Federalism: the state of the debate in South Africa
Hennie Kotzé

Centre for International and Comparative Politics
Department of Political Science
University of Stellenbosch

Abstract
Introduction
The variety of choices for a constitutional model


Federalism: a definition
The federal character of the transitional constitution
Arguments in the federalism debate: a summary
An empirical test of federal preferences
Concluding remarks

ABSTRACT

The main aim of this article is to give an overview of the debate on federalism in South Africa. The historical development of federalism in South Africa is touched upon and the various choices of a constitutional model as proposed by the main political parties are discussed. Elite attitudes towards federalism as a constitutional option for South Africa are analysed. An interesting point made is evidence of a convergence in attitudes by both the ANC and the NP towards federalism and how it should be implemented in a future constitution. The conclusion is reached that in the final analysis, the nature of the transition, the institutions which existed, the transitional constitution which was negotiated and the political culture in South Africa create the opportunity to devise a constitutional system which would be widely acceptable to a great number of the diverse groupings in South Africa.

1    INTRODUCTION

At various times in South Africa's history, political leaders have been faced with different choices regarding regime structures. 1 This article focuses on the outcome of one such choice: the nature of the constitutional dispensation which entrenched the democratic system of government in the transitional constitution of 1993. (The final choice still has to be made before 1996.) To put these regime choices in a historical context and to better understand the motivations for the choices made, a brief overview of the options available since the establishment of Union is given.

The enduring problem in South African politics has had two dimensions: the domination, in both political and economic terms, of the white sector of the population over the black majority, and the potential domination this majority might acquire through their numbers. This problem was entrenched in the South African state that came into being with Union in 1910.

As the name indicates, the Union was a unitary state negotiated only by white political leaders at the National Convention of 1909--1910. (In a united response to their exclusion, black leaders established the South African Native National Congress in 1912 -- later renamed the African National Congress.) Although the representatives from Natal propagated a federal system because it was outstandingly suited to `accommodate political conflicts', the argument by General Smuts was that a federation would not buttress the spirit of unity in the various white communities after the war and unity was needed to fight the 'native question'. The argument of Smuts and his supporters won the day. 2

In 1910 most white males who had attained their majority were enfranchised, as were a fairly affluent but small section of the coloured male population in the Cape Province. (Although some African males in Natal and the Cape had the vote, their numbers remained insignificant.) Within the course of two decades after Union, universal franchise became the norm as far as the white population was concerned. At the same time, however, the political rights of 'non-white' South Africans were being steadily eroded. This led to muted, and later overt, dissent on the part of the affected sector of the population. Meanwhile, the depression of the 1930s and South Africa's involvement in the war in the 1940s led to unprecedented urbanisation and industrialisation.

After the National Party's unexpected victory in 1948 South African politics entered a new era. The unitary type of constitution made it relatively easy for the NP to consolidate and centralise power. Apartheid legislation, and the hardship it caused blacks, led to black organisations actually challenging the white-dominated state. But Verwoerd extended social and political apartheid still further with the creation of `independent' homelands. This further stimulated resistance and led to the emergence of the African National Congress (ANC) and Pan Africanist Congress (PAC) as the direct opponents of the state.

Faced with the prospect of losing its dominating position by extending the franchise to all in South Africa, the white group in particular -- and for that matter other parties fearful of the simple majoritarianism advocated by the ANC and the PAC -- responded in a variety of ways in the 1970s and 1980s. These responses amount basically to an acceptance of the necessity to extend political rights to the whole population (although the terms of this extension have rarely been very generous), coupled to a rejection of a unitary form of state. ƒIn most cases these alternative political options were essentially `conservative' in that they either proposed to break up the existing unitary state to ensure minority domination, or they involved the decentralisation of power away from the central government (allowing substantial lower-level autonomy in the process), or they involved elite decision-making in order to escape the full consequences of mass democracy. These alternative non-unitary constitutional designs can be classified as follows: 3

South Africa's politics underwent more dramatic changes in the 1980s than in any decade since Union was established. The government's aim was to establish a multi-nation/multi-state. After the implementation of the tricameral constitution in 1984, with a state president as executive head of state, the inherent contradictions in the apartheid policy forced the government to look at new options. Burdened by the financial cost of apartheid, increased resistance from the disenfranchised, and the socio-economic challenges which urbanisation, unemployment, housing and education posed, the De Klerk government opened the way in 1990 for a democratic settlement. The acceptance by the NP of the concept of `one nation, one country' meant that the party made its final symbolic break with apartheid in 1990.

In the four years that led to the final acceptance of South Africa's new constitution late in 1993 an entirely new constitution with new political institutions had to be designed. This article focuses on this constitution and its institutional characteristics with specific reference to the vertical division of powers -- national and subnational governments. Firstly, this article briefly traces the shifts in parties' policies and regime choices to arrive at a compromise constitutional model. Secondly, some aspects of the chosen regime structure -- a federal constitution -- will be elaborated on. Thirdly, a selection of points in the debate between the proponents and opponents of a federal model will be summarised. Finally, the results of a recent opinion-leader survey will be used to compare the constitutional preferences of the respondents with their parties' policies.

2    THE VARIETY OF CHOICES FOR A CONSTITUTIONAL MODEL

The full spectrum of choices of constitutional models really came to the fore only in the period after 1990. It became clear in the constitutional debates during the negotiation process that the form of government to be adopted was a contentious matter.

