Afr. j . polit. sci. (1997), Vol. 2 No. 1, 70-8 Elections for Parliament and the Provinces in the New South Africa: 1994-2004 Murray Faure* Abstract This article discusses the arrangements for the election of parliament and the provincial legislatures in the new South Africa. Present electoral provisions in terms of the Interim Constitution of 1993, as well as the prospects for the elections of 1997 and beyond, in terms of the Final Constitution, are covered. The article also reviews the results of the 1994 general election and offers a general assessment of South Africa's electoral provisions. Introduction This article provides an overview of arrangements for the election of parliament and the provincial legislatures in the new South Africa. In addition, it briefly reviews the outcome of the April 1994 election.The electoral arrangements took effect with the adoption of the 1993 Interim Constitution and represent a major break with the preceding political and electoral dispensation. The election of April 1994 institutionalised a new Parliament and nine new provincial legislatures which were elected for a period of five years. The 1996 "Final" Constitution effected important changes to these arrangements, but these changes will not apply in the 1999 election. The transitional arrangements provide for the retention of the electoral system used in 1994; the new Electoral Act, which will provide for an electoral system in terms of the principles laid down in the Final Constitution, is yet to be drafted and promulgated. The overview covers events that led up to the adoption of the Interim Constitution, and, more specifically, the electoral debate which structured and influenced the process. It secondly outlines the features of the electoral system used in 1994 and also highlights the results which the system yielded. In the third section, the electoral provisions of the Final Constitution are examined, and the electoral 1027-0353 © 1997 African Association of Political Science Elections for Parliament and the Provinces in the New South Africa 71 prospects for 1999 and 2004 are assessed. The article is concluded with a general assessment of the electoral arrangements in place, and the changes that may occur in the future. Background to The Interim Constitution Before the unification of South Africa in 1910, a variety of elementary electoral arrangements existed in the Cape Colony, the Colony of Natal and the Republics of the Orange Free State and Transvaal (cf. du Places, 1979 : 49-80; Nieuwoudt, 1966 : 71-80, 1979 : 23-48 and Faure, 1996.a : 89). With unification these were consolidated into a single-member geographical constituency system that functioned on the basis of relative majorities. This system, which was used for the election of members of Parliament and the four Provincial Councils, remained essentially unaltered until 1993. The essentials of the system were also used for elections in the so-called self-governing homelands, as well as the TBVC states — i.e., Transkei, Bophuthatswana, Venda and Ciskei, the so-called "independent" homelands. The three-chamber parliament of 1983 (a distorted application of Lijphart's notion of consociational democracy) also embodied this system of constituency representation (cf. Basson, 1984; Lijphart, 1985; Thompson, 1986 : 12; Faure, 1987 : 3-23, 1988 : 143-169 and 1996.a : 90). Apart from the fact that the constituency system either completely or partially excluded the greatest majority of the population from effectively exercising the right to vote, its main feature was that it severely distorted the ratio between votes cast for, and seats obtained by political parties (cf. Butler, 1983 : 53; Mackie and Rose, 1974 and Faure, 1996.a : 90). The decade that preceded the adoption of the Interim Constitution in 1993 was characterised by a dramatic deepening of the political crisis in the country. Political violence escalated to unprecedented heights and the National Party government was increasingly isolated internationally. Towards the end of the 1980' s it became apparent that major political changes were unavoidable. On 2 February 1990, President de Klerk delivered his now famous speech to Parliament, while Nelson Mandela was released from prison shortly afterwards. The four years that led up to the election of 1994 witnessed the negotiations of Codessa I and II, the debates in the Technical Committee on Constitutional Issues and its interface with the Multiparty Negotiation Council. These debates and negotiations, among other issues, also brought the electoral debate into sharper focus and underlined the growing consensus that electoral reform towards some form of proportional representation (PR) was desirable. Both academics and political actors stated (or restated) their views in this regard (Faure, 1996a : 90-91). Some of the more important events that structured the process of electoral reform are the following (cf. Faure, 1994a : 110-115; 1996a : 91-92). In 1990 contact was made between government and the Electoral Reform Society in the 72 Murray Faure UK. The Society offered to provide expert advice in electoral reform. In Cape Town a conference of the African National Congress (ANC) on electoral matters opted for PR (UWC/CDS, 1990). In the same year, the Committee on Constitutional Affairs of the President's Council underlined the advantages of PR for a future dispensation (President's Council, 1990; cf. Also Faure, 1996.a : 91). During 1991 the Centre for Constitutional Analysis of the Human Sciences Research Council (HSRC) (de Villiers, 1991), the Democratic Party's (DP) Congress (Besprekingsdokument, 1991 : 1-3), and the National Party (NP) in a policy document (National Party, 1991) expressed their preference for some form or the other of PR (cf. Faure, 1996.a: 91). In the course of 1992 three preferences in favour of one or the other form of PR were expressed. These include the ANC's Guidelines for a Democratic South Africa adopted at its National Conference in May (ANC, 1992:3-4), a second report by the President' s Council entitled "Report on a Proportional Polling System for South Africa in a New Constitutional Dispensation" (President's Council, 1992 : 85-86), and Codessa IPs proposal that a 400-member Parliament be elected on a strict proportional basis, one half of the members being returned on a regional basis and the other half elected on the basis of a national list. In this year the government also announced its timetable that would lead up to the election of 1994 (cf. Faure, 1996.a: 90-91). 1993 saw a number of academics explore the merits and demerits of PR for South Africa, notably Frost (1993), Reynolds (1993), Johnston (1993) and Sisk (1993). It was also the year in which the multiparty negotiation process culminated with a set of recommendations on April 1. Following months of further negotiation, the Interim Constitution was finally adopted at the plenary session of the Negotiation Council on November 17. On 22 December 1993, Parliament ratified the Interim Constitution together with related legislation such as the Electoral Act and enacted it as the supreme law of the land effective April 1994 (cf. Faure, 1994.a: 114and 1996.a: 92). The first four months of 1994 were devoted to further negotiations with the aim to achieve an inclusive and democratic election. A major change to the electoral arrangements contained in the Interim Constitution was the decision to grant each voter two ballots — one for the purpose of electing the central legislature, and the other for the purpose of electing the provincial legislatures (Faure, 1994.a : 114). Elections Under The Interim Constitution: The overview of the electoral system used in the April 1994 election is based on the Electoral Act, and more specifically Schedule 2 of the 1993 Interim Constitution. The Electoral System South Africa's Parliament is composed of a 400 member National Assembly and a 90 member Senate which is indirectly elected by parties in the provincial Elections for Parliament and the Provinces in the New South Africa 73 Parliament: Senate 90 Indirectly elected members; 10 per province nominated in terms of the relative strength of parties represented in the provincial legislatures National assembly 400 Members 200 Members elected on a national basis on the basis of a party list 200 Members elected on a regional basis based on a regional party list 21 Seats North West 15 Seats 28 seats Northern Transvaal 20 Seats 4 seats Eastern Transvaal 14 Seats 40 Seats Gauteng 43 seats Western Cape Eastern Cape Northern Cape kwa-Zulu Natal Orange Free State 15 Seats Total 200 Seats Provincial legislatures Western Cape Eastern Cape Northern Cape Kwa-Zulu Natal Orange Free State North West Northern Transvaal Eastern Transvaal Gauteng 42 Seats 52 Seats 30 Seats 80 Seats 30 Seats 30 Seats 40 Seats 30 Seats 86 Seats legislatures. 200 members of Parliament are elected using national party lists, while the other 200 are elected on the basis of provincial party lists with a fixed number of representatives for each province. This ratio also reflects the population distribution in the various provinces. The nine provincial legislatures are elected on the basis of regional party lists; the size of each legislature also reflecting the population size of the province, with the proviso that such a legislature shall have a minimum of 30 members. 74 Murray Faure Members of Parliament and the legislatures of the various provinces are elected in terms of a Continental-like system of proportional representation using closed party lists. Though not as complex as some systems of proportional representation, the "average" voter does not really seem to understand its technical functioning, especially the calculation and allocation procedure of seats according to different formulas. In terms of the criteria that normally differentiate proportional representation systems from each other, the current system displays the following features (cf. Faure, 1994.a : 115-121; 1994.b : 31-50; 1996.a : 92-94; 1996.b : 196-199): It uses both national and sub-national (regional/provincial) party lists. In a certain sense, the election of 200 National Assembly members according to regional party lists make the nine regions/provinces extremely large multi-member constituencies. The election of 200 National Assembly members according to national party lists, like the method used in the Netherlands and Israel, uses the whole country as one very large multi-member constituency. The system allocates seats proportionally at both regional and national levels. While the regional allocation is not the conventional allocation in terms of small multi-member constituencies (but in terms of larger regions/provinces), the national allocation has to take account of regional allocation. Smaller parties with regional support are not disadvantaged by this method, while smaller parties with a country-wide level of support (i.e. not regionally based support) can benefit from this method of national allocation. The threshold, i.e. the minimum votes required to win