The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers

Journal of Hydrology xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol

The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers Magalie Bourblanc a,b,⇑, David Blanchon c a

CIRAD, UMR G-EAU, F-34398 Montpellier, France University of Pretoria, Centre for Environmental Economics and Policy in Africa, Private Bag X20, Hatfield, Pretoria 0028, South Africa c University of Paris Ouest Nanterre, Faculty of Geography, 200 Av. de la République, 92001 Nanterre, France b

a r t i c l e

i n f o

Article history: Available online xxxx Keywords: South Africa Catchment Management Agencies Inter basin transfers Hydrosocial cycle Problemshed

s u m m a r y The implementation of Catchment Management Agencies (CMAs) was supposed to be the cornerstone of the rescaling process of the South African water reform policy. Yet, less than 10 years after the adoption of the National Water Act, the process was suspended for 4 years and by 2012 only two CMAs had been established. Combining approaches in geography and political science, this paper investigates the reasons for the delays in CMAs’ implementation in South Africa. It shows that the construction of interbasin transfers (IBTs) since the 1950s by the apartheid regime and nowadays the power struggles between CMAs and the Department of Water Affairs (DWA) are two of the main obstacles to the creation of CMAs planned by the 1998 National Water Act (NWA). Finally, the paper advocates taking the ‘‘hydrosocial cycle’’ as an analytical framework for designing new institutional arrangements that will include both rectifying the legacy of the past (the specific role of DWA) and acknowledging legitimate local interests. Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction: rescaling South African water policy With the adoption of the National Water Act (Act 36 of 1998), the South African Government officially recognised a ‘‘need for the integrated management of all aspects of water resources and, where appropriate, the delegation of management functions to a regional or catchment level so as to enable everyone to participate’’ (RSA, 1998: 3). Nineteen CMAs were to have been created across the country. A whole chapter (no. 7) was dedicated to the newly created Catchment Management Agencies, which were supposed to be the cornerstone of this rescaling process. Their five initial functions were to:  investigate and advise on the protection, use, development and control over water in the catchment,  develop a catchment management strategy,  coordinate related activities of water users and institutions,  promote coordination of the implementation of the catchment management strategy with development plans resulting from the Water Services Act, and  promote community participation (RSA, 1998: 88). ⇑ Corresponding author. Address: Department of Agricultural Economics, Private Bag X20, Hatfield, Pretoria 0028, South Africa. Tel.: +27 (0)12 420 5228; fax: +27 (0)12 420 4958. E-mail addresses: [email protected] (M. Bourblanc), dblanchon@ yahoo.fr (D. Blanchon).

The CMAs were supposed to be the symbol of a post-apartheid water policy that would shift the power from the highly centralised and powerful Department of Water Affairs in Pretoria to local communities, especially ‘‘disadvantaged persons or communities which have been prejudiced by past racial and gender discrimination in relation to access to water’’ (RSA, 1998: 90). The Water Act states that ‘‘the Minister must promote the management of water resources at the catchment management level by assigning powers and duties to Catchment Management Agencies’’ (RSA, 1998: 84) [emphasis added] and that ‘‘the purpose of establishing these agencies is to delegate water resource management to the regional or catchment level and to involve local communities’’ (RSA, 1998: 85). Yet, 14 years later, the rescaling process of South African water policy is experiencing tremendous challenges. In a press release in March 2012, the South African Minister of Water Affairs stated: ‘‘The Minister decided to reduce the number of CMAs to nine from the original proposal of 19 CMAs. This is due to a number of reasons including the technical capacity required to staff CMAs, and the challenges such a large number of institutions pose to the Department of Water Affairs (DWA) in regulating their performance’’ (DWA 2012). This statement followed the ‘institutional realignment’ process initiated in 2007 at the national level which unilaterally suspended CMAs’ establishment progress for 4 years. Consequently, as is provided in the 1998 Water Act: ‘‘in areas for which a catchment management agency is not established or, if established, is not functional, all powers and duties of a catchment management agency [. . .] vest in the Minister’’ (RSA, 1998: 82) (see Figs. 1a and 1b).

