, allocating seats in direct proportion to the number of votes a party received. This Brief unpacks the components of South Africa’s system in the sections that follow.

There are different ways to configure an electoral system, each entailing trade-offs.[5] For instance, a party-list system severs the link between individual representatives and voters, diminishing the voters’ say in which individual represents them, but generates a representative and proportional assembly. Conversely, a system which increases the ability of voters to hold individuals personally accountable generates an assembly that is less representative and disproportional, but freer to pursue unpopular policies. There is no perfect system, and systems need to be considered in the context of the particular country.[6]

Ideally, as society and societal conditions change, these trade-offs need to be revisited and reconsidered. Indeed, whether South Africa’s electoral system is still suited to the country’s needs is an ongoing debate.[7]

The South African System

Parliament
South Africa’s Parliament [8] consists of 490 seats. Our ‘bicameral’ Parliament has two houses – a lower house and an upper house. The lower house, or the ‘National Assembly’, represents 400 seats. The upper house, or the ‘National Council of Provinces (NCOP)’, represents 90 seats (10 members from each province). Members to the NCOP are provincial delegates nominated by each provincial legislature. The National Assembly, however, is filled in accordance with the votes cast in the General Election by the electorate. Hence, I will leave the NCOP aside.

Elections
The South African electoral system is a closed-list proportional representation system, and a General Election is held every 5 years. The process is facilitated by the Independent Electoral Commission, which is responsible for ensuring free and fair elections.

At the General Election voters elect the national and provincial legislatures simultaneously. Voters are presented with two separate ballots – a national ballot, and a provincial ballot. These ballots list, respectively, all the political parties competing in the election at national and provincial level. Voters cast one (categorical) vote for the party of their choice on each ballot, as they see fit.

In this system, voters vote for parties – not for individuals. It is the parties' prerogative who it wants in the legislatures. Parties submit – to the Independent Electoral Commission – nine provincial lists for the provincial legislatures; and, for the national legislature, nine provincial-to-national (or 'regional') lists and one national list (although a national list is not mandatory). These lists are ‘closed’ and cannot be altered by voters (but are publicised widely for the voter to consider).[9]

At the end of the electoral process, these (ranked) lists are used to fill the seats allocated to each party. The higher up on a list a party member is, the more likely that member is to get a seat.

Leaving the provincial legislature aside, this Brief will now focus on the election of the National Assembly, from regional and national districts, for the national legislature.

The National Assembly
The National Assembly is the decisive house in Parliament and it is here where the executive cabinet originates.

For seats to the National Assembly, only the National ballot is relevant. The National Assembly seats are filled in two tiers: Half (200) seats are regional seats and filled by reference to regional votes and regional lists; the other half (200) are national seats and filled by reference to national votes and national lists (or entirely from regional lists if a party did not submit a separate national list). But first votes must be converted into seats.[10]

Allocating seats
In order to allocate seats, the number of votes a party received is translated into a proportion of the seats in the National Assembly, first regionally and then nationally.

First, the number of votes equivalent to a single seat must be calculated. Each seat then represents a ‘quota’ of votes. The simplest way to do this is by dividing the total number of votes by the total number of seats (i.e. votes/seats).[11] In South Africa we use a version of the Droop Quota method.[12] For regional seats the quota is determined, for each region, by the total number of votes in that region and the total number of seats in that region.  For the national seats, the quota is determined by the total number of votes in the country and the total number of national seats.

Seats are allocated proportionally – the number of seats allocated to a party depends on how many times the party meets a full quota. This is calculated by dividing each party’s share of the vote, regionally and then nationally, by the quotas determined at those levels. During this process the remainders are set aside. If, after this process, there are unallocated seats, the remaining seats are allocated to the parties who have the largest remainder.[13]  And so the 400 seats for the National Assembly are filled.

 

NOTES

[1] http://www.sahistory.org.za/codesa-negotiations
[2] Lodge, T. 2003. How the South African Electoral System was Negotiated. Journal of African Elections 2 (1):71-76.
[3] Teorell, J & Lindstedt, C. 2010. Measuring electoral systems. Political Research Quarterly, Vol. 63, No2. (pp. 434-448).
[4] http://www.gov.za/documents/download.php?f=210704
[5] Norris, P. 1997. Choosing Electoral Systems: Proportional, Majoritarian and Mixed Systems. International Political Science Review, Vol. 18, No. 3, ‘Contrasting Political Institutions’ (pp. 297-312). Sage Publications.
[6] Ibid.
[7] See the 2003 Report of the Electoral Task Team, and the 2009 Report of the Independent Panel Assessment of Parliament. The Democratic Alliance and Agang raised alarm in the media last year calling for urgent electoral reform.
[8] Section 42, Chapter 4 of the South African Constitution
[9] The party-internal controversies regarding party lists are often a public spectacle, and are widely discussed. However, the public have no say in the process.
[10] The votes cast on the provincial ballots (for the election of provincial legislatures in each province) are tallied separately from those cast on the national ballot (for the election of national legislature in each region and for the country as a whole) – They are two separate elections held simultaneously.
[11] This method is known as the Hare quota or ‘the natural quota’.
[12] South Africa employs an ‘STV’ Droop quota which entails adding 1 to the denominator (seats), dividing the total number of votes by this number, and adding 1 to the result [votes/(seats+1)+1]  (see Jonathan Lundell & ID Hill's 'Notes on the Droop Quota'  http://www.votingmatters.org.uk/ISSUE24/I24P2.pdf)
[13] The largest remainder is only applied for the next 5 seats, after which a ‘highest average’ method is used to allocate seats, taking into consideration how many seats a party already has. (See Manuel Álvarez-Rivera's (2010) 'Election Resources on the Internet: The Republic of South Africa Electoral System'. http://electionresources.org/za/system/)

 

Bibliography

  • Gallagher, M and Mitchell, P (2005). Introduction to Electoral Systems. In: Michael Gallagher and Paul Mitchell (Eds), 2005, "The Politics of Electoral Systems" (pp. 3 - 23). New York: Oxford University Press.
  • Gallagher, M and Mitchell, P (2005). Appendix A: The Mechanics of Electoral Systems. In: Michael Gallagher and Paul Mitchell (Eds), 2005, "The Politics of Electoral Systems" (pp. 579 - 597). New York: Oxford University Press.
  • Gouws, A and Mitchell, P (2005). South Africa: One Party Dominance Despite Perfect Proportionality. In: Michael Gallagher and Paul Mitchell (Eds), 2005, "The Politics of Electoral Systems" (pp. 353 - 372). New York: Oxford University Press.
  • Lodge, T (2003). How the South African Electoral System was Negotiated. Journal of African Elections 2 (1):71-76.
  • Manuel Álvarez-Rivera (2010). Election Resources on the Internet: The Republic of South Africa Electoral System.  http://electionresources.org/za/system/
  • Norris, P (1997). Choosing Electoral Systems: Proportional, Majoritarian and Mixed Systems. International Political Science Review, Vol. 18, No. 3, ‘Contrasting Political Institutions’ (pp. 297-312). Sage Publications.
  • Teorell, J & Lindstedt, C (2010). Measuring electoral systems. Political Research Quarterly, Vol. 63, No2. (pp. 434-448).

 

Wim Louw
HSF Researcher
wim@hsf.org.za