After the various divergent points of view had been expressed it gradually became clearer that `regionalism' was coming to occupy an important place in the debates. But one of the most important problems was that consensus could not be reached on the powers to be allocated to regional governments. Yet the acceptance of a form of regional government as a point of departure in the debate already represented an important shift in the policies of two of the most important parties, the ANC and the NP, as will be shown below.

2.1    The ANC and regionalism 5

Initially the ANC in particular took an unambiguous stand on a unitary South Africa. In fact, they emphasised throughout that they advocated a `united, democratic, non-racial and non-sexist country'. They started with a mandate to transform South Africa radically and this could only be done by a strong centralised state. Their members also pointed continually to the balkanising effect of apartheid policies and insisted that a single nation must be built at all costs. Furthermore, the ANC emphasised that a weak central state would not be able to undertake the necessary redistribution of wealth among the regions as well as among the citizens. 6 The `Constitutional Guidelines' and Harare Declaration drawn up in August 1989 again emphasised the ANC's commitment to a unitary form of government (Kotzé & Greyling 1994:56).

Their strong support for a unitary form of government must be read in conjunction with their opposition to group rights. Although the ANC welcomed cultural diversity, the organisation did not want an emphasis on ethnic differences to become the motivation for a form of government with strong regional powers.

Gradually a shift in the role of regional governments came to the fore. The ANC published a discussion document in April 1991 which indicated that elected regional governments should be instituted (ANC 1991). Yet it was clearly stated that these regional governments should operate within the limits set by a central government. What this model amounted to was a central government with all the power vested in the party which won a majority of the votes. Furthermore, provision was made for a bicameral system, with a parliament and senate; proportional representation; a bill of rights; independent judiciary and judicial review; and a president (elected generally or appointed by parliament).

In the policy document, `Ready to Govern' (ANC 1992a), as well as in the discussion document on the proposed regions for South Africa (ANC 1992b), the organisation emphasised that its constitutional model involved a `unitary' South Africa and that the central government had to have overriding powers over the regional governments. The relationship had to be one of `overlapping or concurrent jurisdiction, rather than segmented and competitive powers' (ANC 1992b:18).

After intensive debate on the issue of regionalism, and especially the number of regions to be included, another discussion document on regions was issued late in 1992 (ANC 1992c). In it the issue of dispensations such as federations and unitary states is addressed directly. It is stated quite clearly that only in highly exceptional cases can the central legislature of a state not override the regional legislatures. This is why the central government will always have overriding authority in spite of the `original powers' which regional governments might possess (ANC 1992c:14--15).

Welsh (in Friedman & Humphries 1993:67) makes the following interesting deduction from the ANC's description of the powers of regional governments: `The tenor of the ANC's position strongly suggests an attempt to achieve some extent of common ground between federalist and unitary proposals.'

The PAC, which joined the negotiating process very late, also deserves to be mentioned in the context of support for a unitary state. Its regime model is a populist majoritarian system with a unicameral legislature.

2.2    The NP's switch of models

In the late 1980s the NP still adhered firmly to the idea of group rights. Yet once De Klerk came into power in 1989 there was a shift away from statutory group rights. First of all, the 1989 `Plan of Action' gave such a vague description of groups that most NP supporters found it difficult to explain. But early in 1990 it was explained that groups need no longer to be the essential element in a constitutional model -- although protection of groups was an option that could be built into a new constitution.

However, the shift in NP policy on minorities was at its most obvious in a statement in September 1990 by Gerrit Viljoen, former Minister of Constitutional Development. He indicated that the government intended to scrap the notion of protection of minorities based on race. He also gave a clear indication that little remained of the NP's original policy of `rights for racial groups'. Instead of defining minorities, Viljoen said it was necessary to provide for guidelines and procedures whereby people could voluntarily elect to declare themselves minorities, if preferred (Kotzé & Greyling, 1994:215. See also Sunday Times, 14 October 1990).

In his opening address to Parliament in February 1991 De Klerk announced in a `Manifesto for a New South Africa' that the emphasis would fall on nation-building. For the first time there was no talk of racial groups and their protection. Instead of talking about groups, it was now stated: `We commit ourselves to the creation of a free and democratic political system ... in which ... the rights of all individuals and minorities defined on a non-racial basis shall be adequately protected in the constitution and in a constitutionally guaranteed and justiciable bill of rights' (NP 1991).

This change of direction in NP thinking had important implications for their constitutional model. For the first time in September 1991 they spelled out the institutional components of the plan. 7 Some of the most important elements were the following: on the first tier a unitary state with federal characteristics; on the second tier nine regions each with its own government -- each region would have autonomy over certain matters and have its own tax base; the third level would consist of municipalities with strongly devolved powers; a bicameral legislature; and an `executive college' in which the leaders of three or four of the strongest parties would take decisions on a consensus basis (NP 1991b). This proposed model of the NP was clearly a mixture of federal and consociative elements.

Although the NP would not initially use the concept of federalism, it was openly acknowledged by late 1992 that this was the regime model they were advocating. 8 In Codesa itself the government was insisting on strong regional powers because this was seen as `protection against `simple' majority government'. Furthermore, they argued that the powers granted to the various levels of government should also include `fiscal competency'. 9 However, differences of interpretation, especially between the government and the ANC, created the impression that the government did not defend regional autonomy strongly enough.

