As a political party that repeatedly but democratically wins national elections, a dominant party often controls policy making, political debate and media coverage. Since a dominant party has the ability to control both substance and framing, its electoral defeat marks a critical juncture in previously stable and predictable party systems. Assumptions of stability, while usually justified, nevertheless lay the groundwork for a puzzle whenever a dominant party does lose a national election. Such events were watershed, era-defining moments in places such as India, Italy and Japan. What is responsible for the decline of dominant parties?
With a few exceptions, most previous works largely treat opposition parties as powerless. Many argue that access to state resources gives the ruling party ‘hyper-incumbency advantages’ in controlling spending and messaging, maintaining clientelistic networks and recruiting better candidates (Greene Reference Greene2002: 759–60; Scheiner Reference Scheiner2006). The end of dominance comes from exogenous shocks such as economic crises or from mismanagement of internal party factionalism.
In contrast, I focus explicitly on how opposition parties’ strategies contribute to or erode one-party dominance. Specifically, I look at South Africa, where the ruling party's strategy certainly contributes to its continued dominance (Ferree Reference Ferree2011). However, the strategic choices of the opposition also matter. I show that the opposition has discovered, through a process of trial and error, strategies to gain power at the subnational level. Although these have not so far resulted in national victories, they do represent a weakening of the dominant party and reveal that the opposition is not simply a passive actor. Thus any eventual end of dominance in South Africa is likely to reflect not only exogenous shocks or ruling party missteps, but also the strategic moves of the opposition. Specifically, a party can gain by winning important subnational offices and then creating a governance record that it can use in winning new supporters. Thus, this article also demonstrates the potential value of studying the impact of subnational electoral gains and losses on national politics (see also Gibson and Suarez-Cao Reference Gibson and Suarez-Cao2010; Snyder Reference Snyder2001; Trounstine Reference Trounstine2009).
Dominant Party System Evolution and South African Electoral Politics
In a dominant party system, one political party continually and often overwhelmingly wins control of the national government (Duverger Reference Duverger1954). Sartori (Reference Sartori2005: 173, 205) included ‘predominant’ parties in his classification of party systems; they are distinct from hegemonic party systems and other forms of competitive authoritarianism in which one political party manipulates the political landscape in undemocratic ways while permitting opposition parties to contest elections. Dominant party systems meet minimalist criteria for democracy.Footnote 1
Most general theories of party system change cannot account for the evolution of dominant party systems into more competitive systems (Greene Reference Greene2007: 17–27). In contrast, those who focus on the phenomenon propose several hypotheses. Pempel (Reference Pempel1990: 16, 341, 346) concluded that dominance is identified by the ‘virtuous cycles’ of maintenance by ‘material or symbolic means’. The end of dominance, then, is the failure of these self-reinforcing cycles. However, the key causal factors of dominant party system change are unclear. Most analyses focus on the dominant party's internal factionalism and subsequent failure to maintain broad coalitions. Dominant parties often contain very broad coalitions, drawing together disparate interests through a combination of national unity and the adept use of state resources. Dominant parties are often the successors to independence or democratization movements; they tend to lose their appeals to unity as generational shifts cause fewer voters to give them credit for liberation. Intraparty coalitions are then maintained through clientelism and careful management of access to power. Systemic change comes from changes in resources; that is, either the dominant party loses resources – because of, say, an economic crisis – or an opposition party gains resources independent of its own actions – because of, for example, economic shifts increasing its base of middle-class voters (Bruhn Reference Bruhn1997: 57; Kohli Reference Kohli1998).
However, as Ferree (Reference Ferree2011: 10–16) argues, these studies may overstate the role of clientelism in dominant party systems. Some literature does recognize the opposition parties’ success in building grassroots organizations and co-opting emerging electoral interests in certain localities (see, for example, Brass Reference Brass1990; Bruhn Reference Bruhn1997; Zuckerman Reference Zuckerman1979), but the central focus of the literature remains on factors other than opposition strategies aimed at national electoral growth. The purpose of this article is to illuminate in detail the strategies employed by South African opposition parties.
