First in our series on the May 29, 2024, South African elections is an interview with expert Daniel de Kadt. He provides some background, highlighting the continuities and changes from previous elections and informing us about the key figures in this year’s election and how South Africa’s electoral system works. De Kadt is an assistant professor of quantitative research methods at the London School of Economics and Political Science and the author of many articles on South African electoral politics.
Stay tuned for tomorrow’s interview with Carolyn Holmes, which discusses race and campaigning, potential for electoral violence, and young South Africans’ political engagement.
Joshua Tucker: Why is this election so significant – or why are so many people paying attention to it?
Dan de Kadt: South Africans vote in two different types of elections each with its own election cycle: local or municipal elections and general or national elections. The May 29 election will be a general election. These elections determine the composition of both the National Assembly – the lower house of Parliament – and the provincial legislatures that govern the nine provinces. These provincial legislatures, in turn, determine who holds seats in the upper house of Parliament, the National Council of Provinces.
Since the first post-apartheid election in 1994 – the country’s first fully inclusive democratic election – through the most recent general election in 2019, the African National Congress (ANC) has always won a majority of seats in the National Assembly and Parliament more generally. In fact, until they won only 57.5% of the vote in 2019 they had never held fewer than 60% of the National Assembly’s seats. For 30 years the country has been what political scientists call a “dominant-party system” – institutionally robustly democratic and competitive at the local level, but hegemonic at the national level. If we believe recent polling, there is a very good chance that this will be the first time the ANC loses its majority in parliament.
Now, one thing to remember about the South African political system is that while the head of the executive (and the head of state) is called the president, this person serves in more of a prime ministerial fashion – they are not elected directly but are, along with their cabinet, elevated by the parliament and can be ousted by a motion of no confidence.
The punchline is that if the ANC loses its parliamentary majority on May 29, South Africa will enter a new phase in its political history, where there is no longer a single dominant party that can securely constitute a government on its own. While there have been coalitions of varying degrees of formality in previous parliaments, and some recent governing coalitions at the municipal level, the country has never had real coalition politics in so far as opposition parties could bring down an ANC government. Essentially, these are uncharted political waters.
Who are the main figures in this election?
Politics in South Africa is still largely party-based rather than based on individuals. So it’s mostly a cast of characters who have been part of South African politics for a very long time.
Of course, the ANC remains the most important party in the country. Ideology plays out strangely in South Africa, but in broad terms the ANC is a centrist party, in no small part because it is such a large and thus diverse party. Economically the party is center-left with a strong pro-union bent. In terms of immigration it is broadly pan-Africanist, though there is a lot of internal conflict on this issue.
Incumbent President Cyril Ramaphosa currently leads the ANC. Many South Africans hailed Ramaphosa as a reformer when he rose to power at the end of 2017/early 2018 to replace former President Jacob Zuma. In 2019, Ramaphosa rallied the urban vote even as the ANC’s rural vote softened, but has seemingly failed to get a grip on some of the country’s key problems. He also carries with him a lot of baggage, much of it connected to the role he played in the Marikana massacre of 2012, when the South African Police Service killed 34 mineworkers. Indeed, that event had profound consequences on South African party and electoral politics, costing the ANC an average of 9 percentage points in those communities closest to the massacre in the 2014 and 2016 elections, and sparking the launch of a new opposition party.
That very party now serves as one of two established opposition groups worth paying attention to.
First, the current main opposition party, the Democratic Alliance (DA), represents a center-right alternative to the ANC, advocating for privatization, looser labor protections, and stronger immigration controls. The DA has some governance bona fides, having quite successfully led the Western Cape Province for the past 15 years. The party was founded in the early 2000s as a conglomeration of multiple smaller parties, including the dying National Party. While the party has a somewhat diverse electorate, it remains disproportionately white in its leadership and party list. In no small part due to choices that led extraordinarily talented politicians like Lindiwe Mazibuko, Mbali Ntuli, and Makashule Gana to leave the party, the DA has struggled to escape the identity of a “white party.” This necessarily limits the breadth of its appeal, and the party has so far stalled at around 20 to 22% of the national vote. Sitting atop the DA’s national list is John Steenhuisen, a career politician.
Second, the Economic Freedom Fighters (EFF) is an economically radical left party, advocating for land expropriation, nationalization of the mining industry, state-led development, and pan-African open borders. The party was founded in the aftermath of the Marikana massacre, which provided a political space for ex-ANC politician Julius Sello Malema and others to form a new party. Despite its rapid rise from 2013 to 2019, many feel the EFF may have reached a plateau of around 10% of the vote. Recent polling suggests the EEF may struggle to hit or exceed that mark this year, especially with the rise of some newer alternatives.
Other establishment opposition parties such as the Inkatha Freedom Party (IFP) and Freedom Front Plus (FF+) will likely remain marginal at best because they cater to narrow segments of the population.
