Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 20 June 2019

Electoral Systems in Context: South Africa

Abstract and Keywords

South Africa’s post-apartheid election outcomes demonstrate how contextual factors interact with electoral rules to shape party systems. South Africa’s national electoral system represents one of the most permissive in the world, combining parliamentary rules with an extreme form of proportional representation. These rules were selected to encourage broad representation of parties in the National Assembly. However, South Africa’s party system consistently defies expectations, with a low effective number of seat-winning parties at the national level and dominance by a single party, the African National Congress (ANC). Provincial and municipal outcomes also confound simple institutional expectations. In addition to describing electoral rules and party systems at all three levels of South Africa’s political system, this chapter argues that contextual factors like the salience of racial divisions and the ability of the ruling party to shape institutions and resource flows critically interact with electoral institutions to shape party system outcomes.

Keywords: South Africa, electoral systems, proportional representation, mixed-member plurality electoral rules, single-party dominance

South Africa transitioned to democracy in April 1994 when the African National Congress (ANC) won a landslide election and took leadership of the national government, as well as seven of nine provincial governments. The electoral institutions guiding these elections emerged from a lengthy negotiation process, a key goal of which was the representation in parliament of the country’s many diverse groups and political traditions. Reflecting these goals of inclusion and representation, South Africa’s electoral institutions fell on the far end of the consensual/majoritarian spectrum: they combined parliamentary rules with an extreme form of proportional representation, one that would allow parties with as little as a quarter of 1 percent of the national vote to obtain a seat in the National Assembly. Such a design, endorsed by leading constitutional engineers as the most favorable to democracies with deep social divisions, is generally associated with the consensus-building coalition and minority governments of continental Europe (Lijphart 1999).

In spite of this highly proportional electoral system, South Africa has consistently experienced majoritarian outcomes. In the words of Gouws and Mitchell (2005, 353), it demonstrates “one party dominance despite perfect proportionality.” The ANC has prevailed in national elections since 1994, winning over 60 percent of the vote in all elections to date and forming majority governments from 1999 on. The effective number of seat-winning parties in the national parliament has hovered at 2.3 or lower. The ANC has also won all but a handful of provincial races. The remainder of the electorate votes for a large and fluctuating set of opposition parties, some of which do not cross the (very low) threshold for representation at the national level. The most successful opposition party during the post-apartheid period has been the Democratic Alliance (DA), which typically wins less than 20 percent of the national (p. 944) vote and controls only one provincial legislature. Thus, although South Africa’s highly proportional electoral system was adopted in part to ensure wide representation and coalition governments, South Africa has a low effective number of parties and majority party domination.

A very different institutional design, adopted in 2000, guides South Africa’s municipal elections. These utilize a mixed-member electoral system, with ward-level seats elected according to plurality in single-seat districts, and municipality-wide multiseat districts elected through closed-list proportional representation (PR). Despite the mix of rules, these elections feature ANC dominance in both components of the system. Many parties compete in the races—as they do at the national level—but only a few win seats. The average effective number of parties is low (below two) in both the wards and the PR districts. The 2016 municipal elections suggested some erosion of the ANC’s position as opposition parties pushed it out of majority control in a few large metropolitan municipalities, including Johannesburg. These shifts have rattled the ruling party and may auger larger changes, but should not be overstated: the ANC remains in control of more than three-quarters of South Africa’s municipalities.

The concentration of support in a small number of parties does not confound the institutional logic of PR. As noted by Cox (1997), district magnitude puts a ceiling on the number of parties, not a floor. However, it falls short of the aspirations of many of those who endorsed proportional representation for South Africa and serves as a useful reminder of the limitations of constitutional engineering. Institutional factors may create the broad parameters within which party systems operate, but their effects on outcomes are rarely deterministic. South Africa’s majoritarianism also begs an explanation for why the party system has resisted expansion in spite of highly permissive electoral rules.

Several contextual factors explain single-party dominance in South Africa. First, the formidable salience of racial cleavages in South Africa strongly impacts the party system. In the early post-apartheid period, the opposition parties with the most resources lacked legitimacy with black voters due to their roots in apartheid, which sharply limited their ability to win votes (Ferree 2011). Second, the ANC successfully unified most of the anti-apartheid movement in the early democratic period, starving alternative struggle parties of resources and organization (Ferree 2011). Third, once in power, the ANC exploited secondary electoral institutions like floor crossing and public financing rules to enhance its dominance (Booysen 2006; Gouws and Mitchell 2005; Booysen and Masterson 2009). Finally, like dominant parties everywhere, the ANC has used its control over the state’s fiscal resources to encourage and reward loyalty (Kroth, Larcinese, and Wehner 2016). Combined, these factors enabled the ANC to prevent elite defections, present itself as the only reasonable option to govern the country, and preserve its dominance in spite of electoral rules that place virtually no constraints on the number of parties that can win seats. The South African case thus underlines the importance of putting electoral rules in their context. Electoral institutions shape party systems, but do so in combination with other factors, and sometimes in ways that defy simple expectations.1 (p. 945)

National and Provincial Electoral Rules

South Africa is a parliamentary democracy, with parliamentary institutions at both the national and provincial levels. At the national level, it has a bicameral parliament, consisting of a ninety-seat upper house, the National Council of the Provinces (NCOP), and a four-hundred-seat lower house, the National Assembly. Parliament selects the head of state and government, the president, who then selects the cabinet. Parliament can dismiss the president and/or cabinet through a vote of no confidence, although this has not happened in the post-apartheid period.2 South Africa also has nine provincial parliaments, which range in size from eighty seats in KwaZulu-Natal to thirty seats in Free State, Mpumalanga, and Northern Cape.3 Provincial parliaments select provincial premiers, who select provincial cabinets, called executive councils. Provincial parliaments can dismiss executive councils and premiers through no-confidence votes.4

Elections for the National Assembly and the provincial legislatures have occurred concurrently every five years since 1994. Voters cast two ballots: one for the National Assembly and one for their provincial legislature. Concurrent elections, while the practice, are not constitutionally mandated. Election terms for the National Assembly are five years, but presidents can call for early elections if the National Assembly is dissolved prior to that.5 The NCOP is filled through indirect election via the nine South African provincial legislatures; its primary purpose is to represent provincial interests in national government. Each provincial legislature sends ten delegates, including the provincial premier.