This article focuses on South Africa as a paradigmatic dominant party system. The African National Congress (ANC) has won more than 60 per cent of the vote nationally in each of the four post-apartheid national elections. Many attribute South Africa's one-party dominance, similar to other dominant party systems, to the success of a broad coalition of interests facing weak and fragmented opposition parties. The African National Congress has built an adept electoral combination of business elites, the national union organization and much of both the urban and rural poor who credit the African National Congress with ending apartheid (Butler Reference Butler2008; Seekings Reference Seekings2005a). Conversely, opposition parties are regularly portrayed as niche parties with demographically based appeals too narrow to grow, or they are dismissed as limited vehicles for charismatic personalities and their localized clientelistic networks (Heÿn Reference Heÿn2009; Naidu and Manqele Reference Naidu and Manqele2005; Petlane Reference Petlane2009).
Also similar to other dominant party systems, the African National Congress appears to benefit from continued legitimacy stemming from its role at the forefront of the struggle against apartheid. As a result, South Africa's election results are often understood as a function of demographics (Friedmann Reference Friedman2005; Johnson and Schlemmer Reference Johnson and Schlemmer1996). That is, they are seen as constituting an example of a ‘racial census’ in which voting is an expression of identity rather than policy preferences (Horowitz Reference Horowitz1985). Hence, when a South African politician is asked why Cape Town elections are consistently more competitive than those in other cities, he or she will invariably attribute the difference to the city having a larger proportion of white and coloured voters than other cities. However, a growing body of work convincingly argues that this conventional wisdom is wrong. Race coincides with class, occupation and religion, so it is not surprising that voters from the same race prefer similar parties (Mattes Reference Mattes2004). Additionally, voters may use the perceived racial identities of political parties as a heuristic to understand the likely beneficiaries of parties’ policies (Chandra Reference Chandra2004; Posner Reference Posner2005). Voting patterns hide what survey research shows: nearly half of African voters are independents, not strong partisans of the ‘black parties’ they tend to support in elections (Schultz-Herzenberg Reference Schulz-Herzenberg2009: 30). Furthermore, turnout has declined even as the African National Congress's share of the national vote increased. That is, African National Congress supporters continue to vote but there are fewer of them.
The African National Congress has benefited from facing typically weak and fragmented opposition forces, the ‘successful realization of a political strategy employed by the ruling party to discredit and delegitimize the opposition’ wherein the African National Congress frames opposition parties in the minds of voters as non-credible alternatives (Ferree Reference Ferree2011: 2).Footnote 2 However, in future, the strategies of multiple political parties will determine the competitiveness of South African politics. Indeed, Ferree (Reference Ferree2011: 2) suggests that opposition parties could improve their image and expand their electoral support through ‘winning elections at the local level and developing a reputation for even-handed delivery of services or through alliances with legitimately multi-racial parties’. It is precisely such a dynamic that has been on display in Cape Town in recent years and that I discuss here. To that end, the article traces events in Cape Town and the surrounding Western Cape province between 1999 and 2010. Cape Town presents a challenge to the overall picture of dominance, suggesting areas of vulnerability for the African National Congress and demonstrating how opposition parties might successfully gain a share of political power.
South Africa's Party System(S), 1999–2009
In the lead-up to the 2009 national election, South Africans witnessed the first major post-apartheid split in the African National Congress, culminating in the formation of the breakaway Congress of the People (Cope). Although the African National Congress remained dominant, many observers claimed that the first chinks in its electoral armour had emerged. Media coverage of the 2009 national election depicted the creation of the Congress of the People as being hugely significant because of its potential broad base among black people, constituting the first ‘real’ challenge to the African National Congress. Although the 2009 election is noteworthy, these characterizations ignore the increasing success of longer-established opposition parties, starting from about 2004. As will be shown, opposition strategies have shifted significantly from those employed in the late 1990s. Much as Ferree (Reference Ferree2011) argues that the South African party system is attributable to the African National Congress's political strategies, not just to unconscious identity voting, I argue that opposition parties’ strategies are also responsible for the party system and, particularly, for changes in it over time.
By the second post-apartheid national election, in 1999, the African National Congress had clearly established its dominance. Most opposition parties had declared the transition period over and the advent of ‘normal’ democratic politics in which they would challenge the ruling party's policies as any legitimate opposition would. Thus began the evolution of opposition party strategies to challenge a dominant party.
Dysfunctional Party Mergers: The Creation of the Democratic Alliance, 2000
‘Normal politics’ began with efforts to consolidate opposition bases rather than to expand their support. Facing elections against those whom much of the country considered liberation heroes, the largest opposition parties publicly acknowledged that they were not going to defeat the African National Congress in the foreseeable future. The so-called ‘white parties’ remained fragmented in their pre-transition configurations; the parties’ leaders decided to combine as many non-African National Congress votes as possible under one electoral banner. Hence the New National Party (which had won 6.9 per cent of the national vote in 1999), the Democratic Party (9.6 per cent) and the Federal Alliance (0.5 per cent) became the Democratic Alliance (DA) in time for the nationwide municipal elections in 2000. The Democratic Alliance enjoyed some early success, winning 22.1 per cent of the vote, and gaining control of one of South Africa's largest municipalities, Cape Town.