In response to a growing concern among its own supporters and homeland allies that it was indeed abandoning its preference for a policy of regionalism, the NP organised a conference in September 1992 on various aspects of federalism. To many observers it was ironic that a party which had rejected the concept for many years suddenly became the main proponent of this form of government.

Although conventional wisdom has it that federalism was part of the NP's negotiating agenda at Codesa, the party had never given any real content to its policy of regionalism. It seemed that the NP, under Roelf Meyer, Minister of Constitutional Development at the time, was opting for a `unitary federalism'. In such a system the federal government still has greater powers than in classic federal systems. As far as the larger right-wing parties such as the Conservative Party and others were concerned, there was no question of abandoning the idea of a separate territory for the Afrikaner group. This constitutional option was described as `Afrikaner self-determination'. It was therefore difficult to involve this group in the negotiation process, which was seen as destructive of the `Afrikaner volk'.

2.3    The Inkatha Freedom Party and the Democratic Party: fighting a rearguard action on federalism

In the initial stages of the transition in South Africa only one party proposed a federation as a constitutional model, namely the Democratic Party (DP). From time to time Inkatha leader Buthelezi mentioned federalism as an option, but it was not initially clear what regime model the IFP would find acceptable. It was only much later that they came up with a detailed model with federal characteristics. However, Buthelezi had a long history of support for the idea of regional autonomy.

The most notable proposals about regional autonomy as an option for South Africa have undoubtedly been those thrashed out in the Natal/KwaZulu region in the 1980s. This process culminated in the proposals of the KwaNatal Indaba of November 1986. The Indaba, called by the Natal Provincial Council and the KwaZulu government under Buthelezi's leadership, was an attempt to effect some form of joint decision-making in the Greater Natal area. In terms of the proposals, Natal would assume special autonomous status in relation to the rest of the country. This autonomous area, it was suggested, should be the first unit of a South African federation. 10 In the late 1980s no clear constitutional model was being advocated by Inkatha. It was clear, however, that Buthelezi wanted to convey the message that his organisation was in favour of universal suffrage with the retention of cultural rights. His only qualification was that group domination should not be possible in such a system.

After the establishment of the Inkatha Freedom Party in 1990, this party advocated a clearly defined federal form of government for South Africa. In the early stages of the transition process Buthelezi adopted the view that KwaZulu had to be part of any new government as a region with veto powers.

After an erratic performance at Codesa and on-off bilateral talks with the government and the ANC, the IFP finally agreed in May 1992 to an elected constitution-drafting body. In the Concerned South African Group (Cosag) the only vision shared by the IFP and the white right-wing groups was that of strong regional autonomy. In October 1992 the IFP published a document detailing the `Constitution of the State of KwaZulu/Natal' which provides for this region to be a member of a federal republic of South Africa. While acknowledging the IFP's right to put forward constitutional proposals, both the ANC and the government criticised the fact that the draft constitution had been released as a fait accompli.

The Democratic Party (established in 1989) was probably the party which adopted the most unambiguous point of view in its constitutional preferences. With a core of supporters from the Progressive Party and the later Progressive Federal Party, the DP could already spell out its federal proposals for a constitution in detail at its establishment. In a discussion paper published in August 1991, the party proposed a geographical federation consisting of 8 to 12 federal states, each with its own government. There was also a bicameral parliament, proportional representation, a directly elected president and a prime minister elected by parliament. 11 In summary, it is thus clear that the larger parties -- the ANC, NP and IFP -- expressed clear preferences for regional governments although there were differences about the types of powers that should be granted to such governments. (Another of the parties which was also successful in the 1994 election, the African Christian Democratic Party, was also committed to a federal policy.) In the final phase of the negotiations there was no strong advocate for an `all-powerful centre with Leviathan-like qualities', as Welsh (1992:71) puts it, because the PAC's voice did not carry much weight.

3    FEDERALISM: A DEFINITION

Against the background of the preferences for the different regime models of the largest parties described above, it is clear that the transitional constitution (1993) is a compromise document. Given the divergent views on the relation between the central government and the regional governments -- that is a unitary versus a federal dispensation -- two questions arise immediately: why did the parties adopt this compromise and what is the nature of the constitution, that is, can one detect federal principles?

It is easier to provide an answer to the first question than to the second. Since this is not the focus of the present article, a reference which describes the compromise process in terms of `institutional choice theory' should suffice (Sisk 1994). The process of convergence is described as follows: `Actors reformulate their preferences as they move from their most desired alternative regime structure, the ideal, to the regime structure they perceive as achievable given the preferences of others, the possible. Moving from the ideal to the possible is really a matter of reformulating preferences away from those that can be unilaterally imposed toward the preferences of others through compromise. Following the advent of negotiation, preferences are formulated as not what is ideal, but what is `best attainable'. This occurs when actors converge on a set of institutions that they all perceive to serve converging interests and to be fair given the conditions they face ...' (Sisk 1994:52--53).

To answer the question as to whether there are federal principles in the transitional constitution, the concept of federalism first has to be defined.

