This study was published in 1998 in AFRICAN SECURITY REVIEW.
South Africa’s apartheid regime used boundary demarcations such as ‘homelands’ and ‘group areas’, to maintain and enforce political control by a white élite. The post-apartheid government must demarcate new boundaries to allow formerly repressed peoples more access to land and power. Dramatic changes since 1994 have included the absorption of the nominally ‘independent’ homelands, the cession of Walvis Bay to Namibia, redrawing every metropolitan and municipal boundary in the country, and transforming four old provinces into nine new ones.
The Security Costs of Party-Political Boundary Demarcations: The Case of South Africa
Richard A Griggs, Head of Research, Independent Projects Trust, DurbanPublished in African Security Review Vol 7 No 2, 1998
South Africa’s apartheid regime used boundary demarcations such as `homelands’ and `group areas’, to maintain and enforce political control by a white élite. The post-apartheid government must demarcate new boundaries to allow formerly repressed peoples more access to land and power. Dramatic changes since 1994 have included the absorption of the nominally `independent’ homelands, the cession of Walvis Bay to Namibia, redrawing every metropolitan and municipal boundary in the country, and transforming four old provinces into nine new ones.1
Of these spatial changes, only one has posed a problem of national security: the provincial boundaries. Firstly, the national treasury is being drained to support nine systems of provincial government, some of which are neither viable nor affordable. Secondly, there have been arson attacks, bombings, violent confrontations with security forces and reported deaths in association with boundary disputes left unresolved by the state. This article attributes these problems to the dominance of spatial decision-making by political party negotiations. In 1993, political party negotiators produced too many provinces with too much conflict potential and afterwards prevented local stakeholders from accessing the constitutional mechanisms for resolving them.
TOO MANY PROVINCES WITH TOO MUCH CONFLICT POTENTIAL
South Africa’s political transformation to democracy was facilitated by a series of multiparty talks held between 1990, when the National Party (NP) admitted the failure of apartheid, and April 1994, when the first democratic elections were held. One of the longest debated and most heated issues in these talks was whether South Africa would be federalist or centralist. The outcome would affect the subnational boundaries of the state. A centralist state could be administered through small-scale regional councils (e.g. Namibia)2 but a federalist state would require large `states’ or provinces (e.g. Australia).
By 1993, it was certain that some quasi-federal system of states, provinces, or regions (then called SPRs because each term alone was associated with a political party stance) would emerge as a compromise between centralist, federalist, and even separatist forces. On 28 May 1993, the Negotiating Council of the Multiparty Negotiating Process established a fifteen-person commission to make proposals for new internal boundaries. The Commission on the Demarcation/Delimitation of SPRs (the CDDR) held their first meeting on 8 June 1993 and reached a decision by 31 July 1993.3 It effectively took six weeks of deliberations to propose more than doubling the number of provinces (by comparison, the recent reorganisation of Britain’s counties was negotiated over six years).
Since no meaningful time was allotted for public consultation, the commissioners took as their initial draft the nine planning regions (Figure 1) established by the Development Bank of Southern Africa between 1982 and 1988.4 These planning areas crossed the boundaries of both provinces and homelands (i.e. appeasing the policy of the African National Congress (ANC) on integration) and were demarcated based on functional criteria during the apartheid era (familiar to the NP).5
Only one month of the CDDR’s itinerary was devoted to the gathering of testimony. Many stakeholders missed the notices or could not organise submissions in such a short time. Worse yet, few commissioners read more than a fraction of the reports that were submitted.6In reaction to broad public criticism, the multiparty negotiating committee allocated another three months beginning in August for the commission to receive additional presentations regarding certain `affected areas’. This second phase of commission work was largely a public relations exercise and effected no change in the commission’s recommendations.7 The decision to hear last minute testimony merely to appease the public indicates that local input was deemed less relevant to the political party negotiators at Kempton Park than a base map for their territorial `horse-trading’.8
Figure 2 illustrates that, after the commission submitted its report, the politicians chopped and swapped magisterial districts in order to reach a final party agreement. Of some six major changes to the commission map, four directly fueled conflict and mass protest. These included:
- transferring Kuruman, Hartswater and Postmasburg into the Northern Cape from the Northwest Province;
- incorporating East Griqualand into KwaZulu-Natal leaving the Eastern Cape’s Umzimkulu as an enclave;
- moving parts of KwaNdebele and adjacent areas from Gauteng into Mpumalanga; and
- moving parts of the former Bophuthatswana homeland and adjacent areas from Gauteng into the Northwest Province.