The supposed vote-maximizing strategy of combining opposition parties appeared to have paid off. However, the Democratic Alliance now faced two problems, both of which would haunt it for years. First, the merger of these primarily white-led parties only reinforced the idea that these politicians were not interested in representing Africans. In this way the Democratic Alliance played into the hands of the African National Congress and its efforts to ‘de-legitimate the opposition as a viable competitor for the African, coloured, and Indian vote’ (Ferree Reference Ferree2011: 102). The merger was ‘bizarre’ because of the ideological combinations involved (Seekings Reference Seekings2005b: 22), furthering the Democratic Alliance's image as an exclusionary party, since the racial bases of the parties were the only obvious commonality.
Second, the merger was conducted without due attention either to the development of a coherent platform or to internal organizational issues. One Cape Town Democratic Alliance councillor recalls, ‘We underestimated the clash of organization cultures. It was a completely unmanaged merger.’ While the New National Party's organization and decision-making processes were hierarchical and fairly centralized, the Democratic Party ‘was more like a debating society’, with free-wheeling debate in its caucus.Footnote 3 These difficulties made a deep impression on party leadership and informed political strategies for years.
Rather than consolidating the Cape Town base, the minimal intraparty arrangements collapsed in the face of political and personality disagreements and scandals (Jolobe Reference Jolobe2007). In 2002 new legislation permitted floor-crossing between elections, allowing the New National Party-aligned politicians to defect from the Democratic Alliance and give control of both the Cape Town and Western Cape governments to the African National Congress.Footnote 4 The infighting and controversies surrounding the Democratic Alliance between 2000 and 2004 damaged efforts to mobilize its own base, and in 2004 the African National Congress won elections in the province for the first time. The Democratic Alliance had created a track record in office, but it was one of scandals, fragmentation and gridlock. The first opposition strategy – vote-maximizing mergers – failed; African National Congress dominance was deepened.
Unsuccessful Interparty Cooperation: The 2004 Democratic Alliance–Inkatha Freedom Party Campaign
Although party mergers proved counterproductive, opposition parties still sought to build an anti-dominant party voting bloc, but because of the Cape Town experience they shied away from permanent ties. With the demise of the New National Party, the Democratic Alliance (now mostly former Democratic Party members) and the ethnic Zulu Inkatha Freedom Party (IFP) were the two largest opposition parties nationally. Given the damage to the Democratic Alliance's image in the Western Cape, opposition focus shifted to the other province where competitiveness existed: KwaZulu-Natal (KZN), where the Inkatha Freedom Party had won majority support in 1994 and a plurality in 1999.
For the 2004 national and provincial elections, the Inkatha Freedom Party and Democratic Alliance announced the Coalition for Change, a formal agreement to hold joint campaign events, to share resources in developing common policy positions and to form governing coalitions if doing so would keep the African National Congress from power (Lodge Reference Lodge2005; Piper Reference Piper2005). The Coalition for Change was something of a lifeline for the declining Inkatha Freedom Party, a way of gaining nationwide organizational resources. For the Democratic Alliance, the party sought to ‘negate perceptions that the DA was confined to minority racial politics’ (Leon Reference Leon2008: 577). The lessons from the creation of the Democratic Alliance were learned: the leadership and organizational arrangements were made clear to all involved and, more importantly, a merger or even permanent ties were avoided.
However, despite smooth coordination at the level of party elites, this attempt at pre-election coalition building failed miserably with voters. After 2004, the African National Congress no longer needed the Inkatha Freedom Party as a coalition partner in KwaZulu-Natal. The Inkatha Freedom Party's decline became even more precipitous after this blow; in 2009 its support dropped to 22.4 per cent provincially and 4.55 per cent nationally, both to about half of the level of ten years before. Further, for the African National Congress the results had ‘huge psychological and symbolic significance, but more importantly, would forever lay to rest the fallacy peddled by the apartheid regime that the ANC is Xhosa-dominated and enjoys scant support among the Zulu’ (Gumede Reference Gumede2005: 248). Again, the opposition's strategy benefited the dominant party instead.