0022-1694/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

2

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

2. Hypotheses, methods and paper outline The question that we want to address in this paper is straightforward: why has the rescaling process of South African water management failed so far? There can be several explanations, the most cited one referring to problems of policy implementation that impede almost all of South Africa’s policy sectors nowadays, including the water sector (von Holdt, 2010). More closely related to CMAs, a fairly large number papers have been written been written on aspects mostly focusing on poor administration, mismanagement, lack of training of newly appointed public servants or coordination problems (Gorgens et al., 1998; Pegram and Palmer, 2001; Pollard and du Toit, 2008). For instance, according to Mazibuko and Pegram (2006: 1), ‘‘most water resources managers do not know which directorates or departments to make contact with in a local government organisation to achieve cooperation around a specific issue. This leads to frustration, inefficiencies and inadequate cooperation or consultation between the institutions’’. By focusing on the geo-historical background, as well as on contemporary socio-political aspects of the South African water

management policy, this paper aims at shedding new lights on the failure of the implementation of CMAs in the South African context. Our preliminary hypothesis was that the two major obstacles to the ‘‘delegation of management functions’’ to CMAs are (1) the building of interbasin transfers (IBTs) by the apartheid regime from the 1950s onwards and (2) the power struggles in DWA. The arguments that are presented here are based on two distinct methodologies. The first one relies on work that has been conducted in the DWA archives in Pretoria. Internal reports, internal documents and press releases were analysed, as well as official documents. The point was to have a geo-historical view of the South African waterscape, focusing on the implementation of the huge IBT system. We argue here that this historical background is necessary in order to understand the dynamics of CMAs implementation in South Africa. Indeed, as L. Swatuk (2010: 522) pointed out recently: ‘‘present day water policy, practice and management are the results of historical dynamics not easily displaced by generalised discourses of ‘good water governance’. Therefore, understanding South Africa’s complex history of water management, in particular the web of powerful actors, interests and

200km

The 19 original Catchment Management Agencies (with their official number)

1 Limpopo 2 Luvuvhu/Letaba 3 Crocodile West and Marico 4 Olifants 5 Inkomati

6 Usutu to Mhlatuze 7 Tugela 8 Upper Vaal 9 Middle Vaal 10 Lower Vaal

11 Mvoti to Umzimkulu 12 Mzimvubu to Keiskamma 13 Upper Orange 14 Lower Orange 15 Fish to Tsitsikamma

16 Gouritz 17 Olifants/Dorn 18 Breede 19 Berg

In white : limits of the main primary catchments. Fig. 1a. The 19 original Catchment Management Agencies (with their official number).

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

3

200km

The 9 Catchment Management Agencies as defined in 2011 (with their official number)

1 Limpopo 2 Olifants 3 Inkomati/Usuthu/Pongola 4 KZN Rivers 5 Vaal

6 Orange 7 Eastern Cape Coastal Rivers 8 Western Cape South Coast Rivers 9 Western Cape West Coast Rivers

In white : limits of the main primary catchments. Fig. 1b. The 9 Catchment Management Agencies as defined in 2011 (with their official number).

motives that lie at the heart of decision making, is the necessary starting point for uncovering reasonable ways forward for integrated water resource management (IWRM).’’ The second methodology is based on a research project during which one hundred semi-structured interviews were conducted between June 2010 and March 2011 with various stakeholders’ representatives who had participated in the water fora, with CMA staff and CMA governing board members, with Department of Water Affairs officials, as well as with political elites at local and provincial levels. Our central question was: What are the conditions/factors that favour or hamper the successful establishment of CMAs? We expanded the meaning of this establishment process which, in our understanding, consists of two equally important moments: the pre-establishment phase leading to a CMA being officially gazetted and a post-establishment phase consisting of the time lapse between a CMA officially coming into existence and a CMA actually performing its delegated functions. Combining approaches in geography and political science has proved necessary in a bid to account for the apparent contradiction in the South African water rescaling process. Indeed, it is not

