Parties’ motivations for electoral reform under the democratic transition in South Africa

Institutional change and party motivations[Up]

Electoral systems are not a group of neutral mechanisms to translate voters’ preferences into representation ( ‍Norris, Pippa. 1997. “Choosing electoral systems: Proportional, majoritarian and mixed Systems”, International Political Science Review 18 (3): 297-312. Available at: https://doi.org/10.1177/019251297018003005Norris, 1997). Because the design of the electoral system is the gate for getting legislative seats, political actors try to impose their rational interest on it. Consequently, the design of electoral systems is not a kind of inertia that makes the institutional legacy persists, but rational actors’ choices about whether or not to transform these institutions. Electoral rules have usually proved to resist change until the 1990s^‍‍[2] ( ‍Nohlen, Dieter 1984. “Changes and choices in electoral systems”, in Arend Lijphart and Bernard Grofman (eds.), Choosing an Electoral System: Issues and Alternatives. New York: Praeger.Nohlen, 1984: 218). Notwithstanding the succession of reforms in the 1990s and early 2000s, major electoral reforms remain scarce ( ‍Katz, Richard S. 2005. “Why Are There so Many (or so Few) Electoral Reforms?”, in Michael Galloper and Paul Mitchell (eds.), The Politics of Electoral Systems. Oxford: Oxford University Press. Available at: https://doi.org/10.1093/0199257566.003.0003Katz, 2005). Boix ( ‍Boix, Carles. 1999. “Setting the rules of the game: The choice of electoral systems in advanced democracies”, The American Political Science Review, 93 (3): 609-624. Available at: https://doi.org/10.2307/25855771999) identifies four different phenomena that may lead to the transformation of the electoral arena, when the ruling parties attempt to maximize their representation in a changed electoral arena ( ‍Rokkan, Stein. 1970. Citizens, Elections, Parties. Oslo: Universitetsforlaget.Rokkan, 1970;  ‍Boix, Carles. 1999. “Setting the rules of the game: The choice of electoral systems in advanced democracies”, The American Political Science Review, 93 (3): 609-624. Available at: https://doi.org/10.2307/2585577Boix, 1999). South Africa constitutes one of these examples, with the enfranchisement of a large number of people who had been excluded from legitimate political participation under the British Empire and the Apartheid regime. In this scenario, whether or not the ruling parties embrace reforms depends on two conditions: the new parties’ strength and the ruling parties’ coordinating capacity; that is, whether they are tied in votes (ibid.: 611).

In general, the rational choice approach posits that the relevant actors have a fixed set of preferences and behave entirely instrumentally so as to maximize the attainment of such preferences; that is, they act in a highly strategic manner that presumes extensive calculation. In other words, first, that an actor’s behavior is likely to be driven, not by impersonal historical forces, but by a strategic calculus; and second, that this calculus will be deeply affected by the actor’s expectations about how others are likely to behave as well ( ‍Radnitzky, Gerard. 1987. “Cost-benefit thinking the methodology of research: The economic approach applied to key problems to the philosophy of science”, in Gerard Radnitzky and Peter Bernholz (eds.), Economic imperialism: The economic approach. New York: Paragon House.Radnitzky, 1987;  ‍Ostrom, Elinor. 1999. “Institutional rational choice: An assessment of the institutional analysis and development framework”, in Paul A. Sabatier (eds), Theories of the policy process. Cambridge: Westview Press.Ostrom, 1999;  ‍Peters, B. Guy. 1999. Institutional theory in political science. The New Institutionalism. London: Pinter. Peters, 1999). The model introduced here predicts that political parties involved in the reform of the electoral system during a negotiated democratic transition follow office-seeking and policy-seeking strategies. It also predicts that parties moved by partisan motivations use no partisan motivations as an instrument to generate popular support for their proposals ( ‍Bowler, Shaun and Todd Donovan. 2015. The Limits of Electoral Reform. Oxford: Oxford University Press.Bowler and Donovan, 2015: 4-‍6). This hypothesis, like all hypotheses within the rational choice paradigm, assumes a set of behavioral assumptions: The relevant actors have a fixed set of preferences and they behave entirely instrumentally to maximize the attainment of these sets of preferences in a strategic manner that presumes extensive calculation. In the determination of political outcomes, I assume that political parties evaluate electoral institutions based on the expected effect of each alternative for their own partisan interest.