The Democratic Alliance and Inkatha Freedom Party took two lessons from the Coalition for Change. First, any interparty coordination should be explained very carefully to voters. One politician involved described it as ‘the Collision for Change . . . It actually switched off both our bases and we ended up having a low turnout.’Footnote 5 An Inkatha Freedom Party representative called the coalition the ‘biggest mistake of our lives . . . I remember I was working at the polling station and you could hear our illiterate guys saying, you just cross [on the ballot] whatever – DA, IFP, it's one thing. They didn't quite grasp it clearly, that we are a coalition but we are different. It turned out to be a disaster, as our numbers dropped.’Footnote 6 Afterwards, opposition politicians therefore frequently emphasized their ‘independence’, their ability to act autonomously of other parties in order to represent their supporters. Most of South Africa's opposition parties view official pre-election coalition building warily, usually avoiding it altogether. Only in recent years, after there have been successful post-election governing coalitions (described below), do some opposition politicians cautiously allow that opposition forces ‘are in a much better space now for realignment [of the party system] because we have learnt from experience and can get the conditions right before we move forward. [Any future coordination] will be no shot-gun marriage’ (Zille Reference Zille2009a). Opposition parties are much less willing now to work together electorally, despite the patterns in voter behaviour that suggest that doing so would be beneficial.
The second lesson from 2004 was less explicitly acknowledged but can be observed in both parties’ strategies: no longer do they attempt to expand their support nationwide to the detriment of shoring up their regionally defined bases. Instead, these parties refocused their efforts on maintaining and expanding their support in the subnational areas where they were already strong, to build competitiveness with geographically defined bases rather than demographically defined ones. Doing so would soon pay off spectacularly for the Democratic Alliance.
Capitalizing on a Lucky Break: 2006 Cape Town Elections and Governing Coalitions
A new strategy for operating in a dominant party system – forming post-election governing coalitions – emerged serendipitously after the 2006 municipal elections. The Cape Town electorate fragmented more than ever before, although a significant portion did rally behind the Democratic Alliance's popular mayoral candidate, Helen Zille. The result was a hung council with a Democratic Alliance plurality of 90 seats out of 210, followed by the African National Congress with 81 seats (see Table 1).
Cape Town City Council Seats Won at Elections
Note: The table shows the number of seats for each party awarded from each election. The number of seats held by each party changed between elections due to floor-crossing (defections) and by-elections. The parties included in the coalition also changed between elections; shaded boxes show the original 2006 multiparty coalition government. Dashes indicate that the party did not stand in Cape Town in that year.
The Democratic Alliance and six small parties formed a multiparty post-election coalition, the first in South Africa. The formal agreement reflected earlier experiences. First, no commitment was made to any cooperation beyond the council. This decision avoided the parties appearing to have little ideological common ground and therefore being cast as motivated by the desire to keep the African National Congress – and, by the African National Congress's extension, Africans – out of power. Instead, some argued that the 2006 election results were an opportunity to deepen South African democracy. One member said that the coalition was:
not out of a fundamental disagreement with the ANC. It's also about protecting the space for opposition parties in the future. Because if we don't take the opportunities for opposition parties now, history has shown us that the dominant party may well become the only party. That's not in the best interests of South Africans. So we need to have this contrast.Footnote 7
In short, post-election coalitions do not require ‘buying into [coalition partners’] values, or asking them to buy into your values.’Footnote 8
Second, drawing in part on the experience of the New National Party–Democratic Party merger, the parties formally emphasized their independence from one another in governance. The smaller parties, with 15 seats, feared being swamped by the Democratic Alliance's numerical advantage, so they insisted on caucusing among themselves and only then meeting with the Democratic Alliance to negotiate a single coalition position. It was also this priority for independence that initially kept the third-largest party in the city, the Independent Democrats (ID), from joining the coalition. Even a commitment to a governing coalition was too much for this party in 2006.
These arrangements worked remarkably well – so well that the coalition partners eventually declared it to be a model for elsewhere in the country. If election results so warranted, these parties were willing to form similar agreements. Importantly, the Democratic Alliance apparently put on hold its goal from the first part of the decade of creating one united opposition party. Instead of arguing that ‘a vote for a smaller party is a wasted vote’ as they did in 2004 (South African Press Association 2004), Democratic Alliance leaders began to praise other parties’ leaders whom they deemed to be acting in accord with the Democratic Alliance's goals of transparency and fair delivery of state functions. With varying degrees of explicitness, the opposition parties acknowledged that they had found a path for maintaining subnational power in a dominant party system. Thus a new opposition strategy was born.
