ROGER SOUTHALL: Changes to governing party are rare in SA

There were 20 general elections in SA between 1910 and 1994, albeit held under a racially restricted franchise, and just two changes in governing party as a result. On the first occasion the South African Party lost to a National Party-Labour Party “pact” in 1924. On the second occasion, the United Party lost to the National Party (NP) in 1948, Jan Smuts being the leader of the losing party both times. 

The NP had been ousted from power only in late 1939 by the decision of the governor-general, who under the Act of Union retained the right to appoint and dismiss governments. After prime minister JBM Hertzog lost a vote in parliament on a motion that would have kept SA out of the World War 2, the governor denied him the right to call an election, which he might have won.

Instead, he called on Smuts and the United Party, which had won the war vote in parliament, to form the next government without requiring it to seek the approval of the electorate. Though Smuts won a sweeping victory in an election in 1943, he met defeat at the hands of DF Malan’s reunited NP five years later.

The NP subsequently used various types of electoral trickery to shore up its dominance in parliament. However, this was eventually extinguished by the arrival of democracy in 1994. This saw the ANC sweep into power and subsequently win the next four elections. Only in the last of these, in 2019, was there any serious talk of the ANC possibly losing its majority, though, when it came to the crunch, it was returned to office with more than 57% of the vote.

This time round, on May 29, it really does seem that the ANC’s dominance of the electoral arena is at serious risk. Uncertainty prevails, but polling suggests the governing party’s vote will fall below 50%. If this comes to pass, it will be humiliating for the ANC, yet there is next to no chance that an adverse election result will result in the party being forced from office.

Unless something totally unpredictable happens, it will remain the largest party. In this case, though it will be compelled to turn to one or more opposition parties to form a coalition to secure a majority in the National Assembly, it will remain the dominant party in government and call all the important shots. ANC governance, even if somewhat diluted, is destined to continue after the 2024 election.

Compare this electoral history with that of Britain, where there have been 15 changes in the governing party in about 30 elections between 1906 and 2019 (and another one likely to come about in the election later in 2024). Alternation in power in Britain has been facilitated by its first-past-the-post electoral system, which allows for disproportional outcomes. Candidates need only to secure more votes than their nearest rival to win the election in their constituency, meaning that many of those who win do so by winning a plurality — not a majority — of the vote.

This quite easily translates into a party winning a majority of seats in parliament despite having secured only a minority of the overall popular vote. Indeed, it is rare for governments to take office having won 50% or more of the popular vote in Britain, this having occurred only four times since 1906 (in 1918, 1931, 1935 and 2010, the first two of these being won by cross-party coalitions rather than, as on the two later occasions, by the Conservative Party).

SA inherited the first-past-the-post electoral system from Britain at union in 1910, adding the notorious twist that the franchise was restricted to whites (and initially only white males, bar a small number of qualified black voters in the Cape). However, there was another feature of the system that explained the infrequency of party turnover in government. This was the phenomenon of “loading”, whereby constituencies in urban areas were allocated more voters than those in rural areas.

Because its supporters predominated in rural areas, this gave the NP an inbuilt advantage. When it defeated Smuts’ South African Party in 1924, it had only 35% of the vote but 63 seats, to the latter’s 47% (but only 53 seats), securing its majority in parliament by forming a coalition with the Labour Party as its junior partner. In 1948, Smuts lost the election having won more than 49% of the vote, but again significantly fewer seats than the NP (which had taken just 38% of the vote). Even then, the NP again needed to form a coalition, this time with the small Afrikaner Party, to take power.

Unsurprisingly, given this history, the negotiations that led to democracy abandoned first-past-the-post as allowing for popularly unrepresentative outcomes and adopted a list system of proportional representation in its place. Unlike the predecessor electoral system, this weighted the votes of voters equally, so when the ANC won the 1994 elections with nearly 63% of the vote, it won an equivalent percentage of seats in parliament. 

Debate continues over how important a factor race is in determining the outcome of SA elections, yet few doubt that it remains significant. As the party of liberation with a history going back to 1912, the ANC has long enjoyed majority support among black Africans, who today constitute about 80% of the population. It has had a built-in advantage when contesting elections since 1994. However, factors other than race also influence how South Africans vote. Consequently, because voters have become increasingly critical of the ANC’s performance in power, its share of the vote has been falling steadily since it won about 69% of the votes cast in 2004.

Polling now suggests that the ANC stands on the edge of a precipice, likely to win less than 50% of the vote. Yet so long as it wins more than 40%, it will remain dominant in any coalition it has to form to remain in power. Only when faced in an election by a bloc of opposition parties capable of mustering a similar 40% or so of the vote, and hence equally capable of forming a winning coalition, will it stand in realistic danger of losing office. The multiparty charter bloc of parties aspires to this in the forthcoming election, yet all the indications are that it will fall far short of its ambition.

The 2024 election may well see the ANC having to form a coalition. Yet, realistically, no change of government is in sight, and ANC governance will continue for the foreseeable future — unless for some reason the ANC were to split. Only then are we likely to see that rarity in SA: a genuinely competitive general election.

• Southall is sociology professor emeritus at Wits University.