This case study follows assumptions of partisan strategic rationality and models that operationalize power-maximization in terms of maximizing partisan shares of seats in the legislature ( ‍Brady, David and Jongryn Mo. 1992. “Electoral Systems and Institutional Choice: A case Study of the 1988 Korean Elections”, Comparative Studies, 24 (4): 405-429. Available at: https://doi.org/10.1177/0010414092024004001Brady and Mo, 1992;  ‍Boix, Carles. 1999. “Setting the rules of the game: The choice of electoral systems in advanced democracies”, The American Political Science Review, 93 (3): 609-624. Available at: https://doi.org/10.2307/2585577Boix, 1999;  ‍Colomer, Josep M. 2005. “It’s Parties that choose electoral system”, Political Studies, 53 (1): 1-21. Available at: https://doi.org/10.1111/j.1467-9248.2005.00514.xColomer 2005;  ‍Benoit, Kenneth and John W. Schiemann. 2001. “Institutional Choice in New Democracies: Bargaining over Hungary’s 1989 Electoral Law”, Journal of Theoretical Politics, 13 (2): 159-188. Available at: https://doi.org/10.1177/0951692801013002002Benoit and Schiemann, 2001;  ‍Benoit, Kenneth. 2004. “Models of electoral system change”, Electoral Studies, 23 (3): 363-389. Available at: https://doi.org/10.1016/S0261-3794(03)00020-9Benoit, 2004;  ‍Farrell, David M. and Ian McAllister. 2005. “Australia: The Alternative Vote in a Compliant Political Culture”, in Michael Gallagher and Paul Mitchell (eds.), The Politics of Electoral Systems. Oxford: Oxford University Press. Available at: https://doi.org/10.1093/0199257566.003.0004Farrell and McAllister, 2005). Recent contributions to the institutional change debate from the rational choice theory acknowledge its limitations ( ‍Vowles, Jack. 1995. “The Politics of electoral reform in New Zealand”, International Political Science Review, 16 (95): 95-115. Available at: https://doi.org/10.1177/019251219501600107Vowles, 1995;  ‍Denemark, David. 2001. “Choosing MMP in New Zealand: Explaining the 1993 Electoral Reform”, in Matthew Soberg Shugart and Martin P. Wattenberg (eds.), Mixed-Member Electoral Systems: The best of Both Worlds? Oxford: Oxford University Press.Denemark 2001;  ‍Dunleavy, Patrick and Helen Margetts. 1995. “Understanding the dynamics of electoral reform”, International Political Science Review, 16 (1): 9-29. Available at: https://doi.org/10.1177/019251219501600102Dunleavy and Margaretts, 1995;  ‍Sakamoto, Takayuki. 1999. “Explaining electoral reform: Japan versus New Zealand and Italy”, Party Politics, 5 (4): 419-438. Available at: https://doi.org/10.1177/1354068899005004001Sakamoto 1999;  ‍Shugart, Matthew. 2001. “Electoral Efficiency and the Move to Mixed-Member Systems”, Electoral Studies, 20: 173-193. Available at: https://doi.org/10.1016/S0261-3794(00)00007-XShugart 2001;  ‍Siaroff, Alan. 2003. “Spurious Majorities, Electoral Systems and Electoral System change”, Commonwealth and Comparative Politics, 41 (2): 143-161. Available at: https://doi.org/10.1080/14662040412331310131Siaroff 2003;  ‍Andrews, Josephine and Robert W. Jackman. 2005. “Strategic Fools: Electoral Rule Choice under Extreme Uncertainty”, Electoral Studies, 24 (1): 65-84. Available at: https://doi.org/10.1016/j.electstud.2004.03.002Andrews and Jackman, 2005 ). Vowles ( ‍Vowles, Jack. 2008. “Systemic Failure, Coordination, and Contingencies: Understanding Electoral System Change in New Zealand”, in André Blais (eds), To Keep or To Change First Past The Post? The Politics of Electoral Reform. New York: Oxford University Press. Available at: https://doi.org/10.1093/acprof:oso/9780199539390.003.00072008) notes at the intersection of normative critiques of the existing rules and rational interest of political actors that reform is most likely to occur. Rahat ( ‍Rahat, Gideon. 2004. “The study of the politics of electoral reform in the 1990s: theoretical and methodological lessons”, Comparative Politics, 36 (4): 461-479. Available at: https://doi.org/10.2307/41501712004) proposes a multistage historical comparative approach after analyzing the electoral reforms in Japan, Israel, Italy, and New Zealand. Following Rahat, party perceptions are to be understood as the product of a battle over interpretations between mainly two hard cores of reformers and opponents; party interests result from the adoption of certain perceptions of both the political consequences of the reform itself and the behavior towards it; and party motivations are the end results of these weighed calculations; and these different dimensions should be studied separately during the negotiation of drafting electoral systems (ibid.: 475).