a seat is extremely low. In fact, some thresholds in the system are of the lowest in the world. The threshold (which is, of course, mathematical and not legal in nature) varies for the regional and national lists respectively, and those of the provincial legislatures are about half the size of those required for the regional allocation in the National Assembly. In the case of the Gauteng (formerly Pretoria/ Witwaterstrand/Vereniging) region/province, for example, the threshold for the National Assembly is approximately 2.27%, while the threshold for this region's provincial legislature is approximately 1.14%. In contrast, the threshold for the Northern Cape province in the National Assembly is 25%, while that for the provincial legislature of that province is 3.33%. For both the National Assembly and the regional/provincial legislatures the respective thresholds of the various regions/provinces differ, unless they have the same number of seats respectively. Approximately l/400th of the votes cast for the national party lists of the National Assembly (i.e. about 0.25%) constitutes the threshold, but the number of seats already allocated regionally is subtracted from the seats won in this way, effectively making this threshold about 0.50%. The threshold for the National Assembly as a whole is 0.24938%. Whereas the ballot in some proportional electoral systems allows voters to choose between various candidates (even across party lines if there is more than one ballot), South Africa's present electoral system does not provide for this. Voters are only accorded the opportunity to choose between competing Elections for Parliament and the Provinces in the New South Africa 75 party lists. The voter is bound to the order of candidates as decided by the parties. No candidate preference is provided for, and the only differentiation that voters can exercise is to vote for different parties on the national and regional levels. The Election Results The election of 27 April 1994 was South Africa's first open and free democratic election. In spite of fears to the contrary, the election itself was for the most part free from violence, and it has since gained symbolic significance as the event that brought democracy and liberation to the country. The electoral process itself was marred by many technical problems which the Independent Electoral Commission had to deal with, but as a whole the outcome and its symbolism overshadowed the practical impediments of the process. The electoral system yielded the following results (Tables l-4:Pienaar, 1994; DP Newsletter, July 1994: Constitutional talk, 1995 : 4 and Sunday Times, May 8 1994 : 7. Table 5: de Ville, 1996 : 26. Tables 6-8: Botha, 1996:115-117). Table 1: Estimated Voters and Seat Allocation per Province Province Voters/ 1000 Western Cape Eastern Cape Northern Cape Kwa-Zulu Natal OFS 2,406 3,177 439 4,585 1,637 1,730 2,287 1,586 4,862 Northwest Northern Tvl1 Eastern Tvl2 Gauteng Total 22,709 Table 1 above indicates that almost 23 million voters participated in the election. Gauteng has the highest number of voters, while the Northern Cape has the smallest number of voters. The proportionality of the system is best reflected by the number of seats that each province is entitled to in the National Assembly. Provinces enjoy equal representation in the Senate — a feature typical of federal bicameral arrangements. In terms of the relative sizes of provincial legislatures, the smaller Prov. National Senate Seats Assembly Seats Seats 21 28 4 42 56 30 40 15 15 20 14 10 10 10 10 10 10 10 10 10 43 81 30 30 40 30 86 90 200 425 76 Murray Faure Table 2: Election results — National Assembly and Senate Party Votes/ 1000 ANC NP IFP FF DP PAC 12237.6 3983.7 2058.3 424.6 338.4 243.5 88.1 12 ACDP Remaining Spoilt Papers 147.8 Total 19681.4 provinces are over-represented due to the minimum size of these assemblies which is set at 30 members (Faure, 1996.b : 199-200). Table 2 shows that the election was dominated by one party, the ANC, securing close to 63% of votes cast, while three paries (the ANC, NP and IFP) collectively drew more than 90% of the vote. In terms of the regional party lists of the National Assembly, the African National Congress (ANC), the National Party (NP) and the Inkatha Freedom Party (IFP) won 128,42 and 22 seats respectively, with 3,4 and 1 seat(s) going to the Democratic Party (DP), Freedom Front (FF) and the Pan- Africanist Congress (PAC). In terms of the national party lists of the National Assembly, the ANC, the NP and the IFP won 124, 40 and 21 seats respectively, with 4,5 and 4 seats going to the DP, FF and the PAC (de Ville, 1996: 21,23). The three large parties did better in terms of the "national" component of the National Assembly than in its "regional" component; the reverse is true for the smaller parties. All of them obtained more seats in the "national" component than in the "regional" component. The African Christian Democratic Party (ACDP), for example, won 2 seats in terms of national lists, and none in terms of the regional lists. This underlines the fact that the system does not disfavour small parties which enjoy country-wide, rather than regional support. Seven parties secured seats in the National Assembly, while only five parties are represented in the Senate — with the ANC enjoying a two-thirds majority (Faure, 1996.b : 200). Percentage Senate Seats National Assembly Seats 60 252 17 82 5 43 62.6 20.4 10.5 5 2.2 1.7 1.3 9 7 5 2 0.5 0.7 169.3 3 — — — — — 0.7 90 400 100.0 Elections for Parliament and the