The voters of South Africa—citizens above the age of 18—elect the National Assembly using a two-tier compensatory proportional representation system, with closed party lists. Parties submit ranked lists of candidates to the Independent Electoral Commission (IEC). Parties have complete discretion over their lists, and internal party procedures for generating lists vary. Most consult with local branches to gather preferences, but party leadership retains a veto in all major parties.6 Parties advertise the lists—which can include hundreds of names—prior to the election, but ballots do not include candidate names. Instead they feature the party name, symbol, and sometimes a photo of the leader. Voters cast one party vote and seats are allocated proportionally to parties according to the proportion of national votes won using the Droop quota with largest remainders (LR-Droop).7

In the seat allocation process, parties can submit up to ten lists: a national list (which is optional) and nine regional lists (one for each province). In the regional tier, 200 total seats are distributed over nine provincial multiseat constituencies proportional to provincial population. For example, based on its population, Western Cape received 23 of the 200 regional-to-national seats in 2009, Gauteng received 47 of the seats, Free State 12, and so on. These provincial allocations are distributed to parties proportionally based on their provincial vote totals in the national (p. 946) election using LR-Droop. Parties then allocate their seats to members based on the regional list for that province. The seats a party allocates from its regional lists are then summed across provinces and compared to the allocation of seats it is due based on its total national result. Any difference is topped up using the national list for the party. Thus, in 2009, the ANC won 266 of the 400 National Assembly seats based on its total national vote of 10,601,330. It secured 139 seats from the regional tier, and then received an additional 127 national-tier seats, which it allocated to members on its national list to bring it up to 266.8

The use of the two-tier list system affects how seats are distributed within parties, allowing more candidates to be seated from regional lists in provinces where the party has performed better, but does not alter the overall proportionality of the system as it is completely compensatory and the total number of seats a party receives reflects its performance in the national election. The system therefore acts as if it is one national district with a magnitude of four hundred. Large district magnitude, with no legal threshold for gaining representation, makes South Africa’s electoral system one of the most proportional in the world. In 1994, parties needed just 48,712 votes to surpass the quota for a seat—or about 0.0025 percent of the vote. Based on the seat product model (see the chapter in this volume by Shugart and Taagepera), we would expect this electoral system to result in a very fragmented party system. Given that South Africa has a two-tier compensatory system, we would expect the effective number of seat-winning parties to be around 6.4,9 which far exceeds its level in practice.

The provincial electoral system also employs closed-list proportional representation, with LR-Droop. Parties submit provincial lists to the IEC along with their national and regional-to-national lists. Election terms are five years, but elections can be called early if the provincial parliament is dissolved.10 Provincial ballots resemble national ones, with party names, symbols, and images of leaders.

The interim constitution of South Africa established these electoral institutions in 1993 after a protracted series of negotiations.11 They became permanent after the final constitution of 1996.12 They replaced a strongly majoritarian, Westminster-like parliamentary system that employed plurality rule in single-seat districts. Legendary disproportionality characterized this system and facilitated the rise and consolidation of the National Party during the apartheid period (Faure 1996).13 Adoption of South Africa’s extreme PR rules reflected both dissatisfaction with the earlier majoritarian system and the context of the transition from apartheid.

In the early 1990s, as South Africa transitioned out of the apartheid, it faced deep challenges. Negotiations over a new constitution, including electoral rules for a nonracial democracy, took place against a backdrop of political conflict.14 Race and poverty divided a country among the most unequal in the world. Civil war offered a frightening but plausible alternative to democratic transition. While negotiations primarily involved the ANC and National Party (NP), a wide set of additional actors and parties participated, some with antisystem inclinations. The Freedom Front and other extreme right groups teetered between participation in the new system and more violent strategies, as did parties representing ethnic and regional movements within the black (p. 947) population like the Inkatha Freedom Party (IFP). Moreover, both the ANC and the NP contained elements intent on derailing the process (Johnson and Schlemmer 1996).

While negotiations over the interim constitution periodically broke down over contentious details, the selection of the electoral system was one of the less controversial stages of the process, largely because of an early convergence of preferences of the main players (Lodge 2003). The NP and the representatives of other small political factions favored PR because it would increase their odds of survival (Faure 1996; Lodge 2003). With black South Africans representing 80 percent of the electorate, any party without black support would be a minority party. While the NP hoped to win black votes, it could hardly count on them, and indeed ultimately won few (Giliomee 1994). The ANC, expecting to win a majority of votes based on its popularity with black voters, initially opposed PR, viewing it as a last-ditch effort by white parties to hold on to power (Mattes 1994; Lodge 2003). By the early 1990s, however, the party had switched to embracing closed-list PR (Mattes 1994). This switch likely reflected several factors. The party believed it would obtain a majority independent of the rules, so compromising on this component of the constitution cost it little (Lodge 2003; Gouws and Mitchell 2005; Mattes 1994); Moreover, PR might allow it to secure seats for some of its less obviously electable members—for example, high-ranking whites and Indians with few links to actual black constituencies (Lodge 2003). Closed-list PR would also help party elites maintain discipline over backbenchers (Lodge 2003). And finally, the simplicity of party voting in a closed-list PR system had appeal to a party of newly enfranchised voters with limited education. Voters could find the party symbol and face of Nelson Mandela and put their mark in the correct place (Gouws and Mitchell 2005).

In adopting PR, South Africa followed the endorsements of prominent constitutional engineers like Arend Lijphart, who had longed advocated proportional rules as optimal for countries with deep racial and political fissures.15 Extreme parties with antisystem tendencies might moderate with the promise of parliamentary representation. In contrast, a majoritarian system could freeze them out, stewing in their discontent. The hoped-for outcome of the design was consensual democracy, with a large multiparty system, coalition or minority governments, and the benefits such a system brings: inclusion, representation, moderation, and compromise (see also Reynolds 1998, 1999a; Sisk 1995). The interim constitution’s mandate for a Government of National Unity, which involved leaders of the largest parties serving as deputy presidents in the cabinet, also reflected these goals.

Complaints about the electoral system periodically surface and discussions prior to the 1996 constitution debated the merits of moving to a different set of rules. A main criticism has been the lack of strong linkages between elected officials and constituents and the absence of incentives for constituency service (Mottiar 2005; Barkan 1998). The ANC made efforts to address these issues in 1994 by establishing an informal constituency system for elected members of parliament (MPs) and making resources available to support constituency service. However, absent an electoral linkage, these efforts amounted to little; few MPs keep staffed constituency offices (Barkan 1998). The ANC has not initiated more significant electoral reform. (p. 948)

Municipal Electoral Rules

Municipalities form the third and most local tier of government in South Africa. Legislation beginning in 1998 reformed the existing municipal map and established the structure, responsibilities, and electoral systems of municipal governments.16 The first municipal elections under the new rules occurred in 2000, with subsequent elections in 2006, 2011, and 2016.17 Municipal governments lead the “developmental crusade” in South Africa (Siddle 2011). The Municipal Structures Act of 1998 tasks them with the ambitious goal of overturning apartheid’s legacy of poverty and uneven development. They provide highly valued local public goods like water, electricity, education, health, and housing; set and collect local taxes; and represent the most visible face of government to many voters. During the post-apartheid period, protests about local government performance have been frequent and widespread (Atkinson 2007; Siddle 2011). For these reasons, any treatment of South African electoral and party systems should include a discussion of the municipal level of government.