Some recent works develop explanations of institutional change emphasizing the importance of taking account of coalition dynamics in the study of electoral reform in consolidated democracies ( ‍Strøm, Kaare and Wolfgang C. Müller. 1999. Policy, Office, or Votes? How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press.Strøm and Müller, 1999;  ‍Bawn, Kathleen. 1993. “The Logic of Institutional Preferences: German Electoral Law as a Social Choice Outcome”, American Journal of Political Science 37 (4): 965-989. Available at: https://doi.org/10.2307/2111539Bawn, 1993;  ‍Benoit, Kenneth. 2004. “Models of electoral system change”, Electoral Studies, 23 (3): 363-389. Available at: https://doi.org/10.1016/S0261-3794(03)00020-9Benoit, 2004,  ‍Benoit, Kenneth. 2007. “Electoral laws as political consequences: explaining the origins and change of electoral institutions”, Annual Review of Political Science, 10: 363-390. Available at: https://doi.org/10.1146/annurev.polisci.10.072805.1016082007;  ‍Renwick, Alan. 2005. “Modelling Multiple Goals: Electoral System Preferences in Hungary in 1989”, Europe-Asia Studies, 57 (7): 995-1019. Available at: https://doi.org/10.1080/09668130500301626Renwick, 2005). In his analysis of the forces underlying the choice of a PR electoral system in Germany, Bawn presents a model in which preferences over institutions derive from preferences over policy ( ‍Bawn, Kathleen. 1993. “The Logic of Institutional Preferences: German Electoral Law as a Social Choice Outcome”, American Journal of Political Science 37 (4): 965-989. Available at: https://doi.org/10.2307/21115391993: 967). Using Bawn’s model, Remington concludes that a theory of strategic choice based solely on preferences defined over policy outcomes cannot account for the adoption of the 1993 Russian electoral system or its 1995 successor ( ‍Remington, Thomas F. and Steven S. Smith. 1996. “Political Goals, Institutional Context, and the Choice of an Electoral System: The Russian Parliamentary Election Law”, American Journal of Political Science, 40 (4): 1253-1279. Available at: https://doi.org/10.2307/2111750Remington and Smith,1996: 1256). Uncertainty and miscalculation were critical to the choice of an initial electoral system. The degree to which political actors are motivated either by policy goals or ambition to hold office is not always too obvious Katz ( ‍Katz, Richard S. 2005. “Why Are There so Many (or so Few) Electoral Reforms?”, in Michael Galloper and Paul Mitchell (eds.), The Politics of Electoral Systems. Oxford: Oxford University Press. Available at: https://doi.org/10.1093/0199257566.003.00032005). Actors might take into account factors other than their own partisan interest when deciding on institutional design. Budge and Laver argue that “office can be sought both as an end in itself and as a means to fulfil policy objectives Similarly, policy can be pursued both as an end in itself and as a means to achieve office” ( ‍Budge, Ian and Michael Laver. 1986. “Office seeking and policy pursuit in coalition theory”, Legislative Studies Quarterly, 11 (4): 485-506.1986: 486). Morevoer, evaluations of the system fairness have been found to have had a strong impact on support for the change of the electoral system in the New Zealand in 1993 ( ‍Banducci, Susan A. and Jeffrey A. Karp. 1999. “Perceptions of Fairness and Support for Proportional Representation”, Political Behavior, 21 (3): 217-238. Available at: https://doi.org/10.1023/A:1022083416605Banducci and Karp, 1999).

Explanations based on office-seeking and policy-seeking preferences in the South Africa[Up]