Provinces in the New South Africa 11 Table 3: National versus Provincial Votes National Totals Party 12,237,655 3,983,690 2,058,294 ANC NP IFP FF 424,555 338,426 DP An interesting pattern is shown by Table 3 above. It indicates the degree of differentiation with regard to the two ballots. The smaller parties drew significantly more support at the regional than the national level. The reverse is true for the large parties. Table 4: Provincial Strength of First 4 Parties Per Province (%) Province Western Cape Eastern Cape Northern Cape Kwa-Zulu Natal OFS Northwest Eastern Transvaal Gauteng NP(53.2) ANC (84.8) ANC (49.7) IFP (50.3) ANC (76.6) ANC (83.3) Northen Transvaal ANC (91.6) ANC (80.7) ANC (57.6) Table 4 indicates the provincial strength of parties. The ANC has the majority of popular support in all the provinces except for the Western Cape and Kwa-Zulu Natal where the NP and the IFP have the most support, respectively (Faure, 1996.b: 201-202). Provincial Totals 12,137,307 3,492,467 2,047,083 639,643 538,655 2 1 ANC (33.0) NP (9.8) NP (40.5) ANC (32.2) NP(12.6) NP(8.8) NP(3.3) NP (9.0) NP (23.8) Difference -100,348 -491,223 -11,211 +215,088 +200,229 4 3 FF(2.1) PAC (2.0) DP (1.9) DP (2.2) PAC (1.8) PAC (1.7) PAC (1.3) PAC (1.6) DP (5.3) DP (6.6) DP (2.1) FF (6.0) NP(11.2) FF (6.0) FF (4.6) FF(2.1) FF(5.7) FF (6.2) 78 Murray Faure Table 5: Seats For Parties Compared to their Votes at National Level (%) Party ANC NP IFP FF DP PAC ACDP Table 5 is a dramatic illustration of the degree of proportionality attained by the electoral system. It shows the correlation between the percentage of votes won by the different parties at the national level and the percentage of seats they won (de Ville, 1996 : 25-26). In some instances, this correlation is almost perfect and a far cry from the previous system where by during the 1960's and the 1970's the ruling National Party, on average, recovered about 50% of the vote, but won approximately 75% of the seats (Faure, 1996.a : 90). Table 6: Strength of Political Parties in Terms of Seats Won National Parliament Northern Transvaal Eastern Transvaal Gauteng North West Orange Free State Kwa—Zulu Natal Eastern Cape Northern Cape Western Cape Table 6 provides an overall comparison of the number of seats that each party won in the central and provincial legislatures — a total of 825 seats. Only the ANC and the NP enjoy representation in all of the assemblies. % of Seats % of Votes 63 62.65 20.5 20.39 10.75 10.54 2.25 1.75 1.25 0.5 2.17 1.73 1.25 0.45 NP ANC IFP FF 252 43 82 — 38 1 25 3 — 50 21 26 24 3 — — 26 41 48 — 3 4 9 6 15 12 — 14 9 1 2 5 1 2 — — 2 1 23 — MF PAC ACDP DP 7 — — 5 — — 1 5 — — — — — — — — — 1 2 — 1 1 — 1 1 — — 2 — — 1 — — 1 — — 1 — — 3 Elections for Parliament and the Provinces in the New South Africa 79 Table 7: Voter Participation in the Provinces % voted Province Northern Transvaal Eastern Transvaal Gauteng (PWV) North West Orange Free State Kwa-Zulu Natal Eastern Cape Northern Cape Western Cape National 84 85 86 89 83 80 92 92 87 86 Table 8: Estimated Composition and Turnout of the Electorate By Race Race No. of Registered as voters (millions) Blacks 16.6 Whites 3.4 2.0 Coloureds Indians 0.7 22.7 Total Tables 7 and 8 show the high level of voter participation in the election. Participation was truly country-wide; and not a single province registered a voter turn-out lower than 80%. This high participation cut across racial lines. It is equally remarkable that a low percentage of spoilt ballots was recorded which underscores the fact that the ballot itself was comprehended in spite of the fact that the electoral system is rather complex. % of ballots spoilt 0.7 0.9 0.6 1.2 0.8 1.0 0.5 0.8 0.5 1.0 %of electorate 73 15 9 3 100 % that Number that voted (millions) voted 86.7 14.4 85.3 2.9 90.0 1.8 85.7 0.6 86.0 19.7 80 Murray Faure Elections Under The Final Constitution The constitutional principles The drafting process of South Africa's Final Constitution began shortly after the election of 1994. It is important to understand that the Interim Constitution in its Schedule 4 contained 34 principles which were deliberately designed to structure the writing of the Final Constitution and shape the political society which it would create. These principles can be viewed as political values intended to guarantee a democratic society. With regard to the drafting of the Final Constitution these principles represented the axiomatic framework against which the contents of the new constitution was to be measured and tested. They were viewed as nonnegotiable ("cast in stone"), and covered a wide range of constitutional elements. Furthermore, they were formulated in a rather general way which left (and still leaves) considerable scope for interpretation with regard to specific formulations that were encapsulated in clauses of the Final Constitution, the new Electoral Act still to be drafted and promulgated, as well as the testing of these in terms of the original formulations in Schedule 4 by the Constitutional Court. The references to electoral matters in Schedule 4 are mostly indirect and are spread over a number of principles, notably I, IV, VIII, XIV and XVII. These principles state the following:3 I. The Constitution ... shall provide for ... a democratic system of government committed to achieving equality between men and women and people of all races. IV. The Constitution shall ... be binding on all organs of state at all levels of government. VIII. There shall be representative government embracing multi-party democracy, regular