Census elections like those in South Africa reflect the decisions of millions of individual voters. To understand the origins of the census, we must therefore begin with an exploration of individual-level voting behavior. The question can be framed as this: When an African voter chooses the African National Congress (ANC) over one of its “white” competitors (the New National Party or the Democratic Alliance), what explains her behavior? Is it the need to express her identity as an African? Is it general antipathy to whites and the parties that represent them? Or are her motivations more instrumental, a function of policy preferences and performance evaluations? If they are instrumental, what role, if any, does race play? And, perhaps most importantly, how do these factors interact with one another?

In addressing these questions, this chapter distills and tests three hypotheses from the general voting literature. The first hypothesis (“expressive voting”) holds that racial voting results as voters use the ballot box to express their identities as members of racial groups. The second hypothesis (“politics-as-usual”) explains racial voting through nonracial factors. It suggests that racial voters, like all voters, care what governments do and how well they do it. Convergence of voting behavior within racial groups occurs because group members share common policy preferences and performance evaluations – not a common identity. The third hypothesis (“racial heuristics”) shares the instrumentalism of the second but gives race a more active role in explaining behavior.

On April 15, 2004, South Africa held its third post-apartheid national elections, which the ruling ANC won by a wide margin, taking almost 70 percent of the total vote and winning outright majorities in seven out of nine provinces. In the remaining two provinces – the Western Cape and KwaZulu-Natal – it won pluralities and quickly formed governments with coalition partners. As in 1999, the Democratic Party (now the Democratic Alliance) captured the most votes (around 12 percent of the total) of any opposition party, increasing its size by about 3 percentage points from the earlier election. ANC and DA gains came at the expense of the Inkatha Freedom Party (IFP), which saw support dip from about 8.5 percent in 1999 to 7 percent in 2000, and, more spectacularly, the New National Party (NNP), whose support fell from around 7 percent to under 2 percent (even in the Western Cape, the NNP's support evaporated). Two somewhat odd pre-electoral coalitions – one between the ANC and the NNP and the other between the DA and the IFP – seemed only to strengthen the hands of the ANC and DA. In terms of the smaller opposition parties, the United Democratic Movement (UDM) fell under 3 percent of the overall vote and also lost ground on its home turf (the Eastern Cape).

While the beginnings of South Africa's racial census lie in its past, in the effects apartheid had on voters' beliefs about race and destiny and the reputations parties forged during this period, the enduring imprint of race on elections reflects current politics, in particular the ruling party's ability to use the powers and benefits of office to frame elections and prevent opposition parties from transforming the uncompetitive party images they inherited from apartheid into ones that would give them broader appeal in the electorate. By keeping these parties “white,” by preventing them from evolving in a more multi-racial direction, the ANC has rendered them toothless.

In this final chapter, I consider two additional cases: El Salvador and Israel. I show that dominant parties in both countries used image control and negative framing strategies to discredit their opponents and fortify their own hold on power. The power of the frame, the ability to shape how the electorate views the opposition, is thus a general tool that dominant parties use to continue their rule. While much current research has focused on resource-based strategies for control (especially clientelism), the case material from El Salvador, Israel, and South Africa all point to the importance of broadening our understanding of the repertoires of dominance. Dominant parties most certainly depend on their monopoly over resources to maintain power. At the same time, they cannot buy off all of the voters all of the time.

Given the importance of the transformation of candidate characteristics to party label change, why have South Africa's opposition parties not moved faster in transforming their lists and why, in particular, have they resisted changing the very top positions? In this chapter, I explore how party size affects the ability of parties to transform their candidate demographics, paying particular attention to the role that candidate quality plays in this process. I argue that small parties are especially sensitive to the quality of incoming candidates. Because they control only a small number of seats, any seats they take from existing members for new recruits are necessarily top-level seats, most likely filled with high-quality party leaders. To maintain overall candidate quality, small parties must therefore recruit only high-quality new candidates. At the same time, small parties face significant disadvantages in obtaining these candidates because they are neither able to grow their own nor poach them from other parties. As a result, the challenges of elite incorporation are more difficult for small parties. I argue that these intuitions hold regardless of the internal organization of the party and explain why the pace of candidate change has been slow for South Africa's opposition parties. I then evaluate several observable implications of this argument using the data on candidate career paths introduced in the previous chapter. I also explore the collapse of the New National Party shortly after the 2004 election and the role candidate transformation played in it.