There are many interpretations of the concept of federalism. 12 On the one hand, there is the normative or ideological dimension. This idea stems from the view that conflict arises between people because of their diverse interests. The importance of the federal principle is that, among other things, it creates the possibility of accommodating diversity and unity. 13

Elazar (1987:5) points out that `federal' derives from the Latin `foedus' which means covenant. 14 In essence a federal relationship thus means a `partnership, established and regulated by a covenant, whose internal relationships reflect the special kind of sharing that must prevail among partners, based on a mutual recognition of the integrity of each partner and the attempt to foster a special unity among them'. In its broadest sense federalism is geared towards uniting individuals, groups and `polities' into a stable but limited unity, in such a way that separate goals may be reached but all the parties retain their integrity.

On the other hand, according to Elazar (1987:5) federalism as a political principle has to do with the `constitutional diffusion of power so that the constituting elements in a federal arrangement share in the processes of common policy-making and administration by right, while the activities of the common government are conducted in such a way as to maintain their respective integrities. Federal systems do this by constitutionally distributing power among general and constituent governing bodies in a manner designed to protect the existence and authority of all'.

While recognising the more ideological or idealistic content which could be given to the concept of federalism, it also seems to have a more empirical content. In this case one must look for the features which could give an indication of the most distinctive characteristics found in a federal system.

One salient feature of most definitions of federalism is that the societies or states which form part of the federation strive for unity rather than uniformity. Kriek (1992:15) describes this feature as follows: `To give expression to this striving toward unity within diversity, those matters which are of common importance to all the federal parts are entrusted to the central government, while matters that are vital to the preservation of a separate identity are left to the authorities of the individual units. Accordingly, in a federation, sovereignty is to all intents and purposes partitioned. The central authority is sovereign in regard to the matters entrusted to it, while the states, or communities, or cultural groups, are supreme in their own terrains.'

In this case, in the final analysis `sovereignty' would naturally not include national sovereignty -- this is indivisible as there can be only one sovereignty stipulated by the constitution. In a federation the `sovereign powers' granted to the states or provinces could readily be revoked by the national parliament as is the case in a unitary state, where powers are delegated only to the lower units of the state. In a federal state these powers would be more strongly protected in various ways in the constitution.

King's (1982:77) definition is quoted here to link up with the above description and to operationalise the concept: `Basically we propose that any federation be regarded as an institutional arrangement, taking the form of a sovereign state, and distinguished from other such states solely by the fact that its central government incorporates units into its decision procedure on some constitutionally entrenched basis.'

This definition provides a very useful indication for identifying a federal constitution, namely that when, in addition to having the usual characteristics of a federal constitution, the constitution entrenches the powers of the states/provinces, then it is indubitably a federal constitution.

In conclusion, it should be noted here that the brinkmanship of the IFP played a large role in strengthening the federal character of the transitional constitution. By refusing to participate in the elections until the very last minute they could, for example, insist on changes such as the granting of exclusive powers to provincial governments rather than concurrent powers.

4    THE FEDERAL CHARACTER OF THE TRANSITIONAL CONSTITUTION

As has already been indicated, most political leaders constantly attached a negative connotation to the concept of `federalism'. 15 In some circles it was even referred to as the `f word'. In spite of this it still seems that the Transitional Constitution could be regarded as a federal constitution. In fact, Karpen (1994:7), a jurist from the University of Hamburg, said: `one can't and shouldn't overlook the fact that the interim constitution bears essential elements of a true federative system of government ...'. Kriek (1994:8), a political scientist at Unisa, concurs with this assessment: `Is the new form of state in South Africa a true federation? The answer is yes.'

It was mentioned above that King (1982) regarded the entrenchment of autonomous provinces in the decision-making process as the most important indicator of a federal constitution. Kriek (1994:2), on the other hand, uses the principles required by Riker (1964) in identifying federal constitutions: there must be at least a two-tier system of authority; each level of authority must have at least one functional area of capability in which it is autonomous; and there must be a guarantee in the constitution that each authority can act autonomously within its own sphere of activity.

To summarise, the following principles (they vary in degree of importance) could, among others, be used as important indicators of the federal elements in the transitional constitution: 16

Finally, Elazar (1991:xiv--xv) makes the following important point, which is also applicable to the debate on federalism in South Africa: `Utilizing the federal principle does not necessarily mean establishing a federal system in the conventional sense of a modern federation ... federalism is a phenomenon that provides many options for the organization of political authority and power; as long as the proper relations are created, a wide variety of political structures can be developed that are consistent with federal principles.'

If at this stage one examines the parties' expression of their views on regionalism and federalism, it is clear that a great deal of `reformulation of preferences' has occurred to achieve the `best attainable' regime model. 18 The ANC has moved closer to the NP on the issue of entrenchment of regional governments, whereas the NP has had to give up a number of the consociational elements of its model to arrive at a federal power-sharing regime model. In this process of compromise the smaller parties such as the DP and the IFP undoubtedly played an important role.

5    ARGUMENTS IN THE FEDERALISM DEBATE: A SUMMARY

From the discussion in section 2 it is clear that the policies of the different parties cover the entire spectrum from `unitarism' to `federalism'. These differences were reflected in the so-called regionalism debate which preceded acceptance of the transitional constitution.