sufficient to demonstrate from a geographical viewpoint how river basin management can be at odds with massive state investments in hydraulic infrastructures and plans of river diversion. Molle (2009) recalls how throughout history, especially in the USA in the 19th century, the engineering ethos has given birth to river basin planning approaches, implying that there is no contradiction in principle between these two approaches. Therefore, our ‘‘territorial’’ perspective has to be complemented with an organisational one in a bid to emphasise the contradictions of the inherited South African water management style with the prospects of establishing river basin organisations. After a short presentation of the theoretical background of this study, we will first show that, historically, water resource management has never matched the limits of river basins and now is unlikely to do so, considering the impediments that IBTs are imposing on decentralised resource management. After discussing these impediments in theory, we will turn to the difficulties encountered in the practical implementation of river basin organisations. The objective is to analyse the extent to which the difficulties identified above have materialised in the actual process of CMAs’ creation

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

4

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

and functioning. We will highlight the conflicts between the two competing scales of water resource management, i.e. between the CMAs and DWA. Finally, in the light of these two different approaches of water resource management, we will reflect on the appropriate scale for effectively managing water in South Africa. 3. Theoretical background Although the catchment is seen as the ‘‘natural basis’’ for water management in most IWRM-based water policies across the world, more and more authors, since the seminal work of T. Allan (1998), have criticised this reference to the river basin as, depending on the context, not always being best suited to the situation on the ground. As Venot et al. (2011) have stated: ‘‘A clear discrepancy between the rhetoric and the reality of basin water management worldwide has led many authors to question the river basin approach to sustainable water management on several grounds [. . .] river basins are not neutral spatial units. Contrary to the common hydrologic vision that presents them as natural and non-political, river basins are disputed arenas that overlap political boundaries and are appropriated – or not – differently by different users pursuing multiple, and possibly conflicting, objectives.’’ 3.1. From watershed to problemshed The authors who have denounced the river basin mantra as the new hegemonic discourse have given rise, at a conceptual level, to a criticism of the watershed approach (Swatuk, 2005; Molle et al., 2010). Warner et al. (2008), for instance, have argued that ‘‘river basin boundaries and institutionalarrangements are not natural but matters of choice and contestation. The preference for river basin management is an expression of three fundamental and interlinked choices: (1) the appropriate scale for water management, (2) who decides on the appropriate scale, and (3) how and in which forums these decisions are taken.’’ Similarly, Mollinga et al. (2007: 699) plead a ‘‘problemshed approach’’ and emphasise that in order for a reform to be successful, water management should avoid ‘‘confining the scope of analysis to a hydrologically defined unit. The question regarding the boundaries of a given water management issue, in space, in time and socially, is treated as an open, empirical question in a problemshed perspective, while, in a watershed perspective, boundaries are pre-defined spatially, sectorally and analytically through the primacy of ‘water’.’’ In other words, the hydrological unit of the river basin is not necessarily a panacea in water policies. Criticising the ‘‘social engineering’’ approaches of water management in that ‘‘they do not acknowledge the inherently political nature of reform processes, and do not acknowledge their embeddedness’’ (Mollinga et al., 2007: 705), Mollinga et al. recommend paying particular interest to the historical, social, political, as well as economic context, and then to the institutional embeddedness and power struggle that the introduction of reforms can trigger. In a bid to predict the chance for success of a reform, we can never avoid asking the questions of who benefits and who will bear the cost of change. The boundaries of a problemshed are very likely to exceed the limits of a watershed, and because of the plurality of actors, institutions, values and functions involved in the issue, all of them sometimes overlapping one another, one should be less concerned with the water territory than with the water network (‘‘issue network’’). 3.2. ‘‘Functional spaces’’ In an endeavour to turn to authors that might help us further ground the analysis, we can evoke the concept of ‘‘functional space’’ (Nahrath et al., 2009). As in the problemshed approach,