There are some interesting newer entrants, though the history of new small parties in South Africa suggests caution. The first is uMkhonto weSizwe (MK), a party formed in December 2023 under the leadership of former president Jacob Zuma. The Constitutional Court last week disqualified Zuma from standing as a member of parliament until 2028. The party’s actual platform is quite non-specific, but it appears to hew to Zuma’s predilection for populist, ethno-nationalist appeals. Zuma, despite his tarnished reputation, remains an incredibly popular figure in KwaZulu-Natal, a part of the country the ANC has relied on increasingly since Zuma rose to power in 2007. MK has polled particularly well in this area, but it remains to be seen whether the party’s public support will materialize as votes on election day.
Other new parties of interest include RISE Mzansi, a pro-reform, economically center-left party founded by former newspaper editor Songezo Zibi, which has recruited impressive political talent (including Makashule Gana) in its short existence; and ActionSA, a likewise pro-reform, but economically center-right party founded by businessman and former Johannesburg mayor Herman Mashaba.
How these various parties and characters might come together in a potential coalition government, should the ANC lose its majority, remains unclear. That question will likely occupy South Africans’ attention once the results roll in.
You’ve talked about how many seats different parties might receive, but can you give us some more context on how South Africa’s electoral system allocates these seats?
South Africa has three primary levels of government: national, provincial, and municipal. So in this year’s general or national election, voters are actually participating in two different but simultaneous electoral processes: one to determine the National Assembly and one to determine their specific provincial legislature (this uses a separate ballot paper). The election will be run, as are all elections in South Africa, by the highly regarded Independent Electoral Commission (IEC).
The National Assembly voting process is relatively simple, though recent changes to accommodate the presence of independent (sole) candidates have made things more complex.
For those wanting a very high-level view: South Africa uses a proportional representation system, and the 400 National Assembly seats are allocated to political parties (and, for the first time, potentially independent candidates) roughly proportional to their vote share nationwide. If Party A wins 25% of all votes across South Africa, they get (roughly) 25% of seats in the National Assembly. Members of parliament (MPs) are elected from public closed party lists that are confirmed ahead of the election, though this year there may also be a few independent MPs elected. In general, if Party A wins 10 seats, the first 10 candidates on Party A’s list would be elected. The same general principles apply to the provincial legislatures that govern each of the nine provinces, based on the separate provincial ballot.
Craving more detail on the national system? A few things changed this year! In previous general elections South Africans voted for the National Assembly using a single ballot paper, choosing just a single preferred party (so each voter got two ballot papers: one for the National Assembly and one for each voter’s provincial legislature). Once votes on the national ballot were tallied, the first step was to calculate the number of votes needed to secure a seat in the National Assembly, referred to as a “quota.” South Africa uses the Droop quota, which is equal to the total number of votes cast on the ballot divided by the number of seats available plus 1, rounded down; plus an additional 1.
Essentially, if a party secured this many votes on the national ballot they earned a seat, if they received twice the quota they received two seats, and so on. Dividing each party’s total votes on the national ballot by the quota and rounding down to the nearest whole number established the number of seats won by that party. Any seats unallocated after this first exercise were distributed by the largest remainder method, awarded based on “remaining” votes: those votes that were not “used” to satisfy a quota in the first calculation. This approach – a Droop quota and the largest remainder method on the national ballot – gave the theoretically proportional allocation of seats for each contesting party. Let’s park this theoretical quantity in our heads for a moment.
To actually allocate seats to parties and then select MPs from the lists, things got a little more complicated. The 400 National Assembly seats were divided into two “buckets.” Still using only the national ballot, 200 seats were allocated based on “regional” tallies calculated at the provincial level, while 200 were held in reserve as “compensatory seats” to maintain national proportionality. Note that “region” here just means one of the nine provinces but we use the term “region” to avoid ambiguity. For the first bucket of 200 regional seats a new quota was calculated for each region in the same fashion outlined above. This quota varied by region, given that the number of votes cast would vary by region, as would the number of seats available, based on regional population. Using these regional quotas and the largest remainder method, the first bucket of 200 regional seats were allocated to parties, and then MPs were elected accordingly from separate region-specific lists prepared by each party.
The final step was to allocate the remaining 200 seats from the second bucket, the so-called compensatory seats. The number of seats assigned to each party from the regional bucket was subtracted from the theoretically proportional allocation of seats derived from the total national tallies (that earlier concept we kept in our heads!). The difference was then allocated to the respective parties out of the remaining 200 compensatory seats, thus achieving proportionality based on the national ballot, along with regional representation. For example, if Party A was proportionally due 80 seats (having won 20% of the national ballot), but had already been assigned 42 seats regionally, it was awarded only 38 seats from the national tier, bringing its total back to 80. As complicated as this all seems, this two bucket system had limited impact from the voters’ perspective as they only had a single ballot paper, but were at least provided some nominal form of geographically specific representation.
Here’s what’s different this year
This year things have changed, making the difference between these two buckets of seats (regional and compensatory) much more salient. First, South Africans will now vote for the National Assembly using two separate ballots papers, one for their preferred party nationwide, and another for their preferred regional representative – this can be either a party or an independent candidate. This is in addition to the provincial ballot paper – so voters will now be able to cast three ballots.