elections, universal adult suffrage, a common voters' roll, and in general, proportional representation. XIV. Provision shall be made for the participation of minority political parties in the legislative process in a manner consistent with democracy. XVII. At each level of government there shall be democratic representation. This principle shall not derogate from the provisions of principle XIII (traditional leadership). Except for principles VIII and XIV, the references to electoral matters placed very few restrictions on the nature and type of electoral system that the Constitutional Assembly was to decide upon. Principle VIII, however, is probably the most Elections for Parliament and the Provinces in the New South Africa 81 controversial with regard to interpretation. The crucial question that is raised by this principle is what is meant by "There shall be ... in general, proportional representation?" If it is read in conjunction with principle XIV which states that "Provision shall be made for the participation of minority parties in the legislative process in a manner consistent with democracy", its interpretation becomes more difficult. The crux of the matter seems to be the following: Should the interpretation of this principle be that there can only be a system of proportional representation, or could the interpretation also accommodate an electoral system that is not based on proportional representation (e.g. a majority system), but, which does yield a high proportional outcome? With regard to the latter, scientifically based proportionality indices have shown that, in general, systems of proportional representation yield higher proportionality than majority systems, but that certain majority systems like the ones used for the House of Representatives in the US Congress and the Japanese House of Representatives yield higher outcomes of proportionality than some proportional systems do.4 These systems also ensure minority party representation in the legislature. A reasonable and tenable interpretation of this must be that the principles in question can provide for both the above interpretations (i.e. categories of electoral systems); and even a combination of them, the justification being: (1) That principle VIII uses the adjective "proportional" to describe the nature of representation (noun). It does not prescribe a type of electoral system belonging to the category of electoral systems which is generally known as proportional system of representation. The use of the phrase "in general" seems to strengthen the notion that there is no reference to a specific type of electoral system, but rather to the nature of electoral representation. (2) That there is no clear dividing line between those systems which are generally known as systems of proportional representation and some other systems which closely resemble the features of the former, which are not so called, but which do yield high proportionality and also allow for minority parties to play a role in the legislative process. (3) That there is no clear definition of how small a party can be before it is so insignificant that it should not participate in the legislative process. In this sense, principle XIV is probably more significant with regard to the internal procedures of the legislature itself than to the requirement that minority parties irrespective of size should receive representation. The constitutional principles therefore required that the Final Constitution's electoral provisions be proportional in nature, and not necessarily that a system of proportional representation be used. Electoral provisions The 1996 Final Constitution contains no detailed prescriptions with regard to the 82 Murray Faure electoral system(s) that should be used for election to the central and provincial legislatures of the state. This is in contrast to the 1993 Interim Constitution (Schedule 2) which specified in great detail the procedures that were used to elect the central and provincial legislatures in the election of April 1994. The Final Constitution provides for the election of these legislatures in terms of a number of principles only. Simultaneously, however, the Final Constitution contains a number of transitional arrangements which, among others, specify that the next general election which falls in 1999 must be conducted in terms of the electoral system specified in Schedule 2 of the Interim Constitution. The implication of this is that the election of 1999 will use the same system as the one used in 1994, and that the principles in the Final Constitution will only take effect in the general election of 2004. With regard to the National Assembly these principles state that elections shall be conducted in terms of an electoral system that (a) is prescribed by national legislation; (b) is based on the national common voters roll; (c) provides for a minimum voting age of 18 years and (d) results, in general, in proportional representation. The National Assembly will consist of no fewer than 350 and no more than 400 members. With regard to the provincial legislatures the same principles apply to the electoral system that is to be used, and in addition, such a system must be based on a particular province's segment of the national common voters roll. The exact number of members of the National Assembly and the respective provincial legislatures, which may differ among the provinces, will be determined by a formula prescribed by national legislation. A provincial legislature shall, however, not have