South African municipalities fall into three categories. Metropolitan municipalities (category A, or colloquially “metros”) are large cities with high population densities and complex, developed economies. As of 2016, South Africa had eight metropolitan municipalities, including Cape Town, Durban, Tshwane (Pretoria), and Johannesburg. Local municipalities (category B) are smaller towns and surrounding rural areas. Most South African municipalities (207 in 2016) fall into category B. District municipalities (category C) combine and coordinate several local municipalities into a regional aggregation to capture economies of scale in service delivery. All category B local municipalities nest into overarching district municipalities (of which there were 44 as of 2016), and these two levels share responsibilities and authority for a particular area. Category A metros do not nest into a higher level.18

Municipal councils govern each type of municipality and range in size from three seats (in small local councils) to over two hundred (in large metros). Elections for councils occur every five years. Metropolitan and local municipalities employ mixed-member proportional (MMP) electoral systems, filling half the seats through single-seat plurality ward races, and the other half through municipality-wide closed-list proportional representation.19 Voters cast two ballots in metro or local elections: one for a ward councilor and one for a municipal-level closed party list. Independent candidates can run in ward races, but not for PR seats. The Droop quota for the PR race is determined by dividing the total number of all party votes in the PR and ward races by the number of seats in the council (PR and ward) not filled by independent candidates and adding one.20 To determine the total number of seats won by a party, its total votes (PR and ward) are divided by the quota. If this number exceeds the number of ward seats won, then it is compensated with PR seats using the largest-remainder system. District municipalities, on the other hand, fill 40 percent of their seats through a single district-wide PR election. Representatives selected from local councils fill the (p. 949) remaining 60 percent. Voters in local municipalities thus cast three ballots: one for ward representative, one in the PR race for the local municipality, and one in the PR race for the district council. In contrast, voters in metro municipalities cast only a ward vote and metro PR vote.

Although relatively little has been written on the process leading to these electoral rules, their adoption likely balanced the goal of maintaining proportionality while also injecting a degree of constituency representation into the local government system through ward councilors.21 The rules also opened the door to independent candidates, a significant departure from the extreme party-centered nature of the national and provincial electoral rules (Nijzink and Piombo 2005, 77).22

Like the national and provincial electoral list PR systems, the PR half of South African municipal elections should enable a large number of parties to win seats. District magnitude varies by municipality. Some PR districts involve several dozen seats. Others are small, three to five seats. The mean council size from 2000 to 2011 was around sixteen seats, implying a mean PR district magnitude of eight.23 Duvergerian logic suggests that the effective number of parties should be lower for ward elections versus municipal elections, especially over time as voters gain experience with the system.24

The South African Party System

While the guiding principle of South Africa’s electoral system design was multipartism and consensual democracy, the predominant theme of the post-apartheid South African party system has been ANC dominance at all levels of government. The ANC frequently captures more than 60 percent of the vote in races small and large, and in many parts of the country regularly exceeds 80 percent of the vote. A secondary theme has been opposition fragmentation. The second-largest party typically wins 20 percent or less of the national vote, while a large and shifting set of smaller parties divide the remainder. Recent elections have seen ANC dominance fray somewhat, especially in urban areas, but the party continues to control national government, eight out of nine provinces, and most municipalities outside of Western Cape. Coalitions have been rare in South African government except at the municipal level, especially after the 2016 municipal elections, which produced coalition or minority governments in 13 percent of municipalities, including a handful of large metros like Johannesburg.25

Prior to the 1994 election, the ANC successfully unified the large majority of black voters, approximately 80 percent of the electorate, under its great tent.26 By continuing to hold the loyalty of this group, the party has won national elections in South Africa by wide margins since 1994 (see Table 44.1). Indeed, in spite of South Africa’s highly proportional electoral rules, there has only been one instance of coalition government at the national level: the Government of National Unity (GNU), which drew together the ANC, the National Party, and the IFP during the first parliament (1994–1999), and it transpired even though the ANC had a majority of seats in parliament.27 (p. 950)

Table 44.1 The National Party System






Number of parties (votes)






Number of parties (seats)






Effective number of parties (seats)






Largest party






Largest party vote percent






Second-largest party






Second-largest party Vote percent






Coalition government?






Calculations by author based on data from the Electoral Commission of South Africa’s website at (accessed August 29, 2016).

Table 44.2 The Provincial Party System






Average effective number of parties (seats)






Average number of parties (votes)






Average number of parties (seats)






Average ANC seat percentage






ANC provinces






Opposition provinces






Coalition governments






(*) Western Cape and KwaZulu Natal.

Calculations by author based on data from the Electoral Commission of South Africa’s website at (accessed August 29, 2016).

The ANC has also dominated all but a few provincial legislatures: over five elections in nine provinces, only four provincial governments have featured majority control by a party other than the ANC: Western Cape in 1994, 2009, and 2014, and KwaZulu Natal in 1994. A handful of provincial coalitions also occurred in 1999 and 2004 as power shifted in Western Cape and KwaZulu-Natal (see Table 44.2). Of some note, the ANC’s hold over Gauteng, the industrial heartland of the country with both Pretoria and Johannesburg, slipped to 54 percent in the 2014 election, the lowest ever. ANC support also dropped in traditional strongholds like Limpopo, Mpumalanga, and Free State. Recent erosions of (p. 951) ANC dominance should be put in context, however: the party remained at 70 percent or above in all three provinces and still took an average of 66 percent of the seats across provincial legislatures.

Table 44.3 The Municipal Party System













Average effective number of parties (votes)









Average number of parties (votes)









ANC vote percentage









DA vote percentage









Effective number of parties for 2000–2011 calculated by author using disaggregated data from the IEC; see details in Ferree et al. (2017). Calculations not completed for 2016. ANC and DA PR vote percentages for 2000 election reported by Electoral Institute of South Africa (EISA), “South African Local Election Results 2000: Proportional Representation Vote Comparison by Province,” n.d. ANC and DA vote percentages for 2006 election reported by EISA, “South Africa: 2006 National Overview—Votes by Political Party,” n.d. ANC and DA percentages for 2011 and 2016 on IEC webpage: (accessed August 29, 2016).

Municipal elections present a more varied picture. The ANC dominates these elections in terms of control over municipal councils. Prior to 2016, opposition parties had firm footholds only in Western Cape, where the DA won control over half of the municipal councils in 2011, and KwaZulu Natal, where the IFP and smaller parties prevented the ANC from sweeping all municipalities. In 2016, the ANC lost majority control over a few large and important municipalities outside these provinces—Nelson Mandela Bay, Johannesburg, and Tshwane (Pretoria)—an outcome that filled newspaper headlines and induced jubilation in the opposition.28 Nonetheless, the ANC continued to control over three-quarters of all municipalities in the country. Of some note, average ANC vote percentages have fallen below 70 percent of the vote in both ward and PR races since 2000, and in 2016, it won an all-time low of 53 percent of ward votes and 55 percent of PR votes (see Table 44.3).

The opposition divides a small pie of votes into many slivers. Since 1994, the opposition has featured one main party surrounded by a constellation of much smaller ones. The main opposition party generally wins no more than 20 percent of the vote in national elections, takes at most one provincial legislature, and controls less than 10 percent of all municipal councils.

Initially the primary opposition party was the former apartheid ruling party, the NP, which became the New National Party (NNP) in 1997. The NP won 20 percent of the votes in 1994 and controlled Western Cape.29 In 1999, the Democratic Party (DP)—the descendent of apartheid-era opposition parties—eclipsed the NNP, but only won (p. 952) 10 percent of the vote overall.30 The NNP and DP merged to form the DA prior to the 2000 local government elections,31 and the DA has been the primary opposition party since then. The DA draws support largely from minority voters—primarily whites, but also increasingly coloureds and Indians. It won 12 percent of the national vote in 2004, 17 percent in 2009, and 22 percent in 2014. It has controlled Western Cape provincial governments since 2009, demonstrating its popularity among the white and coloured voters that make up most of the population there. It performed relatively well in the 2000 municipal elections, winning 22 percent of the ward and PR vote and taking control of a number of municipalities in Western Cape. In 2006, its performance flagged as its vote percentages dropped to 16 percent of the vote. It rebounded in 2011 and especially 2016, when it captured 27 percent of the vote and pushed the ANC out of majority control of key municipalities. Recent races suggest that the DA may be making some inroads into the black electorate—strong performances in Gauteng and Eastern Cape would require some degree of cross-over voting by blacks—but the majority of the black electorate, especially in rural areas, continues to avoid the party.