these authors recognise the institutional plurality surrounding environmental issues and consequently they integrate into their analysis different spaces of relevance for tackling the issue at stake. They evoke three different territories, all of them playing a critical role in solving environmental issues, yet not always overlapping: the territory of the problem, i.e. where the problem physically appeared or where it has been raised; the institutional territory, i.e. the territory of politics or of representative democracy; and the policyshed (Cohen and Davidson, 2011), i.e. the sectoral territory of policy regulation. Nahrath et al. introduce the concept of ‘‘functional space’’ to subsume these different territories into a new socially constructed space. According to them, a ‘‘functional space’’ relates to ‘‘a social space, more or less clearly territorialised, organised around rivalries (and therefore around the public regulating of these conflicts) for the access, ownership and allocation of goods and services [. . .] A functional space is a field wherein a collective problem is being framed and recognised as a public problem. It represents the agreed social or geographical perimeter that is perceived as relevant for managing the problem. It also constitutes the legitimate and appropriate political space in terms of presumed policy efficacy, considering both the formal and informal rules that will help arbitrate among competing interests and views over the problem at stake’’ (Nahrath et al., 2009: 6; our translation). However elaborate and convincing the concept of functional spaces (espaces fonctionnels) might appear, it also raises questions. For instance, when evoking water issues, Nahrath et al. refer to the example of the river basin to illustrate what they mean by adequate ‘‘functional space’’. This reveals the limits to their social constructivism. The river basin is very much a given in their analysis and not a constructed object. Yet, critical political ecology considers that natural resources should not be perceived in a deterministic way. Emphasising the social nature of water, they consider, for example, that ‘‘water is a process rather than a thing’’ (Linton, 2010: 4) and also note the fundamentally constructed nature of territories (Norman et al., 2012). 3.3. A political ecology approach of water resources management: the ‘‘hydrosocial cycle’’ As J. Budds (2009: 420) puts it, we ought to re-conceptualise ‘‘water as a socio-nature [. . .and] move away from thinking of water as a resource that is external to social relations, towards one in which social relations are embedded.’’ A few years earlier, in 1999, E. Swyngedouw introduced the Latourian notion of hybridity in water studies in defining the Spanish waterscape as ‘‘a hybrid, a thing-like appearance that is part natural and part social, and that embodies a multiplicity of historical–geographical relations and processes’’ (Swyngedouw, 1999: 445). The concept of hydrosocial cycle has emerged from research on the political ecology of water by authors working in this tradition, such as E. Swyngedouw, K. Bakker, J. Budds and M. Kaika, to name but a few. Recently, J. Linton, in a conference at the University of Paris 10-Nanterre (2011), defined the hydrosocial cycle as follows: ‘‘The hydrosocial cycle represents and analyses the socio-ecological nature of water, in recognition that hydrological processes are shaped by human activities and institutions, that hydrological data and knowledge are constructed in subjective ways, that water is increasingly recognised as culturally specific, and that the material characteristics of water help shape social relations [. . .].’’ According to J. Budds (2009: 428), the hydrosocial cycle concept has a practical and also analytical value: ‘‘The challenge is to better navigate between the material and the socio-political dimensions of environmental change, in order to reveal the power relations that intersect with biophysical dynamics to produce and reproduce political ecologies. The hydrosocial cycle offers an entry point in this respect.’’

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

In this paper, we will use the concept of hydrosocial cycle to describe the South African waterscape, and to explore, as E. Swyngedouw (2009: 59) wrote, ‘‘the close relationship between hydro-social ordering and political economic configurations, or in other words, between the ’nature of society’ and the ’nature of its water flows’’.

