Polls predict that the ANC will finish below 50% in the May 2024 general election. Elections are a contest for power in general and control over legislative and executive authority in particular. The only criterion for success in this struggle is the percentage a party wins. Percentages are important in this respect, but the absolute number of votes cast in favour of various parties is also an important indicator of long-term trends in political support and participation.

The 2024 election may well represent a break with these past trends, but understanding the past is important and may give a good indication of the direction we are heading.

Technical note: The graphs are based on IEC and StatsSA data from four general elections. I have not included the 1994 election in the analysis because it was, in so many ways, a special and unique election. It was politically unique, constituting a founding political moment that is rare in history, but also organised on a different basis from subsequent elections (e.g. there was no requirement for voters to be registered). All the figures below reflect national-to-national votes. While there is some variation in provincial votes, this is generally negligible. When I refer to “population”, I mean the population over the age of 20 years as reported by StatsSA at the end of the year.

In percentage terms, “peak ANC” was reached in 2004 when the governing party reached a two-thirds majority. In successive national elections, the opposition has been catching up, bringing the ANC steadily down towards the 50% mark.

In terms of absolute votes cast, the ANC surged in KZN during Zuma’s 2009 election. This factor may also have led to an increase in ANC votes in Gauteng and Mpumalanga, which are home to a large number of people with roots in KZN. (This Gauteng surge may partly reflect the emergence of COPE, which I discuss below).

In the rest of the country, turnout in favour of the ANC peaked in 2004, and the decline in support has been accelerating, as Gauteng and Mpumalanga reverted to decline in 2014. The combination of significant threats to its KZN base and declining effective turnout in the rest of the country could be fatal for the ANC’s majority.

If the ANC’s support in KZN falls back to pre-Zuma levels, it will have 750 000 fewer votes in that province, a major blow to the ANC nationally.

The national impact of the KZN surge can be seen here by combining all ANC votes excluding that province. This shows that, notwithstanding complementary trends in Gauteng and Mpumalanga, the ANC’s absolute vote outside KZN peaked in 2004. The decline in national turnout in the ANC's favour has been accelerating since then.

Opposition votes increased substantially in absolute terms between 2004 and 2014, but the increase slowed into 2019. We come back to the composition of this opposition support base below.

While the opposition has gained in percentage terms, this does not reflect an increase in support amongst the whole electorate. In the last two elections, the opposition vote barely kept pace with the increase in registered voters ….

… and opposition support fell as a share of the voting population in the 2019 election.

Of course, the ANC’s support as a share of registered votes or as a share of the eligible population has fallen even faster. The result is that active support for the ANC (in the form of voting) has been falling faster than the electoral outcomes suggest (i.e., as indicated by the percentage of votes cast). Less than one-third of the eligible electorate voted ANC in 2019.

In this graph, I have grouped parties into political blocs to give a different view of the trends. This should give us an indication of the extent to which the trends are within broadly defined political traditions or between these traditions.

ANC+ includes all parties led by people who were ANC members in the past or who were part of the liberation movement or the worker's and socialist movement.

It seems that, in addition to the surge in KZN for the ANC, the entry of COPE into the 2009 election boosted the total vote for this bloc. If we exclude KZN from the numbers, we can still see a surge in support for ANC+ in 2009. The combed vote of the ANC and COPE in 2009 was 10.7 million outside KZN, higher than the vote the ANC had received in the 2004 election. In other words, the creation of COPE (which garnered 1.3m votes in 2009) probably had the effect of mobilising new voters into the 2009 election.

On the other hand, the EFF’s debut in 2014 did NOT have such a pronounced effect of drawing new voters out on election day. The combined vote of the ANC+ block fell in that election. In absolute terms, however, the EFF increased its vote by 60 per cent in the 2019 election (from 1.2m to 1.9m), which certainly contributed to the ANC’s decline from 11.4m voters in 2014 to just 10m in 2019. The second factor explaining the whopping 1.4m drop in ANC turnout in 2019 was the reversal of its support in KZN, as is evident in Figure 2.

The DA+ block includes parties that were part of the apartheid state (e.g. NNP), white liberal parties that participated in minority rule, and black collaborators with the tricameral system. While the opposition support grew in 2019, this was mainly due to the expansion of parties outside this bloc (remembering that ‘opposition votes’ in the graph is the total of all opposed to the ANC). In absolute terms, the DA+ bloc appears to have hit its ceiling in 2014, and the increase in opposition votes in 2019 came from outside this grouping.

The IFP+ reflects votes for the IFP, NADECO, and NFP, a combination that has consistently declined in every election.

Finally, it’s important to note that fewer and fewer South Africans are registered to vote….… and of those who are registered, fewer and fewer have voted. Shifts in electoral outcomes, and party control of government, reflect changing allegiances, the capacity of parties to mobilise their supporters to elections. All parties are increasingly failing to do so, and this falling participation in the institutions of political representation weakens the foundations of our constitutional democracy.

Written by Michael Sachs, Adjunct Professor at the Southern Centre for Inequality Studies at the University of the Witwatersrand