This strategy – creating a model of opposition efficacy based on capitalizing on their control of Cape Town – required the platform provided by actually governing a major city. Without having the information with which to assess opposition parties – because those parties had not held many political offices – voters relied on heuristics such as parties’ perceived racial identities (Ferree Reference Ferree2006; Schulz-Herzenberg Reference Schulz-Herzenberg2007). But, as one non-Democratic Alliance member of the coalition said:
I'm the expression of my party in office. To voters we will say, if you want to see how [my party] will manage a political office, then have a look at these examples . . . So here is a track record now that we're able to build up . . . During an election campaign, we need to trumpet what we have done in office [here], because it's applicable anywhere in South Africa.Footnote 9
This councillor also noted that holding office creates additional campaign resources and visibility without much additional effort on his part. As part of the Cape Town leadership, he said, he personally presents the policy proposal that was devised in committee, which is then reported and the public gives him the credit. His being in government rather than in opposition makes voters aware of what he, and by extension his party, does.
Thus, with the Cape Town coalition, these parties sought to ‘demonstrate the alternative’ to African National Congress rule. Campaign materials in 2009 drew on the Democratic Alliance's record in subnational government, making its ‘strong track record in government’ a key piece of electoral appeal (Democratic Alliance 2009a, 2009b). This goal acknowledged one of the dilemmas of voters in a dominant party system: voters may be dissatisfied with the dominant party's rule, but to support another party is to take a risk on an untested and ultimately unknown entity; voters rarely take the risk of handing power to an untested party (Willey Reference Willey1998). Governing a major city provided a very visible illustration of what the opposition parties would do with political power, creating a ‘track record which black African voters can turn to to evaluate the party, before they turn to using the DA's traditionally negative party image [with regards to race]’ (Schulz-Herzenberg 2008: 139, emphasis in original). In short, recognizing that ‘when oppositions lack credibility, voters are stranded on the shores of the dominant party’ (Ferree Reference Ferree2011: 29), the opposition parties made gaining credibility a priority.
To convey its governing record, opposition rhetoric focused on overcoming the racial framing strategy of the African National Congress identified by Ferree. Instead, the opposition sought to reframe political debate to questions about corruption and democracy. Then-Democratic Alliance leader Tony Leon writes that, after the 2004 election, ‘It seemed to us that much of what passed for the other opposition parties . . . were desperately positioning themselves within the ANC's analysis – rather than challenging it or seeking an alternative vision . . . I would point to the dangers of the ANC's hegemonic impulses and the jeopardy it posed to our constitutional order’ (Leon Reference Leon2008: 22). Accordingly, the Democratic Alliance's 2006 local election campaign slogan was ‘Stop Corruption. Start Delivery. Vote DA’ (Democratic Alliance 2006).
This anti-corruption theme continued after the multiparty coalition took power. With several national-level African National Congress leaders embroiled in a high-profile scandal over an arms deal (see Feinstein Reference Feinstein2007) and the contrast between the media coverage of corruption in several municipal governments nationwide and the apparently relatively scandal-free Cape Town government, opposition parties had ample evidence to argue that they were less corrupt. A top Democratic Alliance leader argued, ‘The only way you get rid of identity and race-based politics is by winning somewhere on the basis of issue-based politics, and by demonstrating, from the bottom up, that that's what you can do.’Footnote 10 Holding office somewhere became an explicit part of the strategy to win more offices.
Meanwhile, the African National Congress provincial government of the Western Cape made several legal moves that gave the coalition ammunition to argue that the African National Congress was not committed to democracy (see Faull Reference Faull2006). The Democratic Alliance frequently argued that the dominant party abused its power. Therefore the Cape Town government was not just a viable alternative to the African National Congress, but a better alternative. By 2009 the Democratic Alliance frequently spoke of ‘two political philosophies . . . the “open opportunity driven society for all” versus the “closed, crony society for some”’(Zille Reference Zille2009b). The Democratic Alliance argued that the dominant party was ‘deeply illiberal, authoritarian and fundamentally power hungry’ (Leon Reference Leon2008: 509). In the Democratic Alliance's rhetoric, the African National Congress leadership does not ‘understand the constitutional democracy . . . They think that they have the sole moral right to govern because they were a legitimate liberation movement.’Footnote 11 Therefore, ‘it is imperative that [the Democratic Alliance] keep the ANC below a two-thirds majority nationally, so that it cannot change the Constitution to continue the power abuse designed to enrich its leaders and shield them from the consequences’ (Democratic Alliance 2009a). The narrative the Democratic Alliance related about the previous 15 years accepted that the African National Congress was ‘the defining democratic movement in the country’ but argued that the liberation force had changed, ‘become increasingly self-serving . . . [and] has lost the moral high ground it once dominated’ (Democratic Alliance 2008). The advantage of this emphasis rather than the Democratic Alliance's controversial 1999 ‘Fight Back’ campaign is that an anti-corruption, overtly pro-democracy platform is less easily portrayed by the African National Congress as a racially exclusive position.