While the DA is the primary opposition party, South Africa features a large set of smaller single-digit and “micro” parties (Booysen 2006). The larger and more significant of these parties attract mostly black votes and modestly chip away at ANC control in one or more regions. They tend to flare up for a period of time and then collapse back into the ruling party or persist as one-man or -woman operations. The longest lasting and most significant of these has been the Inkatha Freedom Party (IFP). Under the leadership of Mangosutho Buthelezi, the IFP exploited its deep roots and patronage networks from the KwaZulu Bantustan (KZN) to take majority control over the KZN parliament in 1994.32 It ruled in coalition with the ANC in 1999 and 2004. A significant and controversial player in the period leading up to elections and the early post-apartheid period, its importance has since waned.33 It continues to win small municipalities in KZN (now competing with its own splinter, the National Freedom Party), but otherwise has little impact. Two other modestly successful parties that built on old Bantustan roots were the United Democratic Movement (UDM), led by Bantu Holomisa, and the United Christian Democratic Party (UCDP), led by Lucas Mangope. Both Holomisa and Mangope had previous incarnations as apartheid-era homeland leaders. The UDM drew votes from the ANC in Eastern Cape in 1999 and 2004 before retreating to obscurity (Piombo 1999). The UCDP was the second-largest party in the North West province in 1999 and 2004 but has had diminishing impact since then. Outside of the IFP (which presented a regional but not national challenge), none of these parties seriously threatened the ANC.

Beginning in 2009, a different type of opposition party emerged: one based more on ideology, with a broader, less regionally centered constituency. In 2009, the Congress of the People (COPE) formed out of elite splits in the ANC and opposition to the ascension of Jacob Zuma to party leadership at the divisive 2007 Polokwane ANC party conference. COPE initially appeared poised to draw significant high-level defections from the ANC in the run-up to the 2009 campaigns, and at one point predications suggested it might win as much as 40 percent of the vote (Ferree 2011, 215). It carved out a political (p. 953) space in the center of the spectrum, differentiating itself from the ANC less on policy and more on promises to implement programs more effectively and with less corruption. Although it initially appeared more dangerous to the ruling party than previous contenders, the ANC managed to stem elite defections and ran a strong campaign that ultimately cauterized COPE’s impact, holding it to just 7.4 percent of the national vote and thirty National Assembly seats in 2009 (Ferree 2011). In 2014, COPE’s support fell to under 1 percent of the vote and just three seats.

The 2014 election saw the emergence of the Economic Freedom Fighters (EFF). The populist EFF targets the left side of the political spectrum, campaigning on a hybrid ideology of African nationalism and socialism (Robinson 2014). In so doing, it taps into a deep river of African political tradition in South Africa that was largely quiescent during the early post-apartheid period. The EFF’s leader, Julius Malema, formerly headed the Youth League of the ANC; known as brazen and twice convicted for hate speech, he was expelled from the ANC in 2012 amid controversy. The EFF quickly became popular in urban and semiurban communities in Gauteng, North West province, and Limpopo, where its radical economic message appeals to voters. It became the third-largest party in South Africa in the 2014 election, winning 6.4 percent of the vote and capturing twenty-five National Assembly seats. In the 2016 municipal elections, it increased its support to around 8 percent of the vote. While it is too early to tell whether the EFF will survive further election cycles, it represents a different and potentially more dangerous challenge to the ANC than its predecessors, with an ideological message that strongly resonates with an electorate weary from decades of high unemployment. The two-sided challenge of the EFF on the left and the DA on the right certainly represents something new for the ANC (Southall and Schulz-Herzenberg 2014). While the EFF and DA represent unlikely partners, they do periodically unite to oppose the ruling party and together they may force it into coalition government. Nonetheless, the ANC still has a very formidable hold on the electorate, with 62 percent of the national vote in 2014 and control of eight provinces and more than three-quarters of all municipalities.

In the aggregate picture, the number of parties winning votes in national elections has steadily increased since 1994, when 19 parties won votes, to 2014, when 29 did. The same trend applies to provincial elections: in 1994, an average of 10 parties won votes, whereas an average of 19 did in 2014. In fact, across every provincial legislature, the number of parties winning votes has increased in every single election since 1994. In this sense, South Africa’s proportional system produces the expected outcome: a large (and growing) party system. However, while the number of parties competing in and winning votes has grown, the number of parties winning seats has not changed much at all. There was a big bump in parties winning seats in the National Assembly between 1994 and 1999 (from 7 to 12), but no significant change since. In the provincial elections, an average of 4.3 parties won seats in 1994; in 2014, it was 4.6. In terms of the effective number of parties, very little has changed: in the national party system, the effective number of seat-winning parties was 2.21 in 1994 and 2.26 in 2014, after a dip in the intervening years. In the provincial elections, the (p. 954) effective number of seat-winning parties ranged from 1.66 to 2.11. The low number of parties is not what we would expect for an extreme PR system.

Finally, while we might expect the size of ward party systems (with their single-seat plurality rules) to differ substantially from the size of municipal PR systems, this does not appear to be the case. As Table 44.3 shows, the average effective number of parties for each component of the municipal system was highly similar across the first three municipal elections: 1.7 versus 1.8 in 2000; 1.7 versus 1.9 in 2006; and 1.8 versus 2.0 in 2011. While the PR component produces slightly more parties, the overall pattern remains one of ANC dominance and low effective number of parties regardless of rules. Of some note, racial diversity drives the number of parties in both the single-seat district and PR sides of the system (Ferree et al. 2017), contradicting the well-known interactive hypothesis that social diversity should only shape party system outcomes in permissive electoral systems (Amorim-Neto and Cox 1997; Clark and Golder 2006; Ordeshook and Shvetsova 1994).34Ferree et al. (2017) suggest that the salience of racial divisions in South Africa may interfere with strategic behavior on the part of voters and parties, reducing defections from likely losers and preventing Duvergerian coordination. These results once again demonstrate the limits of institutional theories in South Africa.

Explaining Single-Party Dominance in South Africa

With the exception of the single seat plurality rules that guide ward elections at the municipal level, South Africa’s electoral rules place few constraints on its party system. Yet, while many parties win votes, few win more than a handful of seats, and power remains concentrated in the ruling ANC. Why? To explain single-party dominance in South Africa, one must look beyond the electoral rules to other features of the social and political context.