Furthermore, the CDDR’s worst error placing Bushbuckridge in the Northern Province rather than Mpumalanga was not corrected by the politicians. Thus, the multiparty negotiators helped to produce five disputes and may have prevented only one: moving Clanwilliam, Vredendal and Vanrhynsdorp into the Western Cape (the commission recommended these magisterial districts for the Northern Cape, but they are economically and functionally tied to Cape Town).
Whether one examines the Commission or the multiparty negotiations, the minimum input from ground level resulted in proposed boundaries that contained a high level of conflict potential. Mounting protests were allayed with an `annexure’ to the Interim Constitution that defined fourteen different disputed boundaries (Figure 3).9 The negotiators promised all interested parties that these disputes would be addressed again after the elections (that this was never satisfactorily done according to any of the mechanisms outlined in the Interim Constitution will be addressed later).
From a party-political point of view, the negotiations resulted in demarcations that offered important minority parties a future base of provincial power. The Zulu-dominated Inkatha Freedom Party (IFP) expected to win in KwaZulu-Natal and did. The NP expected to win in two provinces, the Western Cape and the Northern Cape, but lost by a one per cent margin in the latter as a result of the last minute boundary adjustment along the boundary with the Northwest Province that brought in a significant number of ANC voters. The majority party, the ANC, was guaranteed six provinces and took seven. This was significant for its powers of patronage: it awarded hundreds of veterans of the struggle with positions of power and high incomes within provincial governments.
What is good for political party compromises is not necessarily good for national security. Two problems resulted: too many non-viable provinces and four ensuing years of boundary conflicts. Only Gauteng and the Western Cape two provinces with thriving metropolitan regions and no former `homelands’ had the potential in 1994 to generate enough income to finance their own administrations. Perhaps three others (KwaZulu-Natal, Mpumalanga, and the Free State) could achieve self-sufficiency with another decade of capacity-building. The remaining four provinces were and remain non-viable for the foreseeable future. The Eastern Cape and the Northwest Province remain deeply troubled by the absorption of former homelands, while the Northern Cape and the Northern Province are short of resources, infrastructure and capacity.
Since seven out of nine provinces need central government support, the organisation of revenue is necessarily centralised and redistributive. So far, the central government has established development policies and made bulk allocations to provinces based on a formula meant to help the poorer provinces. This approach has set targets without sufficient cash to deliver (e.g. in 1997 the provinces received less than half of what was needed to meet national housing targets) or sufficient powers to manage change (e.g. provinces cannot lay off civil servants, yet more than ninety per cent of their budgets are for salaries).
The result of these financial relations is a large drain on central government revenue accompanied by ineffective provincial government. Firstly, 57 per cent of the national budget after debt servicing goes to the provinces.10 This is partly because provinces have few tax-raising powers. They rely on the central government for more than ninety per cent of their revenue. This also means that overspending provinces can damage national economic policy. By November 1997, the central government had to absorb R11 billion in provincial debt and refused to bail out the provinces any longer.11
Secondly, the provinces are key structures for delivering on central government promises of improved health care, education, housing, and employment. The inability to create workable provincial administrations has slowed the pace of transformation and created discontent. This, in turn, can create incidents of mass mobilisation and other politically and economically harmful responses from the electorate. This is precisely what occurred at Bushbuckridge where tens of millions of rand in infrastructure were destroyed by disputants in the boundary conflict between the Northern Province and Mpumalanga.12 This former area of the Gazankulu and Lebowa homelands had seen virtually no development since the days of apartheid because neither the Northern Province nor Mpumalanga cared to invest in an area that they might lose.