In this way, the opposition parties in South Africa adopted a strategy to challenge a democratic dominant party, similar to what Mexican parties did to challenge the Institutional Revolutionary Party (Partido Revolucionario Institucional – PRI), a hegemonic authoritarian party, in the 1980s: they created a new political cleavage that could divide the ruling party's base. Both Magaloni (Reference Magaloni2006) and Greene (Reference Greene2007) identify a ‘regime cleavage’ in hegemonic Mexico, which shifted the salient political issue away from economic policy – on which opposition parties were divided – and to the question of democratic reforms – on which the opposition parties and even some Institutional Revolutionary Party supporters were united. This regime cleavage united voters against the Institutional Revolutionary Party, which lost the 2000 presidential election. I argue that a similar cross-cutting cleavage can be created with some success in a democratic dominant party system.
Testing the Strategy: Maintenance and Expansion, 2009
Two challenges for the new opposition strategy loomed, both key for further expansion of competitiveness in the party system: they needed to maintain the multiparty coalition itself to continue to build the record of governance, and then to expand the share of opposition support nationwide.
Not all went smoothly within the Democratic Alliance-led Cape Town governing coalition. The defection of the African Muslim Party in early 2007 left the coalition as a minority government. To avoid the imminent collapse of the city government – in its leadership's words, to ‘maintain stability’ in the city – the Independent Democrats contributed their 23 seats to the coalition. While there is no reason not to take the Independent Democrats’ stated motivations at face value, it also seems unlikely that the Independent Democrats would have joined the coalition if the coalition had not shown that such arrangements might be the path to competitiveness. Thus the coalition ‘demonstrated the alternative’ to party elites as well as to voters.
The addition of the Independent Democrats to the governing coalition made the coalition's majority more secure and hence may have contributed to further disintegration of the coalition. That is, without the cost of quitting the coalition being the elevation of the African National Congress to power in Cape Town, some coalition partners found the benefits of working with the Democratic Alliance to be too low and withdrew from the coalition. Combined with some of the smallest parties collapsing (with the majority of their councillors joining the Democratic Alliance), by 2010 only the Democratic Alliance, the Independent Democrats and the United Democratic Movement (UDM) remained in the coalition.
Despite these problems among the coalition partners, what could be called the ‘Cape Town model’ – competition in elections but cooperation in governance, while challenging the dominant party's democratic commitment – became the explicit strategy for opposition party growth in the South African dominant party system. The African Christian Democratic Party (ACDP), Freedom Front Plus (FF+), Independent Democrats, Inkatha Freedom Party and the United Democratic Movement all announced that they would enter coalitions, although only if the electoral results made coalitions necessary (Thamm Reference Thamm2008). The plan was to wrest the Western Cape from African National Congress control in 2009 and then to use that as a springboard for further growth in subsequent elections.Footnote 12
Two factors make such expansion uncertain. First, Cape Town's demographics are sufficiently different from other areas that opposition success there can be incorporated into the African National Congress's strategy of framing the city's governing parties as ‘white’ or at least non-African. That is, the African National Congress can dismiss the success of opposition parties among a population that includes only about 35 per cent Africans. On this count the Democratic Alliance's strategy for growth fails; in 2009 it gained significant numbers of African National Congress voters but not many black African ones (Jolobe Reference Jolobe2009). It may be that the Democratic Alliance cannot expand to parts of the country that have only small non-African populations. If the racial heuristic continues to hold significant value for much of the electorate, then the Democratic Alliance will remain a regional, not national, power. This factor could be overcome with the growth of an opposition party with a base in the black African electorate but it is unclear if any such opposition could emerge through subnational growth, rather than through elite factionalism within the African National Congress, because of the strength of the African National Congress. (After all, even in highly diverse Cape Town, the original multiparty coalition barely had enough seats to form a government). Any African National Congress elite breakaway faces the challenge of sharing the African National Congress's governing record and therefore being unable to distinguish itself from the dominant party. Second, some of the coalition partners were sceptical of expanding post-election coalitions into provincial and national politics. The South African Constitution delineates the issue areas for which each sphere of government is responsible. Smaller parties mention that while they agree on the issues assigned to local government, they have serious differences with the Democratic Alliance on national and provincial issues. Of particular concern are some zero-sum social issues such as abortion and the death penalty.Footnote 13 If potential coalitions need the Democratic Alliance, as the largest opposition party they may struggle to create governing agreements.