Proponents of a unitary system advocated a greater or lesser degree of decentralised powers and functions for regions which were subject to the central government.19 The differences in this camp amounted only to differences as to the degree of centralisation. As opposed to them, there were the proponents of autonomous regions who insisted on the principle of 'non-centralisation'. 20 This debate, however, encompasses much more than simply the powers of the provinces -- that is only one of the most important determinants in the operationalisation of the principle of federalism and is thus summed up here.

It is obviously impossible to decide which is the `best' system on the basis of this debate, because such a choice is governed by a value system. The particular choice made by the negotiators at Kempton Park also reflects simply what they regarded as the `best attainable' structure. 21 Although it is difficult to categorise the different aspects of the debate as many of the arguments are mutually related, an attempt is made below to summarise some of the most important points in the broad debate on a unitary versus a federal regime model: 22

Although the circumstances surrounding regionalism cannot be applied fully to the requirements for federalism, they do provide interesting indicators for the federalism debate. Researchers have shown that normative considerations are more important constraints on successful decentralisation than material considerations. Cloete (1991:28) summarises these findings as follows: `A genuine value commitment to the establishment of subsystem autonomy normally leads to serious efforts to create the material conditions for success. Conversely, even with the advantage of sufficient resource for effective regional autonomy, a lack of real commitment by elites to the principles and goals of regional autonomy will ensure failure as a result of creeping centralisation.'

The commitment to federalism can obviously be tested only in practice, that is, the transitional constitution must first come into effect and the provincial governments begin to operate.

A comparison between the personal constitutional choices of the top opinion-formers in the different parties and the constitutional structure accepted as the `best attainable' could possibly provide a superficial answer to the question of the commitment to federalism.

6   AN EMPIRICAL TEST OF FEDERAL PREFERENCES

A longitudinal survey which has been conducted annually since 1989 among the top opinion-leaders in South Africa has already shown considerable changes in the constitutional preferences of decision-makers. 28 The constitutional preferences of the respondents in the most recent survey emerge as indicated in the table. The survey was conducted during the period July to August 1993 -- that is before the transitional constitution was finalised. 29 From the definition of federalism cited earlier two important indicators as to the type of federation -- weak or strong -- can be deduced. First, the type of powers and, second, how much power the provinces have, may give an indication of the power relationship between the central government and the provinces.

As indicated earlier, by mid-1993 there was a series of preferences which represented the full spectrum of the constitutional models -- from the old apartheid model supported by some of the right-wing groups to the idea of a unitary state based on a class system as proposed by the PAC. The models proposed were operationalised as follows in the questionnaire:

  1. Apartheid -- `a geographically undivided South Africa, entirely controlled by whites, with blacks having only regional and municipal powers subject to a white parliament';
  2. Confederation -- `a number of ethnically based sovereign states within a confederation or commonwealth in which each nation has the right to self-determination';
  3. Volkstaat -- `a geographically divided SA, the smaller region/province/state under Afrikaner control, and the other parts under non-racial control';
  4. Strong federation -- `a federal system which makes provision for strong regional/provincial/state government elected on a non-racial basis by means of a proportional electoral system and with a bill of fundamental rights';
  5. Weak federation -- `a federal system which makes provision for a strong central government and region/provincial/state governments with lesser autonomy and powers elected on a non-racial basis by means of a system of proportional representation and with a bill of rights';
  6. Centralised system -- `a centralised system for a united South Africa which makes provision for majority government with decentralised elected regional administrations which are in all respects subordinate to the central government';
  7. Class state -- `a geographically undivided unitary system in which class distinctions are of primary importance, and where the black working class controls the central government'.
Respondents were asked to indicate their first, second and third choices. It is clear from table 1 that there is no essential difference between the official party policies on constitutional models and the preferences of the parties' supporters. The supporters of the parties which advocate federalism all selected as their first choice the model which can be described as `strongly federal' -- in decreasing order of support, the IFP (93,3 per cent), NP (86,1 per cent) and DP (71,3 per cent). If support for the so-called `weak federation' is included, there is more than 90 per cent support for a federation as a constitutional model in all these cases.

It is interesting that the ANC/SACP supporters do not give the same strong support to a centralised system. While 73,5 per cent do support this, there are about 20 per cent of their supporters who also select a federation as a model, albeit the weak form of federalism. This provides a possible indication that the ANC/SACP had already moved slightly away from a centralised system. The PAC shows a distribution between the official party model of the class state and a centralised state. There is equal support for these among PAC members -- 44,4 per cent support for each of the two models. There are thus clear indications of a shift in the thinking of PAC supporters.

Table 1    Support for regime structures -- first choice 30

Regime structure NP DP CP ANC/SACP IFP PAC Other No party
Apartheid 0,6 -- 6,5 -- -- -- -- --
Confederation 1,7 -- 84,8 -- -- -- 8,3 6,7
Volkstaat -- -- 6,5 -- -- -- -- 13,3
Strong federation 80,1 71,3 2,2 4,1 93,3 -- 50,0 43,3
Weak federation 11,0 25,0 -- 16,3 6,7 11,1 33,3 33,3
Union 0,6 3,0 -- 73,5 -- 44,5 8,3 3,3
Class State -- -- -- 6,1 -- 44,4 -- --
N=447 n=173 n=80 n=46 n=49 n=30 n=27 n=12 n=30

As could be expected the far-right supporters show a preference for the `confederation' model. There is only 6,5 per cent support for a `volkstaat' model. This form of constitutional system, which could be described as the more realistic model of self-determination, initially had the support of only the Boshoff group in right-wing circles. There has thus been a great shift in the thinking of the far right since General Constand Viljoen appeared on the scene. The strong opposition from the other parties towards any form of independent `state' forced the right-wing groups, who prefer a form of parliamentary opposition, to shift to their second choice, namely a `volkstaat'.