Boundary disputes are symptomatic of the larger problem of poorly designed and incapacitated provinces. The 1997 Provincial Review Report cites widespread evidence of incompetence, mismanagement, inadequate accounting, ethnic rivalries, nepotism and corruption.13 Nearly all the provinces have been rocked by continuous scandals. Some 650 cases of government corruption and fraud are being investigated in the Eastern Cape alone.14
In retrospect, a more manageable solution to territorial restructuring in 1994 would have been fewer provinces that inherited some capacity from past administrations.15 Four to six provinces designed by boundary specialists and approved by public referendum are likely to have cost less and been more effective. However, it is not possible to turn the clock back, because the provinces are deeply entrenched. In May 1997, the Minister of Constitutional and Provincial Affairs announced that the government and the collective leadership of the ANC would leave the borders as they are.16
Ironically, the Minister’s announcement came just one month after the Commission of Enquiry into the Taung/Kuruman/Kudumane Border Dispute filed their report recommending a limited referendum.17 NP spokespeople for Namaqualand (and certain other areas of the Northern Cape south of the Orange River) said in 1996 that they would demand incorporation into the Western Cape if the boundary along the Northwestern Province and Northern Cape was changed. Namaqualand was one of the `affected areas’ outlined in the 1994 Constitution and if Taung and Kudumane of the Northwestern Province were admitted into the Northern Cape, it would ensure an ANC victory there in the next provincial elections. With party politics stalemating all mechanisms for resolving this dispute and others (e.g. the Groblersdal dispute between Mpumalanga and the Northern Province), the government leadership acceded to the entrenchment of a poor design.
More than four years after the final provincial map was produced by multiparty negotiations, there are still eight or more active disputes (Figure 3). A clear count is not possible since some of these disputes would remain dormant if they were not linked as trade-offs to resolve other disputes (e.g. the Namaqualand example above). Worse yet, others could arise again (e.g. the viability of the Northern Cape) under new political circumstances. Only the debated split of Gauteng, the possible division of the Eastern Cape along the Kei River, and the status of Sasolburg appear to be permanently resolved.
Three disputes have escalated beyond mass protest to involve violence including killings, arson, land invasions, or wanton destruction of property. The army patrols two of the most violent areas: East Griqualand (between KwaZulu-Natal and the Eastern Cape) and Bushbuckridge (between Mpumalanga and the Northern Province). Military security, police investigations, material damage, and lost tourism revenue have cost hundreds of millions of rands in both cases.
ACCESSING MECHANISMS FOR RESOLVING DISPUTES
The second major argument in this article is that, once the political party negotiators created security problems along the provincial boundaries, they exacerbated these conflicts by discouraging the use of the referendum process as outlined in the Constitution. Many disputes were between ANC dominated provinces so the party leadership discouraged referenda in favour of internal negotiations. The NP fought to ensure that any boundary matter affecting their power base remained unchanged, referring to these disputes as closed-door matters permanently settled at Kempton Park.
On 27 October 1994, the six-month timeframe for referenda expired without ever having been tried. The remaining procedure for negotiating changes to provincial boundaries was given in Section 62 of the Interim Constitution and applied until 1996.18 Government commissions were employed when violence or mass action appeared to intensify in four locations (East Griqualand, Bushbuckridge, Groblersdal, and the Kuruman/Kudumane/Taung area), but these commissions were advisory and could affect no change in provincial boundaries whatsoever. Any boundary changes had to be initiated at a joint sitting of the National Assembly and the Senate and pass by a majority of at least two-thirds in both houses. In turn, this legislation had to be approved in the same manner by the houses of each affected province. Party politics stalemated any attempt to employ this complex process.
Eighteen months after the opportunity for referenda had elapsed, the final Constitution was accepted. All references to mechanisms for resolving boundary conflicts in the Interim Constitution were replaced by a single line: “The boundaries of the provinces are those that existed when the Constitution took effect.”19 This effectively ignored the existence of numerous boundary disputes and suggested a static view of boundaries as opposed to a system for boundary management.
