Addressing the Instability Arising from Hung Legislatures in South Africa

Jun 6, 2023

Hung Legislatures in South Africa

The municipal electoral system

The municipal electoral system has had five municipal elections, all of which have been declared as free, fair, and safe with almost no disruption of the elections. The elections have all been keenly contested, and in the 2021 elections, some 324 parties were involved, including also some 1582 independent candidates contesting wards, with 51 winning a seat, out of the over 9000 councilors elected from parties. The 2021 elections were conducted under exceptionally difficult circumstances due to the COVID-19 pandemic. Across the country, the major parties, ANC, DA, EFF, and IFP, dominated the electoral results.

At the same time, the increasing number of “hung” municipalities from 24 in 2000 to 81 in 2021 has compounded instability.

The high number of hung councils has highlighted the need to strengthen the guidance on the formation and management of coalition councils. If there is persistent instability in such municipalities, provision is made in law for the dissolution of such councils and the holding of fresh elections, although to date this has only happened in a few instances, although COGTA and National Treasury are actively working on a number of such cases.

SALGA – a strong regulatory framework

SALGA has argued strongly for the adoption of a regulatory framework to guide coalitions and address concerns, to improve political stability, ensure oversight and accountability, involve communities, and address the allocation of municipal political portfolios.  SALGA also argue that coalition talks should not discuss issues relating to staff appointments and procurement decisions.  The challenge with this, though, is that such frameworks may end up being worthless as they would be dependent on parties committing themselves to acting in good faith and abiding by such processes.

Two areas for electoral reform

There are two areas in which it is possible to have electoral reform which, based on international evidence, would assist in making hung legislatures and executives more stable. The first is to remove the option of Mayoral Executive in municipalities and only allow Collective Executives.  This could be done very easily and would reduce the focus of coalition agreements to be on positions such as Mayors, Speakers and Chief Whips.  The representation from the larger parties on Executive Committees would remain irrespective of changes in the office bearer positions.

Collective Executive Systems vs Mayoral Executive Systems

Currently, just over 50% of the 257 municipalities have collective executive systems and the remainder have mayoral executive systems. In “executive mayor” systems, executive political authority and the bulk of decision-making power vests in a small number of councillors (the executive mayor and the MMCs, who tend to share party-political/factional/personal loyalties).

Research shows that, while they are in principle accountable to councils, in practice (executive) mayoral committees tend to take decisions behind closed doors and merely report these to council, while answering only to political party leadership structures (located outside of democratic local government).  In addition, and even more worrying, is that “executive” Mayors often act as if they are heads of administrations and accounting officers which are areas that by law are the preserve of municipal managers.  Tensions often arise as this political-administrative interface gets broached resulting in poor governance and financial administration.

Impactful change

1. Executive Committee Systems

However, Executive Committee systems would mean that all major parties could be included in the Executive Committee, irrespective of who the Mayor is, and they would remain in place even if Mayors or Speakers change.

2. Introduce Electoral Thresholds

A second and possibly more impactful change could be to introduce electoral thresholds.  The way thresholds work, for example, is that unless a party gets a certain percentage of the vote they cannot be allocated any PR seats. Some modifications exist such as where parties have won a ward/constituency seat, the threshold does not apply.  Currently, well over 50 countries that have proportional election systems have introduced election thresholds. These generally range between 3-5%.

Work We have Done – City Insight

Examining International Contexts

Work we have done, including examining international contexts, shows that such thresholds do work, even though they do reduce the number of parties elected to legislatures. Some may say reducing parties has negative consequences as having many parties represented in councils could be seen as creating a more inclusive political system. However, it can also be argued that having many parties can create instability if coalitions are formed based on the dictates of what very small parties require.

“One Person, One Vote, One Value”

Our argument, though, is that we should not focus on such subjective issues.  Rather, and more importantly, democracy requires the implementation of the principle of “One Person, One Vote, One Value”. Unfortunately, on the score of “One Value” our electoral system fails us.  However, our research shows that introducing a 1% threshold would largely address these distortions ensuring values for votes are far more equal across parties.

Currently, after the national elections, for example, two parties which have one seat each in Parliament, have average votes per seat which are 34% less than parties which have 5 or more seats in Parliament.  And parties with 2-5 seats in Parliament were elected to the seats based on around 31% fewer votes than the larger parties received for their seats.  This means that smaller parties currently require less votes to get a seat in Parliament compared to the bigger parties.

A similar situation pertains in local government as the following examples show:

  • In Johannesburg, the 8 parties with one seat each received on average 39% fewer votes per seat than parties with 5 or more seats;
  • In eThekwini, the 15 parties with one seat each received 23% fewer votes per seat than the parties with 5 or more seats;
  • In Cape Town, the 6 parties with one seat each received on average 22% fewer votes per seat than the parties with 5 or more seats.

Electoral System currently favours Smaller Parties

Clearly, our electoral system of calculating seats favours small parties and results in these very serious distortions. Introducing a threshold for the number of votes required could address these distortions.

We believe that a threshold of only 1%, if challenged constitutionally, would pass the constitutional principle that our electoral system must result in general in proportional representation.  The word “in general” is important because our electoral systems are not strictly proportional.  In the last national elections, for example, 34 contesting parties (around 2% of the vote) failed to get seats.  These votes are often referred to as “excess” or “wasted” votes because they do not result in a seat. Introducing a 1% threshold would possibly double the excess votes from 2% to just over 4%, but it would reduce significantly the small party seats which have been won through fewer votes compared to bigger party seats.  Many of these small parties become “kingmakers” who are able to exert pressure on larger parties disproportionate to their voter support.  Where their support changes it can result in significant political instability.