It is clear from the distribution of preferences in table 2 that there was a potential for compromise on a constitutional model if one looks at the second choices of the respondents. The supporters of a strong federation in the NP, DP and IFP had a weak federation as their second choice. Most of the ANC/SACP supporters (55,3 per cent) also moved in this direction. There are even ANC/SACP members (17 per cent) who support a strong federation. There is thus a total of 72 per cent of ANC/SACP supporters who indicated one or other form of federation as their second choice. Among PAC members too there is almost 30 per cent (28,6 per cent) support for a weak federation.

Table 2   Support for regime structures -- second choice

Regime structure NP DP CP ANC/SACP IFP PAC Other No party
Apartheid 1,2 -- 4,7 -- -- -- -- --
Confederation 3,0 -- 4,7 -- 19,2 14,3 -- 9,7
Volkstaat 6,7 1,3 76,7 2,1 3,8 -- -- 3,2
Strong federation 11,5 22,8 11,6 17,0 7,7 -- 45,5 32,3
Weak federation 73,3 69,6 2,3 55,3 65,4 28,6 36,4 45,2
Union 3,6 6,3 -- 17,0 3,8 42,9 18,2 9,7
Class State 0,6 -- -- 8,5 -- 14,3 -- --
N = 447 n=173 n=80 n=46 n=49 n=30 n=27 n=12 n=30

As could be expected, far-right supporters indicated a volkstaat as their second choice. No fewer than 76,7 per cent support this model. Once again it is the compromise model for this group. This could thus be the reason that the Negotiating Council succeeded in persuading a large number of far-right leaders in the Freedom Alliance to participate in the election and subsequent constitutional process at the eleventh hour. With the inclusion of the 34th principle in the constitution, creating the possibility of a `volkstaat', General Constand Viljoen, as leader of the Freedom Front, could declare his readiness early in 1994 to participate in the elections. (In some quarters it was suggested that it might be worthwhile for the Freedom Front and Volkstaat Council to investigate the concept of corporate federalism, instead of the geographically linked volkstaat 31 ).

This pattern of attitudes indicates how it was possible to arrive at a compromise proposal in the Interim Constitution that had the features of a federation. However, as indicated there is naturally no consensus on this description among politicians, and from an analysis of the above tables also not much commitment to a federal structure from the members of the majority party in the Government of National Unity.

In the final phase of constitution writing there is thus a possibility of strengthening federal features or weakening them. In such a system in which different centres of power have been created in the nine provinces, each with a number of exclusive powers, the dynamics of autonomous powerbases and political and economic interests could lead to the provincial leadership taking a strong stand on the reduction of their powers -- despite their initial low level of commitment towards a federal structure. The economic development of the various provinces, however, will be decisive in establishing the scope of the exclusive powers they develop.

7 CONCLUDING REMARKS

This review has presented a run-up to the final choice of a regime structure. It is clear that the long historical experience preceding the negotiating process exerted an important influence on it. Two important factors that influenced the debate may be singled out: federalism is seen in some circles as a mechanism to entrench white privilege; and one of the strongest resistance movements had a mandate to transform society in a revolutionary way -- which can be done only with a strong central government. This is why the concept of federalism unleashed political emotions.

After the short theoretical discussion of the concept of federalism, it was argued that South Africa's transitional constitution is in fact a federal constitution. But if the concept seems to be unacceptable to the largest part of the population another name should be found for the regime structure.

It appears from the summing up of the arguments of the `unitarists' and the `federalists' that there is a great deal of overlap in the requirements of the respective regime structures -- sometimes only the terminology differs. A union and a federation lie along a spectrum, with only vague boundaries between some institutions and functions in the various regime models, which makes it easier to reach a compromise on powers and functions. The one important requirement here is the will of the political leaders to make the compromise model work -- this could develop rapidly among the provincial leadership as a consequence of the dynamics and benefits inherent in the acquisition of autonomous decision-making powers.

In the final analysis, the nature of the transition, the institutions which existed, the transitional constitution which was negotiated and the political culture in South Africa create the opportunity of devising a constitutional system which will be widely acceptable to a great number of the diverse groupings in South Africa.

