The 2021 municipal elections: rise of the volatile voter or disaffected citizen?

ABSTRACT

South Africa's 2021 municipal elections present something of a paradox. The results suggested a highly competitive environment, especially in the urban metropoles, with an unprecedented number of hung councils where no single party was able secure 51 percent. The larger parties suffered losses while smaller parties made gains, suggesting voters had abandoned old political homes for new ones. In contrast, while previous local elections enjoyed modest but steady increases in participation, the 2021 municipal elections also witnessed a sharp decline in voter turnout. Less than a third of eligible voters participated. This paper argues that these paradoxical features are likely to remain a feature of South African electoral politics in the medium-term. While many who participated were likely to be volatile voters, switching their votes to a party different from previous choices, their participation was insufficient to compensate for a higher numbers of abstentions, which are a result of growing political system disaffection with electoral actors. These disaffected citizens are positioned at the periphery of politics and will be difficult to entice back to the polls, leaving the competitive aspect of South Africa's elections reliant on a diminishing number of engaged and unpredictable voters.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Final results of 2021 Municipal Elections Address by Chairperson Glen Mashinini, published on IEC website 4 November 2021: See https://www.elections.org.za/pw/News-And-Media/News-List/News/News-Article/Final-results-of-2021-Municipal-Elections-Address-by-Chairperson-Glen-Mashinini?a=AISDGvpz75ps1usOfX7oitXn±ODD68rNS2hKJXpTO8w=

2 The 2021 Afrobarometer public opinion survey was conducted five months preceding the municipal elections. This dataset is used for statistical analysis that allows us to tap the public mood before the elections and to run statistical tests for the determinants of voter turnout.

3 https://www.news24.com/news24/southafrica/news/in-numbers-this-years-elections-sees-the-most-ever-number-of-participating-candidates-20211026

4 This section uses municipal election data from the Electoral Commission website and municipal election data on registration and voter turnout, disaggregated by age, which was provided by the IEC on the author’s request. South Africa’s 2021 census estimates were provided by Statistics South Africa on the author’s request for an accurate calculation of the voting age population.

5 Turnout of the voting age population (VAP) is the standard measure for voter turnout in cross-national research. See Norris (2002, 41), Franklin (2004, 86–87).

6 Figures show both national and municipal election turnout data to show trends over successive elections. However, it is more accurate to compare every second bar, e.g. national with national and municipal with municipal election data.

7 Sources: https://municipalities.co.za/management/3/city-of-tshwane-metropolitan-municipality; https://municipalities.co.za/management/2/city-of-johannesburg-metropolitan-municipality

8 To calculate the Pedersen Index of volatility the differences between the vote-shares for all parties in two consecutive elections are added up and then divided by 2. Therefore the Pedersen score for the 2011 elections are calculated using both the 2006 and 2011 election results, for 2016 the 2011 and 2016 results are used, and so on. Party mergers are splits are taken into account for each election under investigation. While a number of mergers and splits have occurred in South African electoral politics only three directly affected these two municipalities in the 2016 and 2021 elections. In the 2016 election, the National Freedom Party (formed 2011) and Black First Land First (formed 2015) were both splinters from the Inkatha Freedom Party and Economic Freedom Fighters respectively but neither party obtained any vote shares in either municipality. However, for the 2021 elections, Action-SA was treated as a splinter party from its mother party, the Democratic Alliance and the 2021 Pedersen Index calculations for both municipalities take this movement into account. Parties that originated as splinter parties are not treated as splinter parties in the calculations if the party contested an election as a new party in any election prior to the two elections under investigation, and includes the following parties: COPE (split from ANC in 2009 but contested the 2009 elections), EFF (split from ANC in 2013 but contested the 2014 elections), GOOD (split from the DA in 2018 but contested 2019 elections). See more on calculations of the Pedersen index in Bartolini and Mair (2007, 283) and Bogaards (2008, 116). Some scholars critique the Pedersen index for its inability to distinguish electoral change between individual parties from those between groups or blocs of parties (see Bartolini and Mair 2007, 3). However, South Africa lacks clearly defined blocs of parties that represent one or another social cleavage.

9 See the 2000 elections report, 39 (file:///Users/csh/Downloads/elections%202000%20report%20(1).pdf) and 2021 elections report, p. 5 and 55 (file:///Users/csh/Downloads/2021%20Municipal%20Elections%20Report_Large.pdf) respectively.

10 https://www.news24.com/news24/southafrica/news/elections-2021-patriotic-alliance-targets-illegal-foreigners-at-manifesto-launch-20211009; https://amnesty.org.za/south-africa-years-of-impunity-for-xenophobic-crimes-driving-the-latest-attacks/; https://www.iol.co.za/news/right2know-blames-mashaba-ramaphosa-zwelithini-for-attacks-on-foreigners-31739481; https://mg.co.za/article/2019-11-07-00-mashabas-xenophobic-legacy/

11 https://www.businesslive.co.za/bd/opinion/columnists/2022-05-26-jonny-steinberg-the-post-anc-era-has-begun-and-it-has-horrors-in-store-for-us/#

12 Bartolini and Mair found that a voter turnout decline threshold of around of 6 percent should increase levels of volatility noticeably.