fewer than 30 and not more than 80 members (cf. Constitution Chs. 4 and 6, 1996 and Faure, 1996.d : 65 ). The Final Constitution also provides for a second Chamber of Parliament which is known as the National Council of Provinces. Premiers of the respective provinces will head each provinces' 10 delegates to the Council. Each delegation will consist of 4 special delegates as well as 6 permanent delegates. The delegates in a provincial delegation to which a party is entitled, will be determined by multiplying the number of seats the party holds in the provincial legislature by ten and dividing the result by the number of seats in the legislature plus one. If a calculation in terms of this formula yields a surplus not absorbed by the delegates allocated to a party, the surplus must compete with similar surpluses accruing to any other party or parties, and any undistributed delegates must be allocated to the party or parties in the sequence of the highest surplus (cf. Constitution, 1996 Ch. 4 and Faure, 1996.d : 65-66). The Constitution is much more specific and prescriptive with regard to the election of Municipal Councils. For these a system of proportional representation based on a municipality's segment of the national common voter's roll must be used. Such a system may be combined with a system of ward representation, and Elections for Parliament and the Provinces in the New South Africa 83 national legislation can furthermore provide for the appointment of members from another Municipal Council. Several clauses of the Constitution are devoted to matters such as fair ward delimitation and the representativeness of election outcomes (cf. Constitution, 1996 Ch. 7 and Faure, 1996.d : 66). In addition, the Constitution provides for an Electoral Commission, along with a number of other institutions, to strengthen constitutional democracy. The Electoral Commission will manage elections of national, provincial and municipal legislative bodies in accordance with national legislation; ensure that those elections are free and fair; and declare the results of those elections within a period that will be prescribed by national legislation and which will be as short as reasonably possible. The Electoral Commission will be composed of at least three persons and the exact number as well as its powers will be determined by national legislation. The process for instituting the new Electoral Commission has already begun, (cf. Constitution, 1996 Ch. 9 and Faure, 1996.d : 66). Assessment How should these new electoral arrangements be assessed when compared to those contained in the Interim Constitution and used in 1994? At present (April 1997) there is still no indication of what type of electoral system will be embodied in the new Electoral Act for elections after 1999. The fact that the electoral system of 1994 will be used for the 1999 election, apparently without significant changes, must be decried. This decision is the result of a deal struck between the ANC and the NP at the Arniston bosberaad (Calland & Merton, 1996 : 1, 3). It also includes the retention of the much-maligned Section 43(b) of the Interim Constitution which prevents MP's from crossing the floor of the house to change party loyalty. It is, however, significant to note that President Mandela was recently reported to have said that the desirability of PR for South Africa, as well as the prohibition on MP's from crossing the floor needs to be re-examined (Faure, 1996.d:66 and SABC, 16 April 1997). The shortcomings of the present system of PR with closed party lists have been extensively debated in Theme Committee 25 of the Constitutional Assembly as well as in academic circles (cf. De Ville & Steytler, 1996 and Faure, 1996.b : 201-202). While it is strong on proportionality, it is especially weak on voter-accountability. It has even been suggested that the lack of a kind of accountability in the present system may, in fact, be a breach of Principle VI of the Interim Constitution which requires that there should be accountable government (Calland & Merton, 1996 : 3). The^ present system makes MP's responsible to parties only, giving party bosses excessive power in this regard. Whether that is the case or not, the fact is that South Africa will use the present system with its shortcomings again in 1999. The practice of post hoc allocation of "informal constituencies" or "interest regions" to MP's will probably be followed again, but 84 Murray Faure it is a poor substitute to real constituency representation and proper accountability to voters (Faure, 1996.d : 66). The fact that the Final Constitution does not rigidly fix the details of the electoral system(s) to be used for the central and provincial legislatures in 2004 should be welcomed. The Constitution only pronounces on the results that future arrangements must yield, i.e. it/they should result "in general, in proportional representation". Future national legislation in the form of an Electoral Act must determine the nature of the system(s) to be used and the only requirement that the Constitution sets in the preceding phrase is that the system(s) should yield results that can stand the test of fairness, openness and especially representativeness. This is a pragmatic way of dealing with electoral matters. Rather than rigidly fixing the nature of electoral systems in the Constitution, future national legislation can now adapt to experience-in-use, changing needs, changes in demography, and so on. It is also important to note that the requirements stated in the