ENDNOTES

1 Fishman (1990:428) describes a regime as that component within the state which consists of values, norms, rules and procedures which deal with the ordering of power within the different organisational sites of the state. Regimes, democratic and otherwise, are therefore embedded in the wider institutional framework of the state. Stephan's (quoted in Skocpol, 1985:7) definition of the state is preferred; he identifies the state as `the continuous administrative, legal, bureaucratic and coercive systems that attempt not only to structure relationships between civil society and public authority in a policy but also to structure many crucial relationships within civil society as well'.
2 For a brief discussion of this argument see Welsh in Forsyth (1989).
3 It will be appreciated that some of these models are linked to one another in a variety of ways. Confederalism, for instance, is hardly possible if some form of partition does not prevail. The differences between the alternatives also bear mentioning: partitioning is a spatial design, while federation and confederation entail a change in the constitutional form of the state. Consociationalism may be seen as a quasi-democratic form of decision-making.
4 The theorists of consociationalism advocate a form of political representation characterised by cooperation between political leaders of different cultural, racial, ethnic, religious or linguistic segments in a plural society. This almost paradoxically boils down to democracy via elite decision-making. According to Lijphart (1968:71) consociational democracy entails grand coalition government of all leaders; segmental autonomy to groups; mutual veto rights to all groups to assure them sovereignty in matters important to them; and proportional representation in governing institutions.
5 Cloete (1991) uses the following definition by the Council of Europe for regional structures: `all institutions between the level of commune and central government, irrespective of whether they come directly below the level of central government or are nearer to municipalities'. He then comes to the conclusion that regionalism can be logically used either in a descriptive or prescriptive way because `it can be used to describe the regional approach to government, or refer to the theoretical justification for or rationalisation of geographically intermediate or regional structures'.
6 See Dirk Kotzé's summary of the ANC's stand on a unitary South Africa in Kriek, 1992.
7 Humphries (1992:57--58) discusses this change in direction in more detail.
8 Celliers (in Kriek,1992:241), a former NP member of the President's Council, makes the following interesting remark about the use of the concept: `The mere fact that the NP rejected the concept `federalism' for many years did not mean, however, that it summarily rejected its structural and functional components.'
9 Friedman (1993:65--66) gives an excellent account of the negotiations in Codesa on this issue. See also Welsh (1992b).
10 See Bonthuys & Du Toit (1987) and Kriek and Louw (1987).
11 For a more detailed discussion see Democratic Party (1992) and Kriek (1992:234--235).
12 Duchacek (1970), and Burgess in Burgess & Gannon (1993) both give good accounts of what can be described as the fundamentalist or moral dimension and the political dimension of federalism. Kriek (1992:13--16), on the other hand, gives a good summary of the more empirical definitions of federalism.
13 The moral argument favouring federalism was severely criticised by Riker (1969:142--146), who stressed that the argument that a federation is necessary because it fulfils certain functions better than other state forms is false. He stresses that it is the promise of local autonomy that brings people with a regional interest into a federal arrangement. What keeps the central government in a federal system from destroying or maintaining local autonomy is the characteristics of the (regional) political cultures.
14 Kriek (1992:12) describes the meaning of federalism as follows: `it refers to an alliance, a treaty or an oath of allegiance and, by implication, emphasises the right of those who share in the covenant to make their own decisions'.
15 See De Villiers (1993) for a discussion on the controversy surrounding federalism as concept in South Africa. Elazar (in Licht & De Villiers, 1994:77) notes this controversy and writes that South Africans need to find another word.
16 This is a summary of the points mentioned by Kriek (1994) and Karpen (1994).
17 It is interesting to note that there has already been a reference to this stipulation in the debate on housing standards between the Minister of Housing and the Premier of the PWV.
18 In an interview in the Weekly Mail (July 23--29, 1993) Cyril Ramaphosa, Secretary General of the ANC, responded as follows to a question as to whether they had made compromises on constitutional principles with respect to regional government: `We believe these principles are adequate to make regional government effective, and to bring government closer to the people so that the people can enjoy meaningful participation. We have agreed to entrench the powers of regional government in a constitution ...'
19 Cloete (1991:27) identifies three types of decentralised powers in unitary systems: `deconcentrated administrative functions; delegated legislative, executive and administrative discretionary powers and functions; and devolved legislative, executive and administrative powers and functions'.
20 Elazar (1987:173) maintains that noncentralisation is `strengthened in all federal systems by giving constituent polities guaranteed representation in the federal legislature and often by giving them a guaranteed role in the federal process'.
21 Underlying this choice is the preference for process politics.
22 Most of these arguments are distilled from four recent studies reporting on conferences/workshops on federalism in South Africa, namely: Friedman & Humphries (1993); Licht & De Villiers (1994); The American University Journal of International Law and Policy (1992/93); and Development and Democracy (1993).
23 See Sutcliffe in Friedman & Humphries (1993:193).
24 See Horowitz (1991:214--131) for a discussion of this aspect.
25 Elazar (in Licht & De Villiers, 1994:78) expresses concern about what he calls `the reified state'. He writes that the `great revolution in modern democratic republicanism was to get rid of that idea of the state and to see the people as the source of political power'.
26 It is interesting to note that the central government in Canada is more limited in relation to provinces `than the national government of the USA -- this is especially true for economic planning' (Field 1992/1993:450).
27 See Sachs (1992:185) for a discussion.
28 See Kotzé (1991, 1992 & 1993). One of the most important reasons that the project has concentrated on the attitudes of elites or opinion-leaders is that they play an important role in political transitions and in the establishment of a stable democracy. Elites are `those persons who hold authoritative positions in powerful public and private organisations and influential movements, and who are therefore able to affect strategic decisions on a regular basis'.
29 See Kotzé (1993:13--36) for the methodology of the survey.
30 A striking feature of the tables is that NP supporters outnumber ANC supporters. By mid-1993 the ANC had been unbanned for only three and a half years, and thus had not yet been able to establish itself within the ranks of the positional leadership in many sectors.
31 Friedrich elaborates as follows on the advantages of corporate federalism: `Altogether the most important argument for corporate federalism is that it does away with the entire problem of territorial districting, which can then be tackled as a strictly administrative problem without attention having to be given to the complicating issues of population distribution and to cultural and linguistic requirements and aspirations.' For a more detailed discussion see Hugo and Kotzé (1983:181--182).