Following the categories of party motivations for choosing electoral institutions developed by Strøm and Müller ( ‍Strøm, Kaare and Wolfgang C. Müller. 1999. Policy, Office, or Votes? How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press.1999), Bawn ( ‍Bawn, Kathleen. 1993. “The Logic of Institutional Preferences: German Electoral Law as a Social Choice Outcome”, American Journal of Political Science 37 (4): 965-989. Available at: https://doi.org/10.2307/21115391993 ) and Benoit ( ‍Benoit, Kenneth. 2004. “Models of electoral system change”, Electoral Studies, 23 (3): 363-389. Available at: https://doi.org/10.1016/S0261-3794(03)00020-92004,  ‍Benoit, Kenneth. 2007. “Electoral laws as political consequences: explaining the origins and change of electoral institutions”, Annual Review of Political Science, 10: 363-390. Available at: https://doi.org/10.1146/annurev.polisci.10.072805.1016082007), I explore the transformation of the South African electoral system during the transition to democracy, identifying policy-seeking, office-seeking and general interest preferences, incorporating the insights of Bawn model. Bawn model ( ‍Bawn, Kathleen. 1993. “The Logic of Institutional Preferences: German Electoral Law as a Social Choice Outcome”, American Journal of Political Science 37 (4): 965-989. Available at: https://doi.org/10.2307/21115391993: 967-‍968) is based upon three assumptions: (a) parties preferences for electoral systems that result in the most desirable policy outcomes; (b) parties make use of all available information about the electoral’s voting preferences; and (c) parties participating in the choice of electoral institutions know the preferences of the other parties involved in the choice of electoral system. By contrast, explanations based on office-seeking preferences are closely related to winning office. From this point of view, any party will prefer rules that maximize its own share of legislative seatsover the share of seats other parties might gain, regardless of their compatibility in terms of policy goals or ideology ( ‍Benoit, Kenneth. 2004. “Models of electoral system change”, Electoral Studies, 23 (3): 363-389. Available at: https://doi.org/10.1016/S0261-3794(03)00020-9Benoit, 2004: 367-‍370). More difficult to identify is the general interest preference explanations according to the party’s preferences for institutional outcomes that affect the general interest in contrast to the partisan interest. Among others, we can find: party preferences regarding the fairness of the electoral system that entails legislative seats for each group in the minority or majority of the population (inclusiveness); party preferences regarding governability with stable executives; and party preferences about choices that encourage conciliation and conflict management in divided societies.

During the processes of democratic transitions the variable of uncertainty, the absence of reliable knowledge about past, present, future, and hypothetical events facing actors engaged in making institutional choice, can be so great as to prevent choices exclusively based on expectations about the future. The inability to make informed choices is common to actors involved in the design of electoral systems during negotiated transitions to democracies. The absence of a relevant electoral record, rapidly changing conditions, and other factors make estimates of the consequences of alternative electoral choices quite unreliable ( ‍Strøm, Kaare. 1990. “A Behavioral Theory of Competitive Political Parties”, American Journal of Political Science, 34 (2): 565-598. Available at: https://doi.org/10.2307/2111461Strøm, 1990: 588-‍589;  ‍Remington, Thomas F. and Steven S. Smith. 1996. “Political Goals, Institutional Context, and the Choice of an Electoral System: The Russian Parliamentary Election Law”, American Journal of Political Science, 40 (4): 1253-1279. Available at: https://doi.org/10.2307/2111750Remington, 1996: 1276). Under circumstances of regime change, political parties develop preferences about the electoral institutions based on office-seeking and policy-seeking explanations. Partisan motivations are usually justified on normative grounds regarding fairness, inclusiveness and governability; but it is very difficult to determine the real importance political parties give to these normative arguments in the negotiation of institutional change.

Consequently, the model identifies two categories of partisan motivations for electoral reform, namely, (a) maximizing the party’s prospective share of legislative seats according to the different electoral rules, that is, motivations based on the office-seeking explanation; and (b) facilitating the most desirable policy outcomes, that is, motivations based on the policy-seeking explanation ( ‍Benoit, Kenneth and Jacqueline Hayden. 2004. “Institutional Change and Persistence: The evolution of Poland’s Electoral System”, Journal of Politics, 66 (2): 396-427. Available at: https://doi.org/10.1111/j.1468-2508.2004.00157.xBenoit and Hayden, 2004: 398). The empirical analysis that follows traces these two categories of partisan motivations in the strategies developed by the political parties involved in the democratic transition, during the process of negotiation of the institutional change in South Africa.

In February 1990, and as part of the transitional process, the NP Government unbanned, ANC, PAC and SACP, while bringing new players to the electoral arena. The extension of suffrage in South Africa came into reality with the census carried out in 1991 that included previously excluded sectors of the population (see table 1). The black African group became the wide majority, representing more than 76 % of the total population^‍‍[16]. This new electoral scenario meant that the NP would become a minority political party associated with an Afrikaner nationalism ideology. Consequently, after the institutional change its political survival was subject to its capacity to maximize its share of seats in the new democratic parliament. The PR system provided for maximum representation, thereby ensuring the inclusion rather than the exclusion of minority parties and opinions. Overall, the idea of proportionality was seen as vital to allay suspicions that the electoral system would unfairly favor one party over another, as it had notoriously happened via the manipulation of constituency boundaries under the majoritarian systems used in South Africa during the authoritarian regime ( ‍Faure, Murray. 1999. “The electoral systems issue in South Africa”, Occasional Papers, Johannesburg: Konrad Adenauer Stiftung.Faure, 1999). The NP made use of normative arguments from the academic debate about the best institutions to develop under divided societies to justify its office-seeking strategy.