First, there is the formidable salience of racial cleavages in South Africa and the way in which these cleavages shape perceptions about parties. In the early post-apartheid period, the opposition parties with financial resources, ground organization, and campaign experience had little legitimacy with nonwhite voters, who made up the great majority of the country. The National Party had built the apartheid state, one of the most racially repressive institutions in the world, and it would take more than the release of Nelson Mandela and legalization of the ANC to reform its image in the black electorate. The Democratic Party was the latest manifestation of a series of parties that opposed apartheid “from within.” These parties, which started with the Progressive Party in 1959, never captured much of the (all-white) vote, but did manage to periodically shine light on some apartheid-era abuses. While DP leaders hoped that this legacy would win the party black votes, it entered the apartheid period relatively unknown. Those black voters who did know it largely discounted its participation in the apartheid struggle. Thus, (p. 955) while the NP and DP had access to resources that many other opposition parties lacked, they entered the post-apartheid period with party images that significantly limited their appeal to most of the electorate. Most black voters saw them as exclusively “white,” a label the Democratic Alliance inherited when it formed in 2000. As these perceptions of opposition parties fortified the ruling party’s claim to being the only legitimate party to govern South Africa, it fought hard and successfully to preserve them (Ferree 2011).

Of some note, this explanation of how social factors impact the party system is distinct from those offered by prior theories, especially the interactive hypothesis. The interactive hypothesis suggests that large party systems require both permissive rules and significant social diversity. A small effective number of parties in a permissive system might therefore reflect a lack of social diversity. This is clearly not the explanation for South Africa. In addition to its four racial groups, it has thirteen ethnolinguistic groups. If social diversity and permissive rules are enough to generate a large party system, then South Africa should have one. Thus, the effects of social factors are not limited simply to “diversity.” Beyond the size and number of groups, it is important to understand how particular cleavages drive political competition and shape the space of “legitimate” political organizations.

Second, the ANC successfully unified most of the anti-apartheid movement under its sweeping umbrella, monopolizing human capital, grassroots structures, and access to finance. Alternative opposition parties with struggle credentials existed during the early post-apartheid years. The Pan African Congress (PAC) had rich roots in the apartheid struggle and hopes of attracting significant support in the 1994 election, as did Azapo (the Azanian People’s Organization). Without exception, however, these parties lacked financial, human, and organizational resources (Cooper 1994). They won slivers of the vote in early elections before fading into obscurity, prevented from capitalizing on their histories by resource constraints and internal squabbles. Other black opposition parties like the IFP, UDM, and UCDP had roots in the old Bantustan system and could build on existing patron-client networks. While they created greater headaches for the ANC, they were regional in nature, conservative, built on the support of only one ethnic group, and lacking in wider legitimacy. They too faded away as the ANC cemented its control over state structures. Finally, altogether new parties attempting to build themselves out of whole cloth faced formidable challenges on multiple fronts: acquiring money to run campaigns, building structures at the local level, developing attractive party labels, and finding viable candidates. The most obvious route to developing a new party may be to splinter away from an old one, but the experience of COPE, which was formed by ANC defectors who hoped to initiate a wave of additional defections, demonstrates the difficulty of following this path (Ferree 2011). The ruling party has ample means of punishing defectors, banishing them to political obscurity and curtailing their ability to move into lucrative careers in the private sector. More recently, the EFF also formed as a splinter from the ANC. It may ultimately draw more blood from the ruling party, but as of this writing, it is too early to say.

Third, the ANC has used its ability to shape secondary electoral institutions like floor-crossing rules and the public financing of political parties to tilt the playing field (p. 956) in its favor. Floor-crossing legislation in South Africa has followed an arc, from prohibition, to short periods in which it was permitted, back to prohibition. The primary effect of this arc of legislation was to cement ANC dominance and establish the DA as the largest opposition party. Party financing laws also favor the largest parties and deter efforts to form new ones, creating status quo bias in the party system. Both points are developed next.

The interim and 1996 constitutions prohibited floor crossing at all levels of government; MPs who did defect had to give up their seats.35 In practice, the rule may have helped the ANC discipline its own members by eliminating exit options, allowing the party to “redeploy” members at the whim of party leadership (Gouws and Mitchell 2005; Lodge 1999). In spite of this convenience, in 2002–2003 the ruling party cooperated with the recently formed DA to amend the constitution to allow floor crossings at all levels of government. Crossing occurred during specific windows of time determined by the executive.36 The first transfer window at the local level took place in October 2002; in provincial and national parliaments, it took place in March 2003. The ANC gained net seats at all levels of government by absorbing MPs from the NNP and other opposition parties during the initial floor-crossing periods. The DA also gained during the initial crossing windows, attracting the remaining representatives of the NNP to become the largest opposition party. The changes decimated the NNP and many smaller parties (Booysen 2006; Gouws and Mitchell 2005). The elections of 2004 further entrenched these patterns, as did later floor-crossing windows.37 While enhancing ANC dominance, floor crossing also generated tensions within the party as defectors sometimes received higher positions on party lists than party stalwarts. After its contentious national conference at Polokwane in 2007, the party reversed its support of floor crossing and formally abolished it through an amendment of the constitution in January 2009 (Booysen and Masterson 2009).

Public financing of political parties provides another institutional resource for the ruling party. The Public Funding of Represented Political Parties Act (1997, 2(1)) established a fund to provide parties represented in the National Assembly and provincial legislatures with financial support. Parties receive funding proportional to their level of representation. The Independent Electoral Commission administers the fund and has the ability to audit and sanction parties that abuse their use of the money.38 While public financing promotes transparency and regulation (as opposed to the private financing of parties, which is neither transparent nor well regulated in South Africa), it generates status quo bias in the party system. Because public financing is proportional to representation, the ANC’s allocation dwarfs that of its competitors, reinforcing the advantages it already has as the ruling party. In 2009, for example, the ANC received over five times as much public financing as the second- and third-largest parties combined (Booysen and Masterson 2009, 415). Furthermore, because new parties do not receive funding until they gain representation, they face financial challenges during first campaigns. This reduces incentives for politicians to break away from existing (funded) parties to start new ones. Moreover, leaders of parties, even small ones, may be reluctant to merge with other parties if it means giving up an independent source of financing. As a result, public (p. 957) financing creates inertia in the party system, reinforcing existing patterns and levels of support.

Finally, in the tradition of dominant parties around the world (Magaloni 2006), the ANC used its control over the state’s fiscal resources to encourage and reward loyalty. The party has delivered highly desired services to an electorate that previous governments massively underserved and neglected. While this may not guarantee voter loyalty (de Kadt and Lieberman forthcoming), it likely explains some of the electorate’s continued support for the ruling party. The ANC also appears to engage in some resource targeting to areas that supported it in the past (Kroth, Larcinese, and Wehner 2016). Combined with the factors discussed earlier, these moves have helped ensure the ANC’s hold over the South African electorate.


South Africa’s national electoral system, one of the most proportional in the world, was designed explicitly with the goal of ensuring representation in parliament for a wide variety of groups and parties. In spite of these rules, the ANC has dominated all levels of government, and the effective number of legislative parties in national, provincial, and municipal institutions is far lower than predicted by well-established institutional theories. South Africa thus represents a compelling puzzle for institutional theories and illustrates the way in which contextual factors impact the effects of institutions on the party system.

Ferree et al. (2014) distinguish two types of contextual effects. Context may winnow down the number of parties when electoral institutions are permissive. Alternatively, contextual factors may interfere with the strategic behavior necessary to generate coordination and reduce party number in restrictive systems. South Africa primarily demonstrates the former effect. Contextual factors drive down the number of parties, concentrating support in one major party in spite of electoral rules that impose few constraints and allow even very small parties to win seats. South Africa may also demonstrate the latter effect in its single-seat municipal ward elections, which appear to violate Duvergerian expectations when racial diversity is high. South Africa thus illustrates two primary avenues through which context can shape outcomes: winnowing and coordination inhibition.