After four years of resisting referenda, it is clear that both the ANC and hence most central government officials are generally negative toward the idea. Their main arguments have included:
- a pandora’s box thesis public referenda will unleash more and more demands for boundary changes;
- the high cost of referenda; and
- South Africa has soft boundaries and therefore location is irrelevant.20
More likely than any of these reasons is the difficulty of interfering with the power bases of the respective political parties as established at Kempton Park.
A geographic spread of boundary disputes as a result of public referenda seems unlikely. All five popular calls for referenda on boundary disputes since 1994 remain part of the fourteen `affected areas’ outlined in the 1993 Interim Constitution (Kuruman, KwaNdebele, Bushbuckridge, Groblersdal, and East Griqualand). There are more disputes than were officially recognised in 1993 that could surface (e.g. Pongola between KwaZulu-Natal and Mpumalanga), but they are also pre-existing and identifiable.
The costs of referenda and the issue of soft boundaries are also weak excuses that can be better understood in terms of an example. One of the eight areas of protracted dispute is the boundary between the Eastern Cape and KwaZulu-Natal that comprises the historically disputed areas of East Griqualand, Pondoland, and Umzimkulu. Just prior to the so-called independence of the Transkei homeland in 1976, two magisterial districts that might logically have been part of Transkei (Matatiele and Mount Curry) were incorporated into the province of Natal. This aroused resentment and created an anomaly immediately recognisable on any map of provincial South Africa. Umzimkulu became an exclave of the Eastern Cape surrounded by Natal.
Twenty years later, in 1996, the work of the Commission of Inquiry into the Finalisation of the Boundary between the Eastern Cape and KwaZulu-Natal (hence known as the Trengove Commission) was formed. The Trengove Commission heard testimony from affected citizens and made a recommendation: all of the disputed areas should be included into the Eastern Cape. Two of the commissioners, notably those from KwaZulu-Natal, filed a minority report disagreeing with these conclusions.21 Since the start of the Commission, neither the representatives from the Eastern Cape nor those from KwaZulu-Natal ever budged on their initial positions. The final vote was the same as their initial positions, even after eight months of discussion and testimony.
Owing to the longevity of the historic debate over the status of East Griqualand and the tensions associated with this decision, the author recommended to the Commission that a referendum be held and timed to coincide with the May 1996 local government elections. The commissioners responded that their role would be seen as superfluous if a referendum were to be held. Some said that individuals on the ground were not competent to vote on such complex issues because of high rates of illiteracy. The concluding report of May 1996 argued that holding a referendum would be too expensive, logistically difficult, and was not in high demand.22
In mid-May a mass protest of about 2 000 people was held in Kokstad, the main centre of East Griqualand, during which copies of the Trengove Commission report were burned in the streets. White NP supporters marched arm in arm with black ANC members supporting a referendum to redress the lack of consultation with the community. The KwaZulu-Natal legislature, regardless of political party or ethnic affiliation, unanimously rejected the Trengove Commission report. The Eastern Cape parliamentarians favoured it in an equally substantive way.
Thus, a costly commission had nearly no affect on the resolution of the problem. Referenda may be expensive but eight months of commission work is not cheaper. Millions of rands were spent on salaries, security forces, renting halls, secretarial services, publications and press releases, rental cars and four-wheel-drive vehicles, hotels, consulting fees, translators, and airline tickets for both consultants and commissioners to meet around the country. In fact, a question on the boundaries attached to the ballot held in June 1996 would have cost very little in comparison, but the Commission obtained two extensions beyond their original deadline and this cost the opportunity of a referendum in time for the elections.
The use of a commission rather than a referendum also returned the entire decision to a political party process, far removed from people on the ground and fueled a violent reaction. The boundary commission itself heightened tensions by providing a focal point for argument, intimidation, and the mass mobilisation of individuals for and against inclusion in one or the other province.23 Furthermore, costly security problems brought in troops and casspirs. This included numerous petrol-bombing incidents, arson, murder (in one case six people were hacked to death in Umzimkulu),24 land occupations, road blockades, boycotts, and disrupted schools (in the Bushbuckridge dispute, fifteen schools were burnt to the ground).25
The notion that boundaries do not matter also problematises rather than helps to resolve these disputes. Boundaries create the territorial space in which we live, distribute power to people who influence our lives, determine where we vote, create tax bases, construct regional identities, facilitate or impede easy transport, determine access to public services and become blueprints for development planning. Boundaries have enormous social, cultural, economic, and political, ramifications that are felt most deeply in the affected border areas. For instance, twenty years after the incorporation of East Griqualand into KwaZulu-Natal, resentment was still voiced at nearly all the hearings of the Trengove Commission. Cultural memory can last even longer and many ethnic groups, such as the Griqua or Pondo spoke of resentments over boundary disputes dating back to the previous century.