The transitional period was accompanied by a massive rise of crime on streets in key centers, much of it being extremely violent. Since the NP Government carried out an ideology of separate development after passing the Promotion of Bantu Self-Government Act (1959), the leaders of the former homelands increased the patronage and power in the administration of their territories. Organizations like the Inkatha Freedom Party (IFP), a direct continuation of the homeland-based Inkatha movement, still overwhelming supported by Zulu speakers located in KwaZulu-Natal, the second most populated province after Gauteng (former Pretoria-Witwatersrand-Vreeniging province). Political violence, mainly between IFP and ANC supporters, fostered an atmosphere of pre-civil war in the middle of the negotiations. The IFP, a Zulu ethnic-nationalist party led by the former homeland leader Chief Mangosuthu Buthelezi, strongly criticized the negotiation process and accused the ANC of playing a dangerous game by projecting itself as the sole authentic black political force^‍‍[17]. This party claimed to boost its 1,7 million of memberships^‍‍[18]. The mistrust atmosphere around the table of negotiations was fueled by the involvement of the NP Government in political violence after the publication of the Inkathagate case^‍‍[19].

During the mid-80s and beginning of the 90s, in an atmosphere of escalation of violence ( ‍Howe, Herbert M. 1994. “‘The South African defence force and political reform’”, The Journal of Modern African Studies, 32 (1): 29-51. Available at: https://doi.org/10.1017/S0022278X00012532Howe, 1994: 29-‍51), the proposals of constitutional engineering in divided societies were thoroughly debated by politicians in South Africa. The debate within constitutional engineering literature between two competing views ( ‍Lijphart, Arend. 1977. Democracy in Plural Societies. New Haven: Yale University Press.Lijphart, 1977,  ‍Lijphart, Arend. 1985. Power-Sharing in South Africa. Berkeley: University of California.1985,  ‍Lijphart, Arend. 1990. “Electoral systems, party systems and conflict management in segmented societies”, in Robert Schrire (ed.), Critical choices for South Africa. Cape Town: Oxford University Press.1990;  ‍Lijphart, Arend and Don Aitkin. 1994. Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945-1990. New York: Oxford University Press. Available at: https://doi.org/10.1093/acprof:oso/9780198273479.001.0001Lijphart and Aitkin, 1994;  ‍Horowitz, Donald. 1985. Ethnic groups in conflict. Berkeley: University of California Press.Horowitz, 1985,  ‍Horowitz, Donald. 1991. A democratic South Africa? Constitutional Engineering in a Divided Society. Cape Town: Oxford University Press.1991) about the best government structures to attenuate racial and ethnic conflict in divided societies became a debate reflected in the political negotiation dialog with two competing approaches. The literature of ethnic divisions found a positive resonance among the NP party and Government and its idea of constitutional arrangements to protect the minorities and ethnic groups in a divided society. The normative arguments to defend constitutionals arrangements for the protection of the South African minorities, were used to justify the NP office-seeking preferences. By contrast, the ANC committed itself to a non-racial society from the its endorsement to the Freedom Charter^‍‍[20] in 1955, and characterized for retaining racial identity as an organizing feature of the electoral and parliamentary system, and the notion of power sharing, as just “being a mechanism to somehow preserve white privilege”^‍‍[21].

The academic debate around the ethnic divisions and party systems has often been focused on the choice between a proportional and a majoritarian electoral system ( ‍Norris, Pippa. 1997. “Choosing electoral systems: Proportional, majoritarian and mixed Systems”, International Political Science Review 18 (3): 297-312. Available at: https://doi.org/10.1177/019251297018003005Norris, 1997;  ‍Reynolds, Andrew. 1999. Electoral systems and Democratisation in Southern Africa. New York: Oxford University Press.Reynolds, 1999). Various proposals have been put forward by academics as a solution to South Africa’s ethnically and racially divided population. Lijphart measured his theory of Consociational Democracy^‍‍[22] against the reality of what would happen there and concluded that an ideal constitution groups to define themselves. And that would be best done through proportional representation in the electoral system^‍‍[23]. The proponents of proportional systems have argued that a party list system of PR is the most suitable electoral system for divided societies because it allows a fair minority representation in the parliament and the consolidation of democracy. Critics of the use of the PR in divided societies as Horowitz, claim that societies driven along a strong ethnic cleavage tend to develop party systems that facilitate the escalation of ethnic conflict ( ‍Horowitz, Donald. 1985. Ethnic groups in conflict. Berkeley: University of California Press.Horowitz, 1985: 291), and conclude that majoritarian systems resist the inherent tendency of the proportional systems and consociational systems to produce ethnic divisions ( ‍Horowitz, Donald. 1991. A democratic South Africa? Constitutional Engineering in a Divided Society. Cape Town: Oxford University Press.Horowitz, 1991: 173).