Chapter 2 established the centrality of racial credentials to persuasion in South Africa. It demonstrated that party images, which are rooted in the country's apartheid past, powerfully shape African voting behavior. The remainder of the book focuses on the process of party image change. If party images are the lynchpin to persuasion in South Africa, have the opposition parties attempted to change their images in a way that would make them more attractive to the electorate? If they have, what strategies have they used? And, finally, why have they not been more successful?

In this and the following chapters, I will show that the opposition parties in South Africa have understood the link between party labels and persuasion and made extensive efforts during the 1994–2004 period to transform themselves from “white” parties to multiracial ones. They faced daunting challenges in this process, however. For as much as opposition parties have wished to transform their images, the ANC has wanted to preserve them, thereby preventing the emergence of viable competitors for the African vote. The party images of the opposition have therefore formed a central battleground in South Africa's post-apartheid electoral landscape, one in which the ruling party has outplayed the opposition.

In this chapter, I focus on the campaigns of the ANC, NP, and DP during the first post-apartheid election, held April 27, 1994.

The previous three chapters showed the many challenges South African opposition parties have faced in trying to use campaigns to alter party images. I now turn to an alternative method that parties have pursued toward this end: the manipulation of candidate characteristics. I argue that where the ideological reputations of parties are not well developed, candidate characteristics figure prominently in the formation of party labels. Changing candidate characteristics therefore offers parties a means to alter their party labels. This works through two mechanisms. First, a preference effect: Changing the decision makers in the party alters beliefs about the outcomes the party is likely to pursue. And second, a signaling effect: By unseating party stalwarts in favor of candidates with a different set of characteristics, a party sends a potent signal about its intentions.

I test these intuitions in the South African context. I show that parties (as well as journalists, pundits, and academics) frequently discuss the racial balance of candidate lists. Information about the racial balance of lists is therefore common knowledge and widely believed to be important. Using a new dataset that tracks the changing racial balance of the candidates of the ANC, the NNP, and the DA, I then demonstrate that the parties with the strongest incentives to change their labels – the “white” opposition parties – have indeed made extensive changes to their profiles since the end of apartheid. Starting out as exclusively white, these parties were at least 30 percent African by 2004.

From an aerial viewpoint, post-apartheid South African elections bear an unmistakable racial imprint: Africans vote for one set of parties, whites support a different set of parties, and, with few exceptions, there is no crossover voting between these groups, which together make up more than 90 percent of the South African population. Such sharp racial contours of voting have earned South African elections the dubious distinction of being “racial censuses:” Voters line up with their racial groups, seemingly without thought to issues, performance, or any of the other politics-as-usual factors that drive elections in other countries. Indeed, elections look so deeply racial that one wonders if politics has anything to do with it at all. What role can persuasion play if voters simply register their social identity when they enter the polling booth?

However, behind the racial imprint lies a puzzle, for racial identities in South Africa are neither pervasive enough nor unique enough to account for South African voting outcomes. African voters – who comprise around three-quarters of the electorate and drive the census – are a highly diverse group; some place primary importance on race, but many more emphasize nonracial identities. Moreover, liberation jubilation aside, Africans have not been uniformly committed to the ruling African National Congress (ANC). Even in 1994, during the very first post-apartheid election campaigns, African voters expressed uncertainty about the ANC. Surveys consistently show a large group of independent Africans, up to 50 percent at some points in the recent past.