At the basis of all these problems is the fact that no politician, government official or commissioner can actually tell us about the consensus or the will of the affected people. Legitimacy rests with the consensus of the affected people and the best way to measure this is through a referendum. The Commission merely served to delay any resolution, probably into the next century, because after the referendum opportunity elapsed, existing constitutional mechanisms could not solve the problem.
AN ALTERNATIVE: PUBLIC PARTICIPATION BY REFERENDA
An alternative method of boundary construction under time-constrained conditions might be to increase rather than decrease public participation. The advantages of a referendum on long-standing boundary disputes include:
- a decisive answer that can result in a final settlement to a sensitive boundary;
- maximum participation by all affected parties; and
- widely perceived legitimacy.
Where decision-making power is concentrated, political parties will seldom act in the best interests of local peoples if party politics are at stake. A properly structured referendum allows citizens the opportunity to express a democratic will that may differ from their representatives. From empirical studies it is known that public behaviour can be different than private behaviour. Referenda allow individuals to privatise a decision. This makes each voter feel responsible for the referendum result and they subsequently seldom question it. That element of decisiveness should be welcome in situations where there has been long standing political debate or where any recommendation is going to alienate some group to such an extent that it leads to discontent, dissension, mass action, or even violence.
Secondly, referenda automatically include all relevant actors. No one is left out of the decision-making process. This helps to make the result decisive and legitimate. The avoidance of the referendum in East Griqualand means that public debate, acrimony, and related security problems may go on for five or ten more years at great public expense something to remember when calling referenda expensive.
Thirdly, the legitimacy of referenda raises citizen confidence in the democratic system and leads to a pride and stability that would increase the international standing of South Africa. The referendum result is usually accepted owing to perceptions of legitimacy. This is what makes a referendum particularly appealing with regard to the Eastern Cape/KwaZulu-Natal border area. Furthermore, in today’s global village, perceived legitimacy translates into investment, because high levels of participative democracy are linked to stability and economic advancement. In this sense, referenda are a strategic investment with dividends.
The two main disadvantages of referenda are not to the general population, but to politicians and political parties:
- the unpredictability of the result can undermine the strategic plans of political parties; and
- over time, referenda can weaken the political party system to favour direct democracy.
Most international surveys have shown that there is broad support for referenda at the local level, even if the results do not serve the individual questioned.
Throughout the world, referenda are notably unpopular with politicians, because they prefer to keep the decision-making process to themselves. Politicians usually claim that referenda create the divisions that they are meant to resolve, but there are little data to support such claims. However, much of the data indicate that citizens will often make decisions that are unpopular with their representatives. This loss of political power is feared by both individual politicians and political parties, who foresee an erosion in their ability to maintain a hierarchy. Unpredictability can also make the geostrategic manipulation of peoples and territories for political gain more difficult.
The security costs and problems associated with South Africa’s demarcation of nine new provinces resulted from three errors in the manner in which the problem was handled. These included:
- using party-political antagonists rather than referenda to decide on the issue;
- treating boundaries as static rather than dynamic elements; and
- a lack of openness and transparency.
The use of referenda from 1994 could have bypassed all the major problems that characterised both the creation of the provinces and the attempts to resolve ongoing disputes. Referenda, however, are not neutral objects in the construction of bounded spaces. They break the coalition of politicians against public access to the decision-making process and hence could play a role in building a culture of grassroots democracy in South Africa. Therefore, support of referenda to resolve boundary disputes suggests an interest in moving the entire South African polity toward a system of more direct democracy and community empowerment.