In response to Horowitz’s ethnic divisions literature, the Minister of Constitutional Development during the political transition, Gerrit Viljoen, admitted that the NP left its earlier position to protect constitutionally defined groups in racial or color terms, to protect groups defined in terms of constitutional arrangements, and finally to arrive to the protection of minorities through special accommodation of minority groups^‍‍[24]. In the same vein, the then President of the Republic, F. W. de Klerk, supported the electoral reform of FPTP, arguing that the Westminster parliamentary model was unworkable in a multiracial society, and its winner-takes-all-system had to be replaced by a fairer system of proportional representation: “Wherever whites are in South Africa they are a minority. Our call is to be part of a greater South African nation. There do not have to be winners and losers. A system in which everyone can win is possible”^‍‍[25]. After 1948, the NP had won decisive legislative majorities^‍‍[26] under the FPTP system, however the new electoral scenario with the extension of suffrage would place this political party as standard bearer of the minority rights, advocating for a proportional electoral system as inclusive as possible that would avoid to waste votes from the minority groups in the country.

The office-seeking strategy followed by the NP was aim to increase the party’s seat share, but the NP party leaders justified its choice by normative arguments based on the academic discussion about the best institutions for divided societies. “The system of representation that we had prior to the change will not work in the new constitution system. Time was a concern, but [...] how be ensured that the minorities will be better represented, that was a general conclusion among the NP but also in the ANC side, PR would provide for a better overall representation for all communities and for all parties “^‍‍[27].

The ruling party's main concern was to design constitutional mechanisms that would protect its future political position, as minority party, in the new electoral arena embracing office-seeking and policy-seeking preferences. “One of the big challenges we had during talks about talks and during the CODESA process was that the Apartheid regime was not really in favor of one person one vote, of equal value […], because at that time the NP still had the hope that we could have so kind of divided country, they still had some kind of hope that we would had a, perhaps, some another kind of minority veto^‍‍[28]”.

The reasoning behind the ANC constitutional guidelines for a democratic post-Apartheid society was based on promoting “habits of non-racial thinking, instill the practice of anti-racist behavior”^‍‍[29]. Grounded in these basic values, it was argued that “the electoral system must encourage cohesiveness rather than parochialism; centrifugal rather than fissiparous tendencies; unity over narrowness in behavior”^‍‍[30]. In the same vein, a ANC Delegate stated “We wanted as many political parties as many diverse political voices to be represented in parliament and the only way you could do that was through the PR system”^‍‍[31]. Nevertheless, a future electoral scenario where small parties could be leading to divisive strategies of the opposition, would erode the possibility of building coalitions. And that was a good prospective electoral outcome for the ANC. Despite the certainty of becoming the winner in a democratic election, the liberation movement was uncertain about its real electoral strength in its first electoral competition. Consequently, the ANC preferred a system that would open for political allies in the parliament and get the legislative support for policies in the process to transform the country.

The first democratic national elections extended the franchise to approximately 16 million black Africans, 76 % of the total population. The expectations and uncertainties about the future competitive electoral arena were supported by reports and opinion polls. The ruling party was confronted by the loss of its political hegemony in the new democratic dispensation. Looking at some of opinion polls released by the media at that time, there was a strong uncertainty among the party’s preference among the new minority communities, white, colored and Indians. While the NP built its electoral strategy support^‍‍[32] on these minorities. Meanwhile, the ANC was confronted by the challenge to turn itself from a liberation movement into an operative political party and take its constituencies with it. That turned to be a massive task with limited personnel, material and administrative resources ( ‍Waldmeir, Patti. 1997. Anatomy of a miracle: the end of Apartheid and the birth of the new South Africa. New York: W. W. Norton and Co.Waldmeir, 1997: 163-‍164).

Some of the polls on party electoral support affiliation published during 1991, had showed the uncertainty in the new electoral spectrum. The ANC itself is becoming increasingly aware of the fact that there are racial or ethnic problems, and this is underlined by the fact that the ANC has succeeded up to now in obtaining the support from the majority of Africans, but not the majority of Indians and coloreds and minimal support from Whites^‍‍[33].