The South African case also argues for the importance of a wider range of contextual variables. Existing “contextual” explanations for party systems generally focus on one of two factors: the age and/or institutionalization of the party systems and social diversity (Moser and Scheiner 2012).39 Neither offers a sufficient explanation for South Africa. The relative youth of South Africa’s democracy may offer a partial account for its party system. It is possible, even likely, that the dominance of the ANC will eventually crumble. Yet ANC dominance has persisted over twenty years and four national election cycles, and at some point, “youth” offers too thin (p. 958) an explanation for its success in holding on to power. The standard wisdom on how social diversity affects party systems is captured by the interactive hypothesis, which argues that large party systems require both permissive rules and high social diversity. Clearly, this hypothesis offers little insight into South Africa, which has a small party system in spite of both permissive rules and one of the most diverse societies on the planet.

South Africa’s party system thus reflects factors beyond democratic age and social diversity. It refocuses attention not on the size and distribution of groups, but rather on the intensity of certain cleavages and how they shape the political terrain negotiated by parties. Social diversity may not shape the party system in a direct and linear way (more groups, more parties), but the salience of social divisions, specifically race, does clearly matter. Moreover, the South African case points to a fuller set of contextual factors including other types of institutions like floor-crossing legislation and party financing laws. Electoral rules are but one piece in a broad mosaic of rules and institutions that shape party system outcomes. More than anything, the case reveals the ways that dominant parties, once ensconced in power through some fortuitous confluence of factors, fortify their position by manipulating institutions, flows of resources, and perceptions of their opponents to bend the expectations generated by both constitutions and cleavages.

Author’s Note

All tables were prepared by the author specially for this chapter.


Álvarez-Rivera, Manuel. “Election Resources on the Internet: The Republic of South Africa Electoral System,” 2016, (accessed July 18, 2016).

Amorim Neto, Octavio and Gary W. Cox. “Electoral Institutions, Cleavage Structures, and the Number of Parties”, American Journal of Political Science 41 (January 1997): 149–74.Find this resource:

Atkinson, Doreen. “Taking to the Streets: Has Developmental Local Government Failed in South Africa?” In State of the Nation South Africa 2007, edited by Sakhela Buhlungu, John Daniel, Roger Southall, and Jessica Lutchman, 53–77. Cape Town: Human Sciences Research Council, 2007.Find this resource:

Barkan, Joel. “Rethinking the Applicability of Proportional Representation for Africa.” In Elections and Conflict Management in Africa, edited by Timothy D. Sisk and Andrew Reynolds, 57–70. Washington, DC: United States Institute of Peace, 1998Find this resource:

Booysen, Susan. “The Will of the Parties versus the Will of the People? Defections, Elections and Alliances in South Africa.” Party Politics 12, no. 6 (2006): 727–746.Find this resource:

Booysen, Susan, and Grant Masterson. “South Africa.” In Compendium of Elections in Southern Africa 1989-2009: 20 Years of the Multiparty Democracy, edited by Denis Kadima and Susan Booysen, 402–403. Johannesburg: EISA, 2009.Find this resource:

(p. 962) Clark, William Roberts, and Matt Golder. “Rehabilitating Duverger’s Theory: Testing the Mechanical and Strategic Modifying Effects of Electoral Laws”, Comparative Political Studies 39 (August 2006): 679–708.Find this resource:

Cooper, Saths. “The PAC and AZAPO.” In Election ’94: The Campaigns, Results and Future Prospects, ed. Andrew Reynolds, 117–120. New York: St. Martin’s Press, 1994.Find this resource:

Cox, Gary. Making Votes Count. Cambridge: Cambridge University Press, 1997.Find this resource:

Davis, Gavin. “Bridges and Bonds: List Proportional Representation and Campaigns in South Africa.” Working Paper No. 50, Centre for Social Science Research, Cape Town, 2003.Find this resource:

de Kadt, Daniel, and Evan S. Lieberman. “Nuanced accountability: Voter responses to service provision in South Africa.” Forthcoming, British Journal of Political Science, forthcoming.Find this resource:

Electoral Commission of South Africa. “Municipal Election Results.” n.d. (accessed May 5, 2017).

Electoral Commission of South Africa. “2016 Municipal Elections Handbook.” 2016. (accessed August 19, 2016).

Electoral Institute of South Africa (EISA). “South Africa: Political Party Funding.” African Democracy Encyclopaedia Project. n.d. (accessed August 19, 2016).Find this resource:

Electoral Institute of South Africa (EISA). “South African Local Election Results 2000: Proportional Representation Vote Comparison by Province.” n.d. (accessed August 29, 2016).

Electoral Institute of South Africa (EISA). “South Africa: 2006 National Overview—Votes by Political Party.” n.d. (accessed August 29, 2016).

Faure, Murray. “The Electoral System.” In South Africa: Designing New Political Institutions, edited by Murray Faure and Jan-Erik Lane, 89–104. London: Sage Publications, 1996.Find this resource:

Ferrara, Federico, Erik S. Herron, and Misa Nishikawa. Mixed Electoral Systems: Contamination and Its Consequences. New York: Palgrave Macmillan, 2005.Find this resource:

Ferree, Karen. Framing the Race: the Political Origins of Racial-Census Elections. Cambridge: Cambridge University Press, 2011.Find this resource:

Ferree, Karen E., G. Bingham Powell, and Ethan Scheiner. “Context, Electoral Rules, and Party Systems.” Annual Review of Political Science 17 (2014): 421–439.Find this resource:

Ferree, Karen E., Clark C. Gibson, and Barak Hoffman. “Why the Salience of Social Divisions Matters in Party Systems: Testing the Interactive Hypothesis in South Africa.” Party Politics (2017). Published online May 9, 2017.

Giliomee, Herman. “The National Party’s Campaign for a Liberation Election.” In Election ’94 South Africa, edited by Andrew Reynolds, 43–72. New York: St. Martin’s Press, 1994.Find this resource:

Gouws, Amanda, and Paul Mitchell. “South Africa: One Party Dominance Despite Perfect Proportionality.” In The Politics of Electoral Systems, edited by Michael Gallagher and Paul Mitchell, 353–374. Oxford UK: Oxford University Press, 2005.Find this resource:

Greenberg, Stanley. Dispatches from the War Room: In Trenches with Five Extraordinary Leaders. New York: Thomas Dunne Books, St. Martin’s Press, 2009.Find this resource:

Hamilton, Georgina, and Gerhard Mare. “The Inkatha Freedom Party.” In Election ’94 South Africa, edited by Andrew Reynolds. New York: St. Martin’s Press, 1994.Find this resource:

Horowitz, Donald. A Democratic South Africa? Berkeley: University of California Press, 1991.Find this resource:

(p. 963) Johnson, R. W., and Lawrence Schlemmer. “Introduction: The Transition to Democracy.” In Launching Democracy in South Africa: The First Open Election, April 1994, edited by R. W. Johnson and Lawrence Schlemmer, 1–15. New Haven, CT: Yale University Press, 1996.Find this resource:

Legislative Sector South Africa. “Legislatures.” n.d. (accessed July 18, 2016).