Since the provinces are now constitutionally entrenched, the only way forward is in terms of reviewing the relationship between national, provincial, and local government. This is what the Department of Constitutional Affairs is undertaking at present, but it effectively returns South Africa to the drawing boards. The Department and the ANC are strongly backing a redesign of local government structures into large rural municipalities and `megacities’ that reduce provinces to purely administrative units. Regardless of the value of such ideas, new plans come at a great cost both in terms of time and money while the patience of the electorate wears thin.
The financial assistance of the Centre for Science Development (Human Sciences Research Council, South Africa) towards this research is hereby acknowledged. Opinions expressed and conclusions arrived at are those of the author and cannot be attributed to the Centre for Science Development.
- R Griggs, The Boundaries of a New South Africa, Boundary and Security Bulletin, 2(4), Durham, January 1995.
- For a comparison between regional demarcations in Namibia and South African, see Y Muthien, & M Khosa, The Kingdom, the Volkstaat and the New South Africa: Drawing South Africa’s New Regional Boundaries, Journal of Southern African Studies, 21(2) June 1995.
- Republic of South Africa, Report of the Commission on the Demarcation/Delimitation of SPRs, 31 July 1993.
- The National Party government stripped the provinces of their powers in 1986 and replaced them with the short-lived system of regional councils.
- Republic of South Africa, Proposals by the Government of the Republic of South Africa on Boundaries for Regions to Establish Regional Government in South Africa, July 1993.
- R Hulley, Personal communication, May 1994.
- Republic of South Africa, Report on Further Work on the Demarcation/Delimitation of States/Provinces/Regions (SPRs), 15 October 1993.
- M Khosa & Y G Muthien, The Expert, the Public, and the Politician: Drawing South Africa’s New Provincial Boundaries, South African Geographical Journal, 79(1) Cape Town, 1997, pp. 1-12; Y Muthien, Divided We Fall, Indicator SA, 11(3), Durban, Winter 1993, pp. 27-32.
- Republic of South Africa, Constitution of the Republic of South Africa, Act 200 of 1993, Government Printer, Pretoria, Schedule I, Part II, pp. 176-177.
- Now, Let’s Do It, Is Mbeki’s Call, Citizen, 14 May 1997, p. 4.
- L Ensor, Government Absorbs Provincial Debt of R11 Billion, Business Day, 9 October 1997, p. 1.
- R Griggs, Voting Makes a Lot More Sense, Weekly Mail & Guardian, 25 July 1997, p. 27.
- Z Skweyiya, Provincial Review Report, Debates of the National Council of Provinces, 6, 18 September, 1997.
- G Lewis, The Provinces: Can the Crisis be Averted?, RDP Monitor, 3(11), August 1997.
- Interestingly, the March 1993 ANC proposal, prior to multiparty negotiations, was to retain the four provinces until after the 1994 general election. See African National Congress, The Delimitation of Regional Boundaries, Submission to the Delimitation Commission, 107, 1993.
- Republic of South Africa, Debates of the National Assembly, First Session, Second Parliament, 16 May 1997.
- Republic of South Africa, Report of the Commission of Enquiry into The Taung/Kuruman/Kudumane Border Dispute, April 1997.
- Republic of South Africa, Interim Constitution, op. cit.
- Republic of South Africa, Constitution of the Republic of South Africa, Government Printer, Pretoria, 1996.
- African National Congress, ANC National Working Committee Statement, Department of Information and Publicity, 6 May 1997; African National Congress,Statement of the ANC National Executive Committee, Department of Information and Publicity, 14 July 1997; C Madlala, Why the Government Doesn’t Know Where to Draw the Line, Sunday Times, 6 July 1997.
- Commission of Inquiry into the Finalisation of the Boundaries Between the Provinces of KwaZulu-Natal and Eastern Cape, final report, May 1996.
- R Griggs, Cultural Faultlines: South Africa’s New Provincial Boundaries, Indicator SA, 13(1), Summer 1995.
- W Judd, Six People Hacked to Death in Fighting, Mercury, 9 July 1997, p. 5.
- Griggs, 1997, op. cit.