Constitution do not preclude the use of so-called non-proportional electoral systems in conjunction with proportional systems. Since the emphasis is on outcomes that a system yields, it should be repealed that some so-called non-proportional electoral systems can yield higher indices of proportionality than proportional systems themselves. The type of future electoral system(s) is therefore still unresolved and there is an obligation on everyone to campaign for a better dispensation than the pure party list system that will be used again in 1999 (cf. de Ville & Steytler, 1996). Serious consideration should be given to the re-introduction of constituency representation in conjunction with the party list system in order to ensure voter-accountability along with proportionality. Such an arrangement is well demonstrated by the German model (Faure, 1996.d : 67). It is significant to note that a survey carried out by the Parliamentary Information and Monitoring Service during 1994 found that overwhelming support existed among two-thirds of all the parties for a mixed electoral system that included an element of constituency representation. Only the ACDP and the NP registered less than 50% support for such a mixed system. Against this background the retention of the present system could be explained by reports that the ANC is still hesitant to enter such a new dispensation since the majority of its constituency candidates would exclude Whites, Coloureds and Asians. It is also reported that many ANC candidates are not yet ready to take on the added responsibility of formal constituency representation and that members were still on a sharp learning curve (Calland & Meiton, 1996 : 3). Conversely the NP's hesitance for a mixed system is probably to be found in its fear that most constituency candidates that would be elected may be non-White (Faure, 1996.d : 67). With the year 2004 in view, it is indeed a pity that the Constitutional Assembly did very little to reduce the size of the National Assembly and some of the provincial legislatures. The indefensibility of such large legislative institutions Elections for Parliament and the Provinces in the New South Africa 85 have been argued at length elsewhere (Faure, 1996.c : 71-75), and one can only infer that parties in the drafting and negotiation process have retained the large numbers of legislators in their own interests. The National Council of Provinces is a reincarnation of the present Senate, albeit with the exception that representatives from the respective provinces now comprise two categories: the special delegates and the permanent delegates. The former category, which will be headed by the provincial premiers is clearly an attempt to get the most influential spokes-persons of the respective provinces to articulate provincial interests in the Council. Whether this measure will bring new life and purpose to the Council as compared to the mundane role of the previous Senate remains to be seen, but, it definitely will not succeed if the Council does not use its powers to the fullest possible extent (Faure, 1996.d : 67-69). If electoral arrangements for the central and provincial legislatures can be dealt with in an Electoral Act, one could justifiably ask why the arrangements for local government are contained in the Final Constitution. The detailed arrangements are more or less in line with those used in recent local elections during 1995. The relative complexity of these electoral arrangements did not seem to present insurmountable problems and by 2004 the electorate will have had the advantage of experience-in-use of the system (Faure, 1996.d : 68). Conclusion South Africa is approximately less than two years away from its second general election under a new democratic dispensation. Two major challenges exist. The first is to ensure that the election of 1999 is timeously organised. There have already been warnings by Dr. Frederik van Zyl Slabbert and Judge Johan Kriegler on behalf of the Electoral Commission, as well as the Minister for the Interior. Chief Buthelezi, that the time-frames have become very short indeed. Open, free and fair elections are the fountains of legitimacy in all democracies, and South Africa's next general election will not be overshadowed by the symbolic significance which characterised the previous election. It is imperative that the process should run smoothly, and that all the arrangements pertaining to the electoral process be made well in advance. The mistakes of the past election cannot be afforded in 1999. The second challenge is to give effect to the principles of the Final Constitution by embodying a new electoral system in the new Electoral Act. South Africa can justifiable be proud of the system that it has, which it is a major improvement on the old winner-takes-all constituency system that operated on the basis of relative majorities. As mentioned previously, that system grossly distorted voter support when it was translated into seats won by the different parties. However, the present system has a number of shortcomings which can and must be improved upon, especially the lack of voter accountability. All those involved in the process of re- 86 Murray Faure drafting the system have a responsibility to transcend personal and party interests and give South Africa the best possible electoral system that will suit the country's needs. Notes * Professor of Political Science and Head of the Department of Political Sciences, University of South Africa, Pretoria. 1. Also known as the Northern Province. 