The polls published at that time showed 71 % of urban-based black Africans would vote for the ANC, 3 % for the IFP, 4 % for the NP and 4 % for the PAC, according to a poll conducted in the PWV, Durban, Port Elizabeth, East London, and Cape Town area. The ANC had solid support among the 14 million black African rural population, in areas such as the former Ciskei, Transkei, and border regions (now called Eastern Cape). However, the ANC support from the white group of the population -most of whom lived in urban areas- was low. Only 1 % indicated support for the ANC against the 58 % who said they would vote for NP, 21 % would vote for the right-wing Conservative Party (CP) and 10 % for the liberal Democratic Party (DP). From the 3,3 million colored community there were indications that 49,2 % supported the NP, 10 % or 9,3 % the ANC. Consenquently, uncertainties about the future electoral map were high. “Nobody was able to conduct accurate polls^‍‍[34] until the eve of the elections, early in the 1994. In 1992 the predictions were that ANC would have the majority but just over 50 % and the NP even believed the ANC would make it under 50 %. No opinion polls forecasted the ANC would get the 65 % majority that it did”^‍‍[35].

On the side, the NP could feel comfortable with an electoral system that maximizes its seat share.

Those perceptions on their electoral strengths shaped the strategies and preferences for the different institutional alternatives. Although the opinion polls reflected a wide support for the ANC among the black African voters, this did not translate in a wide support among other groups of voters. A sharp up to 50 % of total voters was a tiny majority that did not bring a full confidence in the future electoral outcome and control of the decision-making process. If the low inclusive FPTP system, used under the settler oligarchy, was retained in the new democratic regime, the electoral future for the two main contenders could have been highly uncertain. By contrast, a highly inclusive proportional system with a high proportional electoral formula, a large district magnitude, a low threshold, and a single ballot structure, brought confidence for the two main contenders involved in the negotiation process. The NP evaluated alternative institutions in terms of the utilities they would derive from their share of distributive goods associated with each institution. The ANC, from its side, secured the future policy outcomes with a system that favors its allies in the Constitutional Assembly and shred the opposition party representation in the legislative institutions. The prospect of retaining the FPTP gave rise to challenges in the process to arrive to a settlement.

South Africa is a racially divided society geographically ( ‍Arenstein, Rowley. 1990. “Multi-nationalism: How Many Nations in the New South Africa?”, Indicator South Africa, 7: 12-15.Arenstein, 1990). The Apartheid regime had created white exclusive residential areas and exclusive residential areas for other racial groups and this meant that the new constituencies would quite likey and de facto become race-based. The negotiation of new electoral constituencies, that is, the boundaries of approximately 400 constituencies, would have been a stumbling block to negotiations^‍‍[37]. The Convenor of the Technical Committee (TC) on the Independent Electoral Commission of the Multiparty Negotiating Process, Richard Rosenthal, confirms the serious concern about the logistics and timing without jeopardizing the settlement:

Another serious concern for the ANC leaders was the new voters who would have the opportunity to exercise their votes for the first time. The easier and simpler the system was, better chances it would have to work. “PR system is a much easier system that the FPTP system because in the PR you can just tell the people to vote for the party, put the picture of the leader of the party, then even people who are illiterate or not very educated will find easy to vote”, concludes Valli Moosa about some of the reasons behind the ANC support for a PR system^‍‍[39].

The electoral formula chosen combined with a low threshold opened the legislative door to seven political parties out of the 19 ones that contested the elections. ANC won the first democratic general elections with 62.6 % of votes, that translated into 252 seats out of 400. The ruling party, NP, reached 20.4 % of votes and got 82 seats. The new democratic Constitutional Assembly in charge of drafting the final constitution had to review the interim electoral law^‍‍[40]. During the debates on the electoral system, all the parties expressed their concerns that their representatives were a ‘little bit distant’ from the voters and they needed to find a way of marrying PR with a constituency system that would bring the government closer to the people^‍‍[41]. However, the final outcome of the Constitutional Assembly did not result in any major change in the interim electoral law because there was not consensus on this matter. There was no unanimity among the political submissions, or among individual submissions on this issue^‍‍[42].

