The Battle for the Bay - What happened in the Nelson Mandela Bay Metro in the 2016 Local Government Elections?

Sithembile Mbete, Political Studies; University of Pretoria | 30 August 2016


Nelson Mandela Bay was expected to be one of the most highly contested metropolitan municipalities in the 3 August 2016 local government elections. Before the election ANC had a precarious hold on power in the municipality with a small majority. They held only 62, out of the total of 120 seats. In 2011 local government election the ANC won 52.13 percent of the vote, as compared to 66.85 percent in 2006. The DA won 40.2 per cent in 2011 compared to 25 per cent in 2006. In 2011, the DA pinned its hopes on winning control of the Metro in coalition with COPE, but COPE's decline and literal implosion put paid to those hopes. In the run up to 2016, the DA's projected goal was to win the municipality outright, by capitalising on both - the implosion of COPE as a viable party, and on the deep factionalism within the ANC.

As it came to pass, neither the DA nor the ANC succeeded in winning over 50% of the vote. The DA won 46.71% and the ANC 40.92%. However, in terms of ward seats the ANC won control of more wards than the DA. This difference between performance at ward level and overall share of the vote - prompted some ANC leaders to complain about the electoral system being biased towards the opposition.

Is the municipal electoral system somehow rigged against the ANC? How can we explain the discrepancy between the ANC's ward performance and its total number of seats? Did the DA succeed in 2016 merely by consolidating the minority vote or did it manage to make in-roads in traditional ANC strongholds?

In order to answer these questions we looked specifically at traditional ANC strongholds: Gqebera (Walmer), New Brighton, Kwazakhele, Zwide, Kwanobuhle (Uitenhague) and Motherwell and compared them to two DA strongholds: Summerstrand and Walmer/Newton Park.

Ward breakdown:

Gqebera/Walmer- Ward 4
New Brighton- Wards 14,15,17,18
KwaZakhele- Wards 16, 19,20,21,22
Zwide- Wards 24,25,26,27,28
Motherwell- Wards 23, 54,55,56,57,58,59
Kwanobuuhle- Wards 42,43,44,45,46, 47
Summerstrand- Wards 1 and 2
Walmer/Newton Park: Wards 3, 6,7

Voter Turnout

Voter Turnout (2011 compared to 2016)

It has frequently been said that declining voter turnout and/or a voter stay away impacts negatively on the ANC's performance in an election. In other words, the ANC receives lower percentage support when voters choose not to vote, whether out of apathy or disgruntlement with the party. In a previous article, Potgieter, Berkowitz and Fakir (2016) found that in Nelson Mandela Bay, the correlation between voter turnout and voter support was stronger in national and provincial elections than in local government elections. When the results were disaggregated by ward, the authors found that there was a negative relationship between ANC support and voter turnout and a positive relationship between DA support and voter turnout. However these relationships were weak and were not consistent across all wards.

In all the areas under consideration voter turnout declined between 2011 and 2016, except for Gqebera, the only ANC support base in which turnout increased somewhat. This drop in turnout can be explained by the factional politics in the Tripartite Alliance in general and within the ANC in particular. Kwanobuhle is a township in Uitenhague, the centre of the auto industry in Nelson Mandela Bay and traditional support base for the National Union of Metalworkers of South Africa (NUMSA), that was expelled from COSATU in 2015. The effect of NUMSA's expulsion on the election results in the area will be explained below.

While there was a noticeable drop in turnout in ANC strongholds, we will see below that this drop is not significant enough, to explain the fall in the ANC's support in the city.

Voter Turnout (2011 compared to 2016)

Turnout in the traditional DA strongholds increased significantly from 2011 to 2016 to well above the average voter turnout in Nelson Mandela Bay as a whole. Voters in Summerstrand, whose turnout was below the municipal average turnout in 2011, spiked in 2016 and it increased in Walmer/Newton Park as well. It may thus, be that much of the increase in support for the DA in 2016 in Nelson Mandela Bay, was driven by the increase in turnout among its traditional constituencies.

Party Support Patterns

Party support

We can assess the ANC's decline in its strongholds by looking at the electoral contest in more detail. The ANC's support in these areas decreased significantly from 2011 to 2016. The smallest drop was in Gqebera, though in spite of the drop the ANC's percentage support remained above 80%. In New Brighton, Zwide and Kwanobuhle the ANC's support dropped by over 17% giving it below 70% of the vote in New Brighton and Zwide and below 60% in Kwanobuhle. This represents a significant decline in support in areas that have a long culture of ANC activism both before and after apartheid. The ANC experienced a smaller decline in Motherwell where the party's support fell by -12,7%.