Lijphart, Arend. Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. New Haven, CT: Yale University Press, 1999.Find this resource:

Lodge, Tom. Consolidating Democracy: South Africa’s Second Popular Election Johannesburg: Witwatersrand University Press, 1999.Find this resource:

Lodge, Tom. “How the South African Electoral System Was Negotiated.” Journal of African Elections 2, no. 1 (2003): 71–76.Find this resource:

Lodge, Tom, and Bill Nasson. All Here and Now: Black Politics in South Africa in the 1980s. London: C. Hurst and Company, 1991.Find this resource:

Magaloni, Beatriz. Voting for Autocracy: Hegemonic Party Survival and Its Demise in Mexico. Cambridge UK: Cambridge University Press, 2006.Find this resource:

Mare, Gerhard. “The Inkatha Freedom Party.” In Election ’99 South Africa: From Mandela to Mbeki, edited by Andrew Reynolds, 101–113. New York: St. Martin’s Press, 1999.Find this resource:

Mattes, Robert. “The Road to Democracy: From 2 February 1990 to 27 April 1994.” In Election ’94 South Africa, edited by Andrew Reynolds, 1–22. New York: St. Martin’s Press, 1994.Find this resource:

Moser Robert G., and Ethan Scheiner. Electoral Systems and Political Context: How the Effects of Rules Vary Across New and Established Democracies. Cambridge: Cambridge University Press, 2012.Find this resource:

Mottiar, Shauna. “Elections and the Electoral System in South Africa: Beyond Free and Fair Elections.” CPS Policy Brief 39. Johannesburg: Centre for Policy Studies, 2005.Find this resource:

Ndletyana, Mcebisi. “Municipal Elections 2006: Protests, Independent Candidates, and Cross-Border Municipalities.” In State of the Nation: South Africa 2007, edited by Sakhela Buhlungu, John Daniel, Roger Southall, and Jessica Lutchman, 95–113. Cape Town: HSRC Press, 2007.Find this resource:

Nijzink, Lia, and Jessica Piombo. “Parliament and the Electoral System: How Are South Africans Being Represented?” In Electoral Politics in South Africa: Assessing the First Democratic Decade, edited by Jessica Piombo and Lia Nijzink, 64–86. New York: Palgrave, 2005.Find this resource:

Norris, Pippa. Electoral Engineering: Voting Rule and Political Behavior. Cambridge: Cambridge University Press, 2004.Find this resource:

Ordeshook, Peter C., and Olga V. Shvetsova. “Ethnic Heterogeneity, District Magnitude, and the Number of Parties”, American Journal of Political Science 38 (February 1994): 100–123.Find this resource:

Piombo, Jessica. “The UCDP, Minority Front, ACDP, and Federal Alliance.” In Election ’99 South Africa: From Mandela to Mbeki, edited by Andrew Reynolds, 133–146. New York: St. Martin’s Press, 1999.Find this resource:

Piper, Laurence. “The Inkatha Freedom Party: Between the Impossible and the Ineffective.” In Electoral Politics in South Africa: Assessing the First Democratic Decade, edited by Jessica Piombo and Lia Nijzink, 148–165. New York: Palgrave, 2005.Find this resource:

Piper, Laurence. “Inkatha Freedom Party: The Elephants’ Graveyard.” In Election 2014: The Campaigns, Results, and Future Prospects, edited by Collette Schulz-Herzenberg and Roger Southall, 89–103. Auckland Park, South Africa: Jacana Media, 2014.Find this resource:

Republic of South Africa. Constitution of 1996. n.d. (accessed July 18, 2016).

(p. 964) Republic of South Africa. Municipal Demarcation Act (No. 27 of 1998). n.d. (accessed August 23, 2016).

Republic of South Africa. Local Government Municipal Structures Act (No. 117 of 1998). n.d. (accessed August 23, 2016).

Reynolds, Andrew. Election ’94 South Africa: The Campaigns, Results, and Future Prospects. New York: St. Martin’s Press, 1994.Find this resource:

Reynolds, Andrew. “Elections in Southern Africa: The Case for Proportionality, a Rebuttal.” In Elections and Conflict Management in Africa, 71–80. Washington, DC: United States Institute of Peace Press, 1998.Find this resource:

Reynolds, Andrew. Electoral Systems and Democratization in Southern Africa. Oxford: Oxford University Press, 1999a.Find this resource:

Reynolds, Andrew. Election ’99 South Africa: From Mandela to Mbeki. New York: St. Martin’s Press, 1999b.Find this resource:

Robinson, Jason. “The Economic Freedom Fighters: Birth of a Giant?” In Election 2014: The Campaigns, Results, and Future Prospects, edited by Collette Schulz-Herzenberg and Roger Southall, 72–88. Auckland Park, South Africa: Jacana Media, 2014.Find this resource:

Siddle, Andrew M. “Decentralisation in South African Local Government: A Critical Evaluation,” Doctoral Thesis, Graduate School of Business, University of Cape Town, South Africa, 2011.Find this resource:

Sisk, Timothy D. “Electoral System Choice in South Africa: Implications for Intergroup Moderation.” Nationalism and Ethnic Politics 1, no. 2 (1995): 178–204.Find this resource:

Southall, Roger, and Collette Schulz-Herzenberg. “The Party System and Political Prospects in the Wake of Election 2014.” In Election 2014: The Campaigns, Results, and Future Prospects, edited by Collette Schulz-Herzenberg and Roger Southall, 228–240. Auckland Park, South Africa: Jacana Media, 2014.Find this resource:

Kroth, Verena, Valentino Larcinese, and Joachim Wehner. “A Better Life for All?” Democratization and Electrification in Post-Apartheid South Africa.” Journal of Politics 78, no. 3 (July 2016): 774–791.Find this resource:


(1.) On context and electoral rules, see Ferree, Powell, and Scheiner (2014) and Moser and Scheiner (2012).

(2.) See Republic of South Africa, Constitution of 1996, chap. 4 (Parliament) and chap. 5 (The President and National Executive). The constitution can be found online at (accessed July 18, 2016).

(3.) These seat counts reflect the situation after the 2014 election. Information on the provincial legislatures can be found at Legislative Sector South Africa, “Legislatures,” (accessed July 18, 2016).

(4.) Constitution of the Republic of South Africa, 1996, chap. 6 (Provinces).

(5.) Constitution of the Republic of South Africa, 1996, chap. 4 (Parliament), Sections 49 and 50. According to Section 50, the president must dissolve the National Assembly if three years have passed since the assembly was elected and the assembly adopts a resolution to dissolve with a supporting vote from a majority of its members. Section 50 also indicates that an acting president must dissolve the National Assembly if there is a vacancy in the office of president and the assembly has not elected a new president within thirty days of the occurrence of the vacancy.

(7.) The Droop quota is the number of valid votes divided by one plus the number of seats to be allocated. After application of the quota, the parties with the largest vote remainders (their votes minus votes used up already on quota seats) obtain the remaining seats. See the introduction to the volume for details.