4. Results 4.1. A catchment no more: South African watersheds and IBTs 4.1.1. The complex legacy of apartheid era IBTs In 1994, the new ANC government had to manage the legacy of a huge network of IBTs built from the 1960s to the late 1980s. This network was conceived of in the 1950s – although some plans were drawn earlier – in a bid to cope with the rapid industrial and urban development of post-war South Africa, especially in the metropolitan areas of Johannesburg, Cape Town and Durban. The main outcome of our research in the DWA archives was to discover that, far from being a rational and global plan, South African water policy was always vehemently discussed at the highest level throughout the apartheid era. There was no consensus between supporters of ‘‘water demand management’’, who opposed any further development in water scarce regions, and those who favoured ‘‘water supply management’’ and advocated the transfer of water wherever it was needed. The first option was expounded in 1949 by L.A. Mackenzie, Director of Irrigation, in an address to the Federated Chambers of Industries, who declared: ‘‘the transport of water over great distances may be impracticable or too costly, so we have to go where we have adequate water’’, which meant Zululand, the Limpopo river basin and Transkei. He concluded that ‘‘no new industries requiring large quantities of water should be established along the Vaal River’’. This choice would have necessitated robust water demand management and the development of the future Bantustans. It was coherent with the plan of keeping the African majority out of the white cities. In contrast, D.C. Midgley (1961: 125–126), Professor of hydrology at the prestigious Witswatersrand University, retorted: ‘‘The urban complex centred on the Witswatersrand is the heart of and nerve centre of the whole continent. It would be folly to allow its growth to be restricted by a shortage of water’’. This option was based on water supply management, which implied that the African population would be attracted to the Pretoria–Witswatersrand–Vereeniging region by industrial development, a phenomenon that eventually occurred. Those who favoured water transfers were divided between the white commercial farmers (mainly Afrikaners, who supported, for example, the Orange River Development Project) and the predominantly anglophone capitalist elite who wanted water to flow towards the booming industrial regions, i. e. Johannesburg and other major cities. According to our research in the Department of Water Affairs archives, this debate was never clearly settled. This helps explain the complexity of the South African water infrastructure network and its contradictions with other apartheid spatial policies, such as the development of Bantustans, the forced removals of population, and the implementation of industrial de-concentration points. Finally, water was due to flow towards cities such as Cape Town and Johannesburg, and in the meantime the government planned to remove non-white inhabitants from these cities and send them back to their so-called homelands in the infamous Bantustans. As a consequence, major dams and IBTs were not constructed according to a carefully planned water policy, but for reasons that were loosely linked to water problems. For instance, it could be

5

argued that the Orange River Development Project (ORDP) was partly planned and constructed in response to the Sharpeville uprising in 1960. For more than 12 years (1948–1960), despite local political support, farmers of the Great Fish and Sundays River valleys never managed to place their diversion project from the Orange River on the DWA agenda. Yet the whole project was planned within a few weeks after the uprising. As S.P. Botha declared during the inauguration of the H.F. Verwoerd Dam 10 years later: ‘‘the project was tackled as an act of faith at a time when attacks were being made on us both from outside and within and our achievements have convinced even those who are against us that we are determined to go forward’’ (Botha, 1972). T.P.C. van Roebboeck, who was in the beginning of the 1960s a young engineer, confirmed the link between Sharpeville and the ORDP many years later: ‘‘In the first place, the main reason why the ORDP was undertaken was to restore confidence in the country after Sharpeville [. . .] The decision to embark on it was political, and those entrusted with the task of preparing the White Paper in question (one cannot even call that planning) had to produce their ‘plans’ within weeks’’ (WCD, 2000, Appendix 1). Twenty years later, the Lesotho Highlands Water Project was also included in a geopolitical ‘total strategy’ which was not directly linked to water management: the main objectives for South Africa, then under international sanctions, was to control Lesotho – the treaty was signed after a coup in Lesotho supported by South Africa – and to have access, through World bank loans, to Western banks and contractors (Blanchon and Turton, 2005) (see Fig. 2). The socio-political history of the South African IBT network helps explain the complex geography of the South African ‘‘plumbing system’’ and its contradictory directions. The first generation of IBTs (i.e. the ORDP and connected canals) directed water from the Orange-Vaal to the periphery, such as the Great Fish/Sundays watershed), whereas the second generation (Tugela-Vaal and LHWP) transferred water from the peripheral catchments towards Gauteng. As a result, 17 of the 23 South African primary catchments are now interconnected. For instance, the Rand Water Board, which is the main bulk water supplier for Gauteng, is drawing water from 9 primary catchments in South Africa and also in Lesotho. If one takes into account the smaller Riversonderend–Berg IBT, only 6 major primary catchments are to be considered as ‘‘independent’’ from IBTs in South Africa (the Mzimvubu, Gamtoos, Gourits, Kei, Keiskamma and Olifants). If one takes the original 19 CMAs proposed in 1998, generally speaking, only in the coastal regions do the CMAs’ boundaries follow the limits of the primary catchments. In the country’s interior, the situation appears more problematic since CMAs 9, 10, 14 and 15 are interconnected by the ORDP, CMAs 7, 8, 9, 3, 4, and 5 are linked by the LHWP and Tugela-Vaal projects, and CMA 18 and 19 are joined by the Riversonderend/Berg IBT. Only CMA nos. 12 and 17 are not affected by IBTs; they represent 17% of the total ‘‘natural’’ flow of South Africa, and supply less than 12% of the population. Altogether, the volume of water transferred in South Africa represents 4 km3 per year, 10% of the total ‘‘natural’’ flow regime of South Africa, and, more significantly, 40% of the total water used. To the best of our knowledge, no CMAs have been implemented in such a context in any other country. 4.1.2. IBTs in the post-apartheid era There has been no sign of IBT decommissioning in South Africa since 1994, and on the contrary, new projects are being discussed. According to a DWA internal report, the department should implement the second phase of the LHWP (DWA, 2009). A previous report, in December 2006 (DWA, 2006), also contemplated the augmentation of the inter-basin transfer from the Tugela to the Vaal. Interestingly enough, both reports underline the point that the Vaal is the backbone of almost all of the South African water