Despite these potentially limiting factors, in 2009 the Democratic Alliance was the only party in the National Assembly to increase its share of the national vote and won outright control of the Western Cape provincial government. In the 2011 municipal elections, the Democratic Alliance and the Congress of the People posed the most serious challenges ever to the African National Congress in its stronghold of the Eastern Cape. The African National Congress won only 51.91 per cent of the vote in the province's largest city, Port Elizabeth. Clearly, the country's political landscape was becoming more competitive.
The Congress of the People created a test of the degree to which the opposition leadership had learned the lessons from the earlier efforts at party coordination. Questions of alliances and coalitions arose from the start. The Congress of the People and the Democratic Alliance performed a delicate campaign in which they spoke positively of one another but refrained from endorsing each other, while both employing similar rhetoric about the African National Congress's commitment to democracy. Congress of the People chairman Mosiuoa Lekota said, ‘If it is to defend the democracy, we will join or strike an alliance with the DA’ (in Makinana Reference Makinana2008). However, the Congress of the People struggled to create a clear identity for itself, as other opposition parties before it had. The African National Congress and its allies cast the Congress of the People as ‘the Black DA’ (Andrew Reference Andrew2009). Perhaps as a result, after the 2009 elections Congress of the People leader Mbhazima Shilowa said, ‘Alliances, at this point in time, are out of the question, but we will cooperate with others on specific issues . . . Cope will not go into a coalition which it does not lead, where its identity will be subsumed by others’ (in Booysen Reference Booysen2009: 112). Similarly, the Democratic Alliance emphasized its desire and ability to win elections outright. Party leader Zille said, ‘We are planning to win alone where we can, but where we cannot, we are happy to enter into alliances and coalitions . . . We plan to win the Western Cape alone next year’ (in Makinana Reference Makinana2008). Although the Cape Town coalition was barely mentioned and the Coalition for Change not at all, it is clear that the Congress of the People was leery of being cast as just another stooge of the Democratic Alliance, just as the Democratic Alliance was leery of tying itself to another disorganized party with significant political baggage.
While the smaller parties joined the Cape Town coalition highly cognizant of the danger of being overshadowed by the Democratic Alliance, by 2010 the Democratic Alliance had largely consolidated opposition forces in the city under its banner alone. The process culminated when the Independent Democrats’ leadership agreed to merge with the Democratic Alliance. The lessons of earlier coalitions and mergers can be seen in how the two parties’ leaders responded to a question of how the two parties still differed. The Democratic Alliance's Zille, reflecting the experience of the Democratic Party–New National Party merger, answered, ‘We have different political histories. Our parties have different organisational structures. We are finalising a joint set of local government policies’ (in Faull Reference Faull2010). In contrast, Patricia de Lille, the Independent Democrats’ founder, emphasized the disagreements on substantive issues that had long concerned the Democratic Alliance's partners, saying, ‘On some economic issues we differ. We are social democrats, while the Democratic Alliance are liberal democrats’ (in Faull Reference Faull2010). However, both parties sought the approval of their activists before embarking on the formal agreement, recognizing that they could not impose cooperation on their members (Faull Reference Faull2010). By the end of 2011 reports surfaced that the Congress of the People's leadership sought a similar ‘engagement’ to eventually merge with the Democratic Alliance (Majavu Reference Majavu2011).
Whether the South African electorate is more prepared for such arrangements within the opposition than it was before the Cape Town coalition remains to be seen. What is clear is that small opposition parties face a potentially untenable choice. Those smaller parties that joined the 2006 Cape Town coalition have been either absorbed by the Democratic Alliance in one manner or another, or have broken publicly and bitterly with it and have experienced severe declines in support and access to power. Other opposition parties must be wary of similar fates. The success of the Democratic Alliance may impede its ability to build future coalitions and therefore its ability to repeat the Cape Town strategy in other cities and provinces.