2. Also known as Mpumalanga. 3. Cf. Schedule 4 — Interim Constitution, 1993. 4. Compare the indices of Rose. (Elections and electoral systems: choices and alternatives) in Bogdanor, & Butler, (1983 : 41). 5. The Constitutional Assembly sub-divided itself into a number of Theme Committees — 1, 2, 3, 4, etc.; each with specific tasks with regard to the drafting of the Final Constitution. References ANC policy guidelines ..., 1992. AN C policy guidelines for a democratic South Africa. As adopted at National Conference, 28-31 May 1993. Basson, D.A. 1984. Kiesstelsels van proporsionele verteenwoordiging (Report POL 6). Pretoria: HSRC. Besprekingsdokument, 1991. Besprekingsdokument oor 'n nuwe kiesstelsel vir Suid-Afrika. Demokratiese Party. Bogdanor, V. & Butler, D.E. (eds.) 1983 Democracy and elections: electoral systems and their political consequences. Cambridge: Cambridge University Press. Botha, S. 1996. South Africa's party system, in Faure, A.M. & Lane, J-E. (eds) South Africa: designing new political institutions. London: Sage. Butler, D.E. 1983. Variants of the Westminstermodel, in Bogdanor, V. and Butler, D.E. (eds.), Democracy and elections: Electoral systems and their political consequences. Cambridge: Cambridge University Press. Calland, R. & Merton, N. 1996. Changes in electoral system put off to 2004, in Parliamentary Whip. May 17 : 1,3. Constitution of the Republic of South Africa, 1996. Constitutional Talk, 1995. Official newsletter of the constitutional assembly. 17 March — 6 April, No. 5. de Ville, J. 1966. The electoral system in terms of the South African interim constitution, in de Ville, J. & Steytler, N. (eds.) Towards 1999: Choosing an Electoral System. Durban: Butterworths. Elections for Parliament and the Provinces in the New South Africa 87 de Ville, J. & Steytler, N. (eds.) 1996. Towards 1999: choosing an electoral system. Durban: Butterworths. de Villiers, B. 1991. An electoral system forthe new South Africa. Pretoria: HSRC Centre for Constitutional Analysis, Year 2, No. 2. DP Newsletter, 1994. April 1994 Election: A brief Analysis. Pretoria: July, du Places, A. 1979. Die Britse tydperk (1795-1803; 1806-1910), in Nieuwoudt, C.F. etal, (eds.), Die politieke stelsel van Suid-Afrika. Pretoria: Human & Faure, A.M. 1987. Demokratiese kiesstelsels: ontwikkeling, tipes, toepassings, in Faure, A.M. 1988. Kiesstelsels in die demokrasie, in Faure, A.M. et al, (eds.) Rousseau. Politeia. Vol. 6(1): 3-23. Suid-Afrika en die demokrasie. Pinetown: Owen Burgess. Faure, A.M. 1994.a. The electoral system, in de Villiers, B. (ed.) Birth of a constitution. Kenwyn: Juta & Co., Ltd. Faure, A.M. 1994.b. Suid-Afrika se nuwe proporsionele kiesstelsel — Hoe werk dit?, Wat is die kenmerke daarvan?, en Hoe moet dit beoordeel word?, in Politeia, 13(1): 31-50. Faure, AM. 1996.a. The electoral system, in Faure, AM & Lane, J-E (eds) South Africa: designing new institutions. London: Sage. Faure, A.M. 1996.b. The electoral system, in Journal of theoretical politics. 8(2) April : 193-208. Faure, A.M. 1996.c. Proposal 1: Prpoprtional representation and single-member constituencies, in de Ville, J. & Steytler, N. (eds.) Towards 1999: choosing an electoral system. Durban: Butterworths. Faure, A.M. 1996.d. Electoral matters, in Politeia, 15(2): 65-69. Frost, M. 1993. Choosing an electoral system, in Die Suid-Afrikaan. No. 43, February/March. Interim Constitution of the Republic of South Africa, 1993. Johnston, R.W. 1993. PR at work: a case study, in Die Suid-Afrikaan. No. 43, February/March. Lijphart, A. 1985. Electoral systems, party systems, and conflict management in segmented societies. Paper prepared for presentation at the International Conference on Intergroup Relations of the Human Sciences Research Council. Pretoria: September 10-11. Mackie, T.T. and Rose, R. 1974, 1976,1982. International Almanac of Electoral History. (1st and 2nd editions) London. National Party, 1991. Constitutional Rule in a Participatory Democracy (the National Party's framework for a new democratic South Africa). National Party. Nieuwoudt, C.F. 1966. Die konstitusies van die Vrystaat-en die Transvaalse Republieke, in van Jaarsveld, F.A. & Scholtz, G.D., (eds.), Die Republiek van 88 Murray Faure Suid-Afrika: Agtergrond, ontstaan en toekoms. Johannesburg: Voortrekkerpers. Nieuwoudt, C.F. 1979. Die staatsreelings van die Voortrekkers en die Boere, in Nieuwoudt, C.F. etai, (eds.), Diepolitieke stelsel van Suid-Afrika. Pretoria: Human & Rousseau. Pienaar, P. A. 1994. Results of the April 1994 election. Unpublished manuscript. Pretoria: Department of Geography, University of South Africa. President's Council, 1990. Report of the Committee for Constitutional Affairs on Decision-making and Conflict Resolution Mechanisms and Techniques in Constitutional Systems. Cape Town: Government Printer. President's Council, 1992. Report of the Committee for Constitutional Affairs on a Proportional Polling System for South Africa in a new Constitutional Dispensation. Cape Town: Government Printer. Reynolds, A. 1993. Voting for a new South Africa. Cape Town: Maskew Miller Longman. SABC 16 April 1997. RSG News Bulletin 13:00. Schedule 2 of the Interim Constitution of the Republic of South Africa, 1993. Sisk, T.D. 1993. Choosing an electoral system: South Africa seeks new ground rules, in Journal of Democracy. Vol. 4 No. 1, 79:91. Sunday Times 1994, May 8. Thompson, G. 1986. Change without bloodshed. The Star, 12 February 12. UWC/CDS, 1990. Workshop: Electoral Systems: A Discussion Document. Cape Town.