BIBLIOGRAPHY

ANC 1991.
Structures and principles of a constitution for a democratic South Africa. Johannesburg.
ANC 1992a.
Ready to govern. Johannesburg.
ANC 1992b.
Ten proposed regions for South Africa. Discussion Document. Bellville: CDS UWC.
ANC 1992c.
Draft discussion document on regional policy. Johannesburg.
Bonthuys, L & Du Toit, P 1987.
Die grondwetlike voorstelle van die KwaZulu-Natal Indaba as raamwerk vir die beskerming van groepsbelange, Politikon, Vol 14(1).
Burgess, M & Ganon, A G (eds) 1993.
Comparative federalism and federation: competing traditions and future directions. New York: Harvester Wheatsheaf.
Cloete, F 1991.
Regionalism in South Africa: constraints and possibilities, Africa Insight, Vol 21(1).
De Villiers, B 1993.
Regionalism in South Africa: the debate continues, Johannesburg: Konrad Adenauer Stiftung Occasional Papers.
Development and Democracy. 1993.
The regional question. Johannesburg: Urban Foundation. The American University Journal of International Law and Policy 1992/93, Vol 8(2&3).
DP 1992.
Constitution proposals: policy discussion paper. 23 August.
Duchacek, I D 1970.
Comparative federalism. New York: Holt, Rinehart & Winston.
Elazar, D J 1987.
Exploring federalism. Tuscaloosa: University of Alabama Press.
Field, M A 1992/1993.
The structures of federalism. The American University Journal of International Law and Policy, Vol 8(2&3).
Fishman, R M 1990.
Review article: rethinking state and regime: Southern Europe's transition to democracy, World Politics, Vol 42(3).
Freidrich, C J 1975.
The politics of language and corporated federalism. In Savard J G & Vineau, H R (eds). Multilingual political systems, problems and solutions. Quebec: L Vagneault.
Friedman, S & Humphries, R 1993.
Federalism and its faces. Johannesburg: CPS.
Friedman, S (ed) 1993.
The long journey. Johannesburg: Ravan Press.
Horowitz, D 1991.
A democratic South Africa. Berkeley: University of California Press.
Hugo, P J & KotzéH J 1983.
Suid-Afrika: oorlewing in politieke perspektief (2nd revised edition). Johannesburg: Jonathan Ball.
Humphries, R 1992.
National Party and state perspectives on regionalism, Africa Insight, Vol22(1).
Karpen, U 1994.
Federalism: an important instrument for providing pluralism in the new democratic South Africa. Johannesburg: Konrad-Adenauer Stiftung Occasional Papers.
King, P 1982.
Federalism and federation. London: Croom Helm.
KotzéD 1992.
The new South African state: why it should not be a federation. In Kriek D (ed). 1992. Federalism: the solution? Pretoria: HSRC Publications.
Kotzé, H J (ed) 1993.
A Negotiated Democracy, Research Report No 2, Stellenbosch: Centre for International and Comparative Politics.
Kotzé, H 1991.
Elites & democratization, Report No 1, Stellenbosch: Centre for International and Comparative Politics.
Kotzé, H J 1992.
Transitional politics in South Africa: an attitude survey of opinion-leaders, Research Report No 3, Stellenbosch: Centre for International and Comparative Politics.
Kriek, D J 1994.
The new form of state in South Africa: is it a true federation? ISSUP Bulletin, No 3.
Licht, R A & De Villiers, B (eds) 1994.
South Africa's crisis of constitutional democracy: can the US Constitution help? Cape Town: Juta.
Louw, A & Kriek, D J 1987.
Die KwaZulu-Natal Indaba: 'n oorsig van die werksaamhede. Politikon, Vol 14(1).
NP 1991b
Demokrasie in 'n Regstaat. Pretoria: NP Inligtingsdiens.
NP 1991
Manifesto for a new South Africa. Pretoria: NP Information Service.
Riker, H 1964.
Federalism, origin, operation and significance. Boston: Little, Brown & Co.
Riker, W H 1969.
Six books in search of a subject or does federalism exist and does it matter? Comparative Politics, Vol 2(1).
Sachs, A 1992.
Watch out -- there's a constitution about, South African International, Vol 22(4).
Sisk, T D 1994.
The Elusive Social Contract: democratization in South Africa. Cornell University Press (forthcoming).
Skocpol, T 1985.
Bringing the state back in: strategies of analysis and research. In Evans, et al (eds). Bringing the state back in. New York: Cambridge University Press.
Sunstein, C R 1992/1993.
Federalism in South Africa? Notes from the American experience. The American University Journal of International Law and Policy, Vol 8(2&3).
Welsh, D 1989.
Federalism and the problem of South Africa. In Forsyth M (ed). 1989. Federalism and Nationalism. Leicester: Leicester University Press.
Welsh, D 1992b.
Turning point towards democracy, MPD, Vol 1(1).
Welsh, D 1992.
Federalism and South Africa: The future is not yet written. In Friedman, S & R Humphries. 1992.
Federalism and its faces. Johannesburg: CPS.

Go back to Politeia Contents page