Technically, these changes have some important implications. First, the quota used to derive the theoretically proportional allocation of seats will now be based on the combined total votes cast both on the national ballot and the new regional ballot, then divided by the total number of seats in the National Assembly plus 1, plus an additional 1. Second, the 200 regional seats will now be filled using only the regional ballot tallies and region-specific quotas. If an independent candidate wins multiple seats in a region, or a party wins more seats than it has candidates on its regional list, those excess seats must be forfeit and returned to the regional pool. The region-specific quota will then be recalculated, ignoring the votes cast for that individual or party, and ignoring the seats already allocated to them, and so on, until all 200 seats are allocated. Once these 200 seats are allocated, the remaining 200 compensatory seats will be allocated by subtracting the regionally allocated seats from the theoretically proportional allocation of seats, with the intent of again achieving proportionality to the sum of national and regional ballots won by each party.
The separation of the two ballots raises the interesting possibility of split-ticket voting, where voters may diversify their choices by supporting local candidates and smaller parties on the regional ballot, while still supporting one of the three major parties (the ANC, DA, or EFF) on their national ballot.
The provincial legislatures essentially follow the same process as above, but without the added complexity of the two buckets. That is, a quota is established in the same fashion, seats are allocated, the quota is recalculated if any seats have been forfeit, remaining seats are reallocated based on the new quota, and then the unassigned seats are assigned using the largest remainder method. Provincial legislatures then nominate representatives to the upper house, the National Council of Provinces, proportional to the party composition of the provincial legislature.
Once Parliament – both the National Assembly and the National Council of Provinces – are constituted, it’s time to select the executive. And that’s where things will get very interesting this year.
Why, despite sustained and high levels of unemployment, three decades of dismal economic performance, publicly exposed corruption on a grand scale, regular interruptions in the electricity and water supply, and so on, is the ANC’s support still so robust? To outsiders, it seems that there is no shortage of compelling explanations for why the ANC is losing support; the question is really why they aren’t losing even more?
I’m going to throw out three major factors – primary cleavages in South African politics.
The first major factor is the intersection between political geography, party capacity, and party history. South Africa is a reasonably big country, and one of the (numerous) legacies of apartheid is that many parts of the country are simultaneously very hard to physically penetrate and, because of apartheid planning and forced resettlement policies, remain very populous. So there are a lot of votes to be won in hard-to-reach communities. The ANC, more than any other party in the country, has the physical and financial capacity to put operatives in the farthest reaches of the country. Simply put, the ANC’s political machine is strong, well-funded (though it undermined itself in this regard with the recent Political Party Funding Act, which it amended this month), and technically robust. Add to this the fact that the ANC has built a generous welfare state and has a strong political identity grounded in anti-apartheid struggle narratives to which many, especially older and thus more politically active, South Africans remain deeply attached.
A second and connected major contour of the past 15 years has been the ANC’s strong ruralization of its base, especially into KwaZulu-Natal, which happened under the leadership of President Jacob Zuma. For those interested in further reading, Horacio Larreguy and I trace and quantify this shift, while Lungisile Ntsebenza’s book offers a magisterial analysis of these politics.
During this 2007 to 2014 period the ANC gradually lost support in urban South Africa and began to rely much more significantly on rural voters in KwaZulu-Natal, as well as its more established rural base in the Eastern Cape and other parts of the country. This was a viable electoral strategy precisely because of the country’s electoral system; the system ensures proportionality and so it does not matter where votes come from.
Ramaphosa did a little to win back some urban voters in 2019, but with covid-19, the increasingly severe electricity and water crises, and the newly passed National Health Insurance Act affecting urban voters very directly, it is quite possible that urban South Africa will look for alternatives this year. If the new MK party also severs the ANC’s connection to rural KwaZulu-Natal, the ANC could be in real trouble.
One final factor that I would emphasize, in part because it is something I have been thinking about for over a decade, is the role of generational cohort replacement in South African politics. Young South Africans are quite extraordinarily unlikely to participate electorally. This has been the case throughout the country’s democratic history, and there is little evidence this will change in 2024.
Older voters are already registered at extremely high rates. From a registration perspective, that means parties have to rely on younger voters to boost their vote share. In a 2013 analysis the IEC estimated that between 91% and 100% of all eligible voters aged 50 and over were registered to vote, while only 52% of voters aged 20 to 29 were registered, and only 10% of 18 to 19-year-olds.
Indeed, the growth of the overall electorate has slowed since the 2014-2019 cycle. While there are many newly registered voters who are, mechanically, mostly young people, the reality is that young voters remain stubbornly unlikely to actually vote on election day whether they are registered or not.
This trend is likely to be compounded by the fact that, for the first time, South Africans must vote at their local voting station unless they acquire an exemption from the IEC. Given that young people are more likely to be mobile than older voters, this suggests another barrier for the youth vote.
So part of the ANC’s slow decline is slow generational replacement – younger voters are slow to displace older voters, simply because young voters are not voting at high enough rates.