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

6

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

The South African waterscape

Main watersheds Vaal sub catchment

Water demand industrial urban 8M. 3,2 M.

Main dams A : Bloemhof B : Vaal C : Grootdraai D : Sterkfontein E : Van der Kloof F : Gariep G : Katse

200 000 . Agricultural water demand Most watered region

Inter basin transfers 1 : Orange River Project 2 : Eastern-Transvaal 3 : Tugela-Vaal 4 : Lesotho Highlands Water Project 5 : Transfert vers Johannesbourg et Pretoria

Watersheds connected to Orange-Vaal

Le Cap

0

200km

Fig. 2. The South African waterscape.

complex and that any study of the ‘‘Integrated Vaal River System’’ should take into account the catchments of the Komati, Usutu, Thukela, Crocodile and Senqu Rivers. As a matter of fact, as shown in Fig. 3, the Vaal dam is still the core of the South African waterscape. Built in 1938, 77 km south of Johannesburg, it receives water coming from the Tugela, the Buffalo, the Usutu and Orange Rivers; major users, such as the Rand Water Board (the bulk water supplier of Gauteng), SASOL (Suid Afrikaanse Stenkool en Olie – originally founded in order to transform South African coal into oil, and now defines itself as an ‘‘integrated international energy and chemicals company’’) and Eskom (the main electricity producer in South Africa) depend on it. Finally, the waste water of these consumers is released in the Crocodile, Marico, Olifants, and Limpopo rivers. It seems inconceivable that the Vaal dam could be managed by the Upper Vaal CMAs: both of the above DWA reports also underline the importance of the ad hoc steering committee for the ‘‘Vaal River System: Large Bulk Water Supply Reconciliation Strategy Study’’ in order to successfully address all the issues of this complex system. Consequently, the major water infrastructure in South Africa is likely to be managed in the future at the national and international (as far as the Lesotho is concerned) levels. In the next section, we examine how the complexities of this waterscape might have impacted on the actual process of CMAs’ creation and functioning.

controlled at a national level’’ (Appendix D, National Water Resource Strategy, DWA, 2004). As was to be expected, in the revised 9-CMAs strategy published in 2012, the Vaal system, which previously was split into 3 WMAs in the initial 19-CMAs plan, has been reunited with the Orange river system. In addition, in the appendix of the NWRS where the scenario for ‘‘interventions for the reconciliation of requirements and the availability of water’’ is presented, the DWA very often mentions the ‘‘national importance’’ of a given river development and the fact that ‘‘large-scale development [. . . should] be made subject to authorisation at national level’’. If we were to give just one illustration of the burden that IBTs can impose on the fate of a CMA, let us mention the case of Crocodile West and Marico WMA which is very dependent upon Vaal transfers. For this reason, the Crocodile CMA ended up encompassing the Johannesburg area in its WMA although it makes little sense to have them part of the same WMA since the North West province and Gauteng have very different economic activities. Local stakeholders have mentioned this lack of homogeneity in the interests at stake and the distance to Johannesburg as an impediment to the mobilisation and collective action in the early phase of the establishment process (interviews, Rustenberg, March 2011).

4.2. IBTs and the challenges of establishing CMAs

4.3. One organisation responsible for water resource management, no more?

Very noticeably, in water management areas (WMAs) where massive IBTs are in place, such as in the Vaal and Orange water systems, CMAs are still in a pre-establishment process. Eliminating any attempt at claiming ownership over strategic resources, the National Water Resource Strategy clearly indicates that ‘‘management of water resources in the Vaal River System is to be

Because of the complexities of the South African ‘‘plumbing system’’, one actor has imposed itself as being particularly central in water resource management in a bid to support this sophisticated architecture: the Department of Water Affairs. In the following paragraphs, we question the role of DWA in the process of establishing decentralised river basin organisations.

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

7

Fig. 3. An illustration of South African interbasin transfers: the Vaal Water System.

With only two CMAs being operational out of the 19 that were initially planned under the National Water Act, it is fair to characterise the introduction of CMAs as a protracted and difficult process, so far. It is not a unique situation as far as the implementation of the NWA is concerned; the implementation of other provisions of the Act is also lagging behind schedule (Kemerink et al., 2012). However, we have argued earlier that the difficulties encountered in the implementation of this CMA provision have a lot to do with the impractical nature of any attempt at creating river basin organisations in a country characterised by massive IBTs. We now turn to another explanation: the perceived threat that such decentralised organisations might represent to DWA institutional power and its centralised resource management routine. 4.3.1. A delayed process of establishing CMAs as an illustration of DWA’s reluctance towards them? Examining in more detail the exact progress in terms of establishing CMAs to date is illustrative of DWA’s reluctance towards implementing CMAs. In 2004 when the DWA set out a more detailed programme for establishing CMAs, its National Water Resource Strategy (NWRS) stated that: ‘‘five water management areas have been identified where the establishment of Catchment Management Agencies is urgent. These are the Inkomati, Olifants, Breede, Crocodile West and Marico, and Mvoti to Mzimkulu.’’ As criteria for prioritising the establishment of CMAs, the NWRS mentions elements such as ‘‘the extent to which service delivery will be enhanced by the establishment of an agency’’, ‘‘anticipated revenue from water resource management charges to fund the agency’s operations’’, ‘‘stakeholder preparedness’’, ‘‘priority for compulsory

licensing’’, ‘‘priorities under the Integrated Rural Development Programme’’. Therefore, we can see that institutional difficulties that might undermine the establishment process were considered from the very beginning and that DWA cannot be blamed for any excess of optimism in that respect. A few years later, in 2006, the Minister of Water Affairs announced that ‘‘four processes are at the point where the Agencies are close to being established after having completed a period of public comment on the establishment proposals. These four are the Thukela, the Usuthu to Mhlatuze, the Gouritz and the Olifants-Doorn. They will be established later this year and the Advisory Committee processes and appointment of the Governing Boards will be completed in 2007’’ (Internal question paper no. 13 replying to parliamentary question no. 486, 26 May 2006). Yet, 4 years later the new Minister of Water Affairs could only state during a parliamentary session that the already gazetted ‘‘Mvoti, Crocodile West and Marico, Olifants-Doorn, Gouritz, Thukela and Usutu CMAs [. . .] have finalized their advisory committee processes. We are currently in the process of recommending which sectors should be represented on the board’’ (Internal question paper no. 14 replying to parliamentary question no. 1575, 21 May 2010). In other words, no progress had been made since 2006. According to NWRS (2004: 120), in 2004 eight CMAs had been gazetted, suggesting that considerable progress had been made since such a formal establishment can only take place after a viability study has been performed and after a public participation process with water forums and stakeholders. This has been confirmed by some of our interviewees from two of the most pro-active DWA regional offices which had started on the

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

8

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

establishment process quite early: in Mpumalanga province, Olifants was making good progress but its progression was stopped by the head office in step with the initiative on the institutional realignment. The same was the case for Thukela and Mvoti, which were both very advanced in the process and had submitted their proposal for the set-up of a governing board to Pretoria, but which have received no feedback on their plan since 2006 (interview, Nelspruit, February 2011; interview, Durban, August 2010). This apparent reluctance from DWA was confirmed by several of our interviewees when trying to explain the delay in establishing CMAs or the delays in the delegation of full functions to more advanced CMAs such as BreedeOverberg. These interviewees mentioned a perceived reluctance from either regional or head offices of DWA (interviews, Worcester, August and November 2010). However, it is not completely clear whether or not this reluctance originates in the administrative arm of DWA. It seems that the political level also played a critical part in the delaying process. For instance, a principled defiance towards CMAs can be noticed in the fact that internal audits have always been ordered after a new political team has come into power in the Department of Water Affairs. Nevertheless, the results of these audits, conducted either by external consultants or DWA internal experts, always concluded that the CMAs were relevant. This at first sight seems to indicate that such defiance came predominantly from the ruling party, and less from the management side within the DWA head office. Indeed, it is well-known that ANC functioning is very centralised and lacks trust towards organisations such as parastatals which grow outside the state and are perceived as a potential threat to its authority (Gumede, 2005). This partly explains the initiative on institutional realignment in the water sector which can largely be understood as a political process that not only concerns CMAs but also Water User Associations or institutions responsible for national water resources infrastructures and especially Water Boards. The number of these institutions had to be reduced in order to ensure ‘‘economies of scale’’ and improved ‘‘regulation’’ or ‘‘clear accountability’’ as represented in the official discourse. The process started in 2007 and ended in 2011 with the decision to retain only 9 CMAs. Yet there is hardly any evidence that attention was paid to the issue of compatibility between these competing approaches of water resource management in the choice of these 9 CMAs. In contrast, what can be observed is a constant attempt at undermining CMAs’ future potential power. In that respect, if these 9 CMAs correspond to the number of provinces, CMAs’ boundaries do not match provincial boundaries at all. This mismatch can be understood as a way for the national level to hamper any potential influence of provincial governments over these institutions: the influence of a provincial government over a CMA straddling several provinces is very likely to be reduced while the national level appears as the designated referee that arbitrates any potential coordination problem. More specifically, DWA reluctance towards CMAs can be observed at the design level of CMAs boundaries and in the division of tasks between DWA and CMAs. 4.3.2. A strategic delineation of CMAs boundaries The list of Water Management Areas (WMAs) was published in the Government Gazette in 1999 (1 October 1999, Notice no. 1160). It is interesting to pay attention to the debate that preceded the design of these WMAs – the territory a Catchment Management Agency is responsible for. Official criteria mentioned comprise not only ecological (‘‘watercourse catchment boundaries’’) but also economic (‘‘social and economic development patterns’’), efficiency (from both an expertise and financial viewpoint) and collective action levels (‘‘the existence of a communal interest in the area’’) (www.dwaf.gov.za/CM/Docs/WORD/finalwma.doc). South Africa is far from being a unique case where the ecological criterion would be counter-balanced with other criteria in the design of

‘‘river basin’’ management area. However, the debate within the DWA over the number and therefore over the size of CMAs is indicative of the power dimension that was taken into account in this particular case and this added to the usual criteria considered elsewhere in the design of river basin organisations. Officially, DWA policy documents mention the fact that too small and therefore too many CMAs having to report to DWA would make the situation unmanageable, in addition to the financial cost of such a solution (www.dwaf.gov.za/CM/Docs/WORD/finalwma.doc). In addition, the risk of having ‘‘special interest groups dominate’’ CMAs is also evoked as a counter-argument against having too small CMA units (DWA, August 1999). Authors, such as Brown (2013), have alleged that DWA’s willingness to defend the interests of small-scale farmers and its intent to speak on their behalf rather than letting them speak for themselves (de Lange, 2004) explain the attitude of DWA in not being willingto delegate its prerogatives to CMAs, considering the risk of ‘‘interest capture’’ (Woodhouse, 1995). On the other hand, DWA did not have any incentive for having too big, and thus too powerful, CMAs challenging its authority. We can interpret in