The Reaction of the African National Congress
The African National Congress was not passive in response to these opposition strategies. It ‘successfully framed the 2009 election as a face-off between well-off blacks and whites on the one hand and the poor black majority on the other – rather than on an examination of the government's record in power’ (Gumede Reference Gumede2009). The implication that the African National Congress knew itself to be vulnerable if the campaign focused on its record, given widespread protests over failures to deliver government services, is not farfetched.
However, during this period, the African National Congress also faced some of the most extreme intraparty factionalism of its existence, to the benefit of opposition. The divisions that led to the creation of the Congress of the People were arguably deepest in the Western Cape, resulting in dysfunctional and discredited provincial party structures that at one point even failed to register candidates for by-elections on time (Dawes Reference Dawes2008; Joubert Reference Joubert2009). Undoubtedly the Democratic Alliance was lucky in its subnational governing success coming in the same province where the African National Congress was ill-prepared to respond organizationally. Still, it should be noted that these events are endogenous: the Democratic Alliance's success is both a contributing factor to and result of the African National Congress's dysfunction. A weakened dominant party is not sufficient to cause party system change; former dominant party voters need an alternative to vote for, or they will simply drop out of the electorate. Despite most sources of African National Congress internal strife existing for much of the previous decade, it was only in 2009 – with the establishment of an opposition record of governing a major city that overcame the racial heuristics of political parties – that the South African electorate shifted in ways that showed potential for systemic change.
Conclusions and Implications
This article traces events in Cape Town and the surrounding Western Cape province between 1999 and 2010, showing how opposition parties’ actions led first to the consolidation of dominant party power and later to the growth of opposition. In doing this, I show how party strategies reflect learning by party elites, as opposition parties used looser and looser arrangements in coordinating their efforts against the dominant party. After difficult lessons in the pitfalls of both party mergers and pre-election coalitions, the Democratic Alliance in particular used a highly visible post-election coalition to position itself to build more permanent arrangements towards an oft-stated goal of party system realignment.
I also show how subnational electoral shifts may affect national politics, as South African parties’ fortunes nationally were often foreshadowed by their fortunes in municipal and provincial elections. While opposition parties struggled in the wake of failed coalitions and scandals in the first few elections, the success of the coalition in Cape Town heralded further opposition growth throughout South Africa. This pattern demonstrates the importance of a track record for voters; that is, the evidence here suggests that voters do want direct evidence of parties’ and candidates’ skill in governing rather than only campaign rhetoric. Survey research should explore directly whether this is true, as it is potentially a vital point for the evolution of dominant party systems.
Observers of African politics argue that if opposition parties combined resources, through mergers and other agreements, they might find greater success in electoral contests against dominant parties (Mozaffar and Scarritt Reference Mozaffar and Scarritt2005; Rakner and van de Walle Reference Rakner and van de Walle2009). The original merger creating the Democratic Alliance suggests that this advice should be adopted with caution. If opposition is fragmented by real ideological disagreements, mergers only reinforce the dominant party's arguments that opposition leaders are opportunistic power-grabbers. If opposition is fragmented by personality divisions, mergers could result in behaviours that only worsen voters’ perceptions of the opposition. In light of the results of the Democratic Alliance–Inkatha Freedom Party Coalition for Change, even pre-election coalitions should be approached cautiously for fear of alienating voters. Even the latest mergers – while slow and systematic and careful – constitute a risk on the part of South Africa's opposition parties, as voters may still see the mergers as ideologically odd and therefore reject them as merely crass office-seeking.
Whether or not they lead directly to the end of dominance, the causes of shifts in competitiveness show where the African National Congress – and, by extension, dominant parties as a class – are vulnerable and how opposition parties might successfully gain a share of political power. South African parties increasingly focus their criticisms not on the ideological or policy differences between themselves and the dominant party, but instead on the contrasts they perceive in governance, corruption and tolerance for dissent. A coalition of voters united in the common cause of good governance can cut across other cleavages and thus potentially overcome the problems of fragmentation that face opposition parties in dominant party systems.
Acknowledgements
The research presented here was supported in part by the Mershon Center for International Security Studies at the Ohio State University. Many thanks to Miryam Farrar Chandler, Michael Cohen, Ana Espinal-Rae, Richard Gunther, Dennis P. Johnson, Marcus Kurtz, Goldie Shabad, Christina Xydias and the journal's two anonymous reviewers for their comments on this or earlier versions.