(8.) For a clear explication of the arithmetic behind list allocations, see Manuel Álvarez-Rivera’s “Election Resources on the Internet: The Republic of South Africa Electoral System,” (accessed July 18, 2016). See also Faure (1996).

(9.) The Shugart-Taagepera formula is NS = 2.5t(MSB)1/6, where t is the share of total seats allocated to the upper tier, and MSB is the product of the average magnitude of the basic tier and the total number of seats in that tier. In South Africa, both tiers have two hundred seats, so t = 0.5. The average magnitude of the basic tier is 22.22 and the total number of seats in that tier is two hundred.

(10.) Provincial legislatures are dissolved in a fashion similar to that of the National Assembly. The premier must dissolve the provincial legislature if three years have passed since it was elected and a majority of the legislature supports a resolution to dissolve.

(11.) More specifically, Schedule 2 of the interim constitution—Act 200 of 1993—and the Electoral Act of 1994. See Faure (1996).

(12.) The constitution of 1996, approved by the Constitutional Court on December 4, 1996, calls for proportional representation but does not stipulate details (Gouws and Mitchell 2005, 358, fn.10). It indicated that the 1999 election should be conducted according to the interim rules but required parliament to enact legislation for elections after 1999. The Electoral Act 73 of 1998 (Schedule 1) and Electoral Laws Amendment Act of 2003 provided this legislation.

(13.) Faure (1996, 90) notes that in the 1948 election, which put the National Party in power and marked the beginning of the apartheid period, the NP (and Afrikaner Verbond) won 42 percent of the vote and seventy-nine seats, while the United/Labour Parties won 52 percent of the vote and only seventy-one seats. During the 1961–1981 period, the NP held 75 percent of the seats in parliament with a far smaller percentage of the vote.

(14.) The negotiations were conducted at the Convention for a Democratic South Africa (Codesa), which ran between December 1991 and May 1992, when they broke down over disagreements between the ANC and NP. New negotiations started in April 1993 at the Multi-Party Negotiating Forum and culminated in the interim constitution and elections of April 1994. See Lodge (2003).

(15.) The choice of PR was not without intellectual critics, however. Horowitz (1991) questioned the consensus on proportional representation. He argued that an alternative vote (AV) system was preferable because it would create incentives for politicians to build coalitions and pursue “vote pooling,” which would help bridge racial and ethnic divisions. Barkan (1998) registered a second critique, arguing that a constituency-based system was more appropriate for Africa’s rural electorates, which needed a local, embedded elected representative. PR risked disconnecting the state from the population. Finally, Davis (2003), building off of Horowitz (1991) and Norris (2004), argued that PR encourages parties to engage in “bonding” campaign strategies (i.e., appealing to a narrow base), as opposed to “bridging” strategies that attempt to appeal to many groups and suggested PR has had this effect in South Africa.

(16.) The primary pieces of legislation include the Municipal Demarcation Act (No. 27 of 1998) and the Local Government Municipal Structures Act (No. 117 of 1998). There have been several amendments building on this legislation. Legislation can be accessed at (accessed August 23, 2016).

(17.) South Africa’s first post-apartheid local government elections occurred in 1995, but these still were considered transitional as they used old boundaries.

(18.) For an overview, see Electoral Commission (2016).

(19.) Very small councils do not have wards, and hence the system here is pure PR. If there is an odd number of seats on the council, the number of ward seats is rounded up. Hence, a council with twenty-one seats would have eleven ward and ten PR seats. See Electoral Commission (2016, 72).

(20.) The formula is VMC+1, where V is the total number of valid votes cast for all parties (in ward and PR races), M is the number of seats on the metro or local council, and C is the number of independent ward councilors elected in the election. See Section 12(1) of the Local Government Municipal Structures Act (p.70) and Electoral Commission (2016, 72).

(21.) Reynolds (1998) notes that options on the table for reforming the South African national electoral system in the late 1990s included a mixed-member proportional system like Germany’s (see Zittel’s chapter in this volume). This reform was obviously not adopted for the national system, but was reflected in the electoral system adopted for municipal elections.

(22.) The challenges facing independent candidates have nonetheless been significant. See Ndletyana (2007) for case studies of two candidates who stood as independents in the 2000 municipal elections, attempting to forge viable paths outside the main parties.

(23.) See Ferree, Gibson, and Hoffman (2017, Appendix Table 44.1).

(24.) The literature is somewhat divided on whether Duverger’s law should hold in the district tier of mixed-member systems. Some have argued that parties may run extra candidates in the districts to help advertise and draw votes to their PR lists (see Ferrara, Herron, and Nishikawa 2005). Evidence has been mixed (Moser and Scheiner 2012). See Ferree et al. (2017) for an application to South Africa that suggests contamination effects are insignificant in that context. See the chapter in this volume by Herron, Nemoto, and Nishikawa for an extensive discussion of these issues.

(25.) Unless otherwise indicated, election results for this section come from the website of the Electoral Commission of South Africa:

(26.) For more on the 1994 election, see Reynolds (1994) and Johnson and Schlemmer (1996). For a history of political opposition during the 1980s, see Lodge and Nasson (1991). Greenberg (2009) provides a fascinating inside account of the great difficulties the ANC faced in unifying the African electorate in the early 1990s.

(27.) An explicit effort to bring minority parties into cabinet through a rule guaranteeing any party with at least twenty (out of four hundred) seats the option of joining the government with one or more cabin portfolio, the GNU was a creature of the interim constitution. The final 1996 constitution of South Africa did not include similar provisions, and the ANC’s strong margins have allowed it to form a series of majority governments from 2009 on.

(28.) Party control over municipal councils can be calculated from data available on the webpage of the Electoral Commission of South Africa: (accessed August 29, 2016).

(29.) See Giliomee (1994) for an account of the NP’s 1994 campaign.

(31.) Booysen (2006) offers a useful history of this period.

(32.) For accounts of the IFP see Hamilton and Mare (1994), Mare (1999), Piper (2005), and Piper (2014).

(33.) Violence between supporters of the IFP and supporters of the ANC resulted in hundreds of deaths in the province over more than a decade of conflict that started before the 1994 elections.

(34.) But see the chapters in this volume by Shugart and Taagepera and by Moser, Scheiner, and Stoll for caution about the applicability of this interactive hypothesis.

(35.) See Annexure A to Schedule 6 of the constitution, Item 23A(1) in the transitional constitution, and Section 47(5) of the final constitution.

(36.) Both the DA and the ANC had strategic motivations for pursuing the constitutional change. The DA supported relaxing the prohibition on floor crossing to allow its MPs—who still sat as members of either the DP or NNP—to officially convert to the newly formed DA. The ANC agreed, gambling that it could capture some of the support abandoning the sinking NNP. See Booysen (2006).

(37.) During the 2005 and 2007 windows, the ANC increased its parliamentary presence from 69.69 in 2004 to 74.25 percent by 2008. See Booysen and Masterson (2009).

(38.) Parties are not permitted to fund business ventures, pay representatives’ salaries, or run campaigns with this money. Rather, the goal is to promote “continuous, vital links between the people and the organs of the state” through outreach, education, mobilization, and provision of information. EISA African Democracy Encyclopaedia Project, South Africa: Political Party Funding, (accessed August 19, 2016).

(39.) Also see the chapter by Moser, Scheiner, and Stoll in this volume.