The decline in Kwanobuhle is particularly noteworthy when viewed in the context of the expulsion of NUMSA from COSATU and the impact this had on the Tripartite Alliance. This is the first election since 1994 in which the ANC has not had access to COSATU's organizational network for electoral campaigning. One of the effects of the alliance tensions in Kwanobuhle was voters staying away from the polls as seen in the 8,65% decline in voter turnout between 2011 and 2016. However this is not the only explanation for the fall in ANC support. It appears that the EFF was the main beneficiary of the ANC's losses as it won 11.2%, a good showing for its first local government election. Another beneficiary was the United Front of the Eastern Cape (UFEC), a party linked to NUMSA. It won 5,66% of the vote. The DA's had a 5,03% gain between 2011 and 2016. This likely came from COPE, which had a 7,75% loss. This indicates that COPE voters did not go back to the ANC but rather voted for another opposition party or stayed away.

The DA gained support in all six areas under consideration. It's most significant gains were in New Brighton and Zwide where it's percentage of the vote went up by 11,88% and 10,16% respectively. Most of this gain can be explained by the implosion of COPE, which fell by 7,38% and 6,73% in each area. However, this does not account for 4,5% and 3,43% that the party won in New Brighton and 3,43% in Zwide respectively. We can presume that this means the DA won those votes from the ANC, which means the party is beginning to make some inroads, not only among black voters but also into the ANC's traditional support base in Nelson Mandela Bay. Nevertheless, the ANC won the wards in all of these areas showing that the DA has some way to go before it can shift significant proportions of the ANC's electoral support its way.

Party support

When we examine party support in two traditional DA strongholds, we see that the DA received a massive proportion of the vote in these areas thus boosting its overall share of the vote in the metro. In both Summerstrand and Walmer/Newton Park the DA received over 85% of the vote in 2016. This is similar to the support it received in 2011. However given the much higher turnout rate in these areas, the DA's share of the vote was augmented compared to the ANC's relatively weaker performance in its strongholds.

Seat allocation

2016 Seat allocation

The ANC won the most number of wards in Nelson Mandela Bay- 36 out of 60. As demonstrated above, even if the party's percentage support declined in its traditional support areas, it still won enough votes to win those wards outright. However the drop in voter turnout and the decline in the ANC's proportion of the vote in those areas had an effect on the party's share of the vote in the municipality as a whole.

The DA received 23 ward seats, 13 less than the ANC. However its performance in its traditional strongholds enabled it to retain those wards with next to no competition from other parties. In addition, voters in suburban areas turned out at significantly higher rates than those in townships. This raised the DA's proportion of the vote enabling it to win the larger share of PR (party) seats, 57 compared to the ANC's 50. This provided the boost the party needed to form a minority government in coalition with other smaller parties.


The municipal elections system is not rigged against the ANC but rather aims to balance direct representation in the ward system with the principles of proportional representation on which South Africa's whole electoral system is founded. There may be a case to be made for discounting votes cast in a ward towards being counted in the overall tally of a parties PR seats, but this would require a minor amendment to the electoral system in which ONLY the PR ballots cast would count towards a parties PR seat allocation. But this would disadvantage smaller parties and go against the grain of the very good rationale (for fair proportional test of support for a party, inclusivity and diversity) for why the system was adopted. In any event - the extant system worked to the benefit of the ANC until this local government election in 2016. This system as currently configured is intended to ensure the representation of all people in a diverse society.

In previous elections the ANC could rely on high levels of support among the electorate to control both the majority of wards and to win the highest proportion of party seats. However since the split from COPE in 2011 the party has been in steady decline in Nelson Mandela Bay. The 2016 election is an indicator of how far the proportion of ANC support has fallen in the metro, relative to opposition parties. This has been exacerbated by the intra-alliance conflict that resulted in NUMSA's breakaway from COSATU. This corroborates the contention made by Brunette, Kotzen, Rampedi and Mukwedeya's (2016) in a pre-election report that the battle for Nelson Mandela Bay was in fact fought within the ANC. Factionalism represents a far greater threat to the party than the steady encroachment of opposition parties, who have benefited disproportionately from the ANC's internal battles.

As was witnessed in other metros, the DA consolidated its support in its traditional support bases, where it succeeded in getting voters out to the polls in very large numbers. If the DA wants to eventually win control of the Nelson Mandela Bay Metro outright, without the need for a coalition, the DA evidently need to do more to attract voters in the townships. It has made a good start in terms of increasing its percentage support in all the township areas on a steady, but sustained basis. Results in the townships demonstrates that voters are quite sophisticated - and have been willing to give their vote to a party other than the ANC. Whether the DA can capitalise on this depends on how well it proves itself in Government in the Nelson Mandela Bay Metro.