South Africa presents a one-party dominance system^‍‍[44], despite its almost perfect proportionality (see table 2). During the first democratic legislature, 1994-‍1999, opposition politics in South Africa developed in a multipolar direction, rather than along purely race-based^‍‍[45] lines ( ‍Barkan, Joel. 2005. Emerging Legislature or Rubber Stamp? The South African National Assembly After Ten Years of Democracy. CSSR Working Paper, 134. Rondebosch, South Africa: CSSR.Barkan, 2005: 4). Upon withdrawing from the GNU in may 1996, F. W. de Klerk argued that his party would consolidate its position^‍‍[46] as a party based on values rather than ethnic affiliation ( ‍Reynolds, Andrew. 1999. Electoral systems and Democratisation in Southern Africa. New York: Oxford University Press.Reynolds, 1999: 50). Ethnic and racial parties drastically declined in both number and significance^‍‍[47]. The major ethnic parties, DP, NP, and IFP, attempted to shift their appeals into an ideological or multiethnic tier during the first ten years of democracy, and many of the smaller, purely ethnic parties, ceased to exist as credible parties ( ‍Piombo, Jessica and Lia Nijzink. 2005. Electoral politics in South Africa: Assessing the first democratic decade. New York: Palgrave Macmillan. Available at: https://doi.org/10.1057/9781403978868Piombo and Nijzink, 2005: 67). The Zulu ethnic-nationalist party, IFP, had declined in the number of votes and seats in the National Assembly (see table 2). In 1999, the NP, was renamed New National Party (NNP), was reduced to the third position. And in 2004, the NNP’s electoral performance was disastrous. The NNP’s involvement in a number of coalitions^‍‍[48] confused its image among voters. The NNP turned away from the ethnic tie of Afrikaner mobilization, in favor of a social conservative approach based on community rights and Christian values. The NP/NNP that had dominated South Africa politics for 50 years by September 2005 had dissolved.

Six years after the ratification of the Constitution of the Republic of South Africa, the ANC Government established an Electoral Task Team in 2003 (ETT) to review the electoral system and suggest whether or not any changes should be made. The ETT considered three types of electoral systems: (a) the current party-list system of PR; (b) a mixed system with single-member constituencies balanced by proportional elements; (c) and a constituency system, including single and multi-member constituencies to ensure proportionality. Finally, the Cabinet decided to retain the current electoral system. The argument used by the ANC to reject the ETT recommendations came from its NEC meeting in July 2002, the country still needed “to harness our inclusive political system in the interest of nation-building and national unity” (Ibid.: 76). Since then, the calls for an electoral reform have been initiated primarily by academics and the civil society, not by political parties and not by the present ruling party, the ANC: “For a long time to come, it is not going to be reformed (electoral system), certainly for another generation will not be reformed and that simply because it works very well for us”^‍‍[49]. An important aspect of stability of electoral institutions is that political actors who have the power to change them are those who benefit enough from the existing electoral system to hold seats in the legislature ( ‍Bawn, Kathleen. 1993. “The Logic of Institutional Preferences: German Electoral Law as a Social Choice Outcome”, American Journal of Political Science 37 (4): 965-989. Available at: https://doi.org/10.2307/2111539Bawn, 1993: 987). Consequently, the electoral system will not reform as long as it suits the office-seeking and policy-seeking preferences of the main political actors.

The NP during the Apartheid regime was against any reform^‍‍[50] of the electoral system that would had reduced their own legislative share of seats, instead they favored the FPTP system arguing that was a general interest for the governability factor in the political system ( ‍Faure, Murray. 1999. “The electoral systems issue in South Africa”, Occasional Papers, Johannesburg: Konrad Adenauer Stiftung.Faure, 1999: 2). The introduction of the universal suffrage in South Africa entailed the transformation of the ruling party under the authoritarian regime to a minority party in a future democratic regime. At that stage, because the NP needed to maximize its seat share, instead to favor the retaining of the old electoral regime, the NP shifted its position of defending the governability factor that a FPTP system produces, to defend the inclusive factor that a PR system produces. In both cases, against and for the reform of the electoral system, the NP used normative arguments to favor their proposals despite that they were moved by office-seeking strategies.

If the FPTP system was retained in the country, the delimitation of 400 single-member districts would have blocked the negotiation process, and “would have to argue for the next year or two on how to draw the map of boundaries and time was of essence”^‍‍[51]. The ANC was concerned with the map of boundaries of districts and the possibility that NP could manipulated it. The design and implementation of the last census had revealed in evidence that the NP used any opportunity for maximizing its political outputs. Finally, the ANC chose an electoral system that maximized its allies’ seat share in the Constitutional Assembly, SACP, as well as small parties on the basis of the reform dimension’s alignment of MPs’ policy preferences. At the same time, a system that minimizes the likelihood that its main rival, NP, would increase the number of seats and its chances of leading an opposition coalition and shaping policy. In addition to the party list system -that allows party leaders to exercise a control over their electoral organizations by controlling the placement of candidates on the party list. An attribute that suited an ANC in an embryonic state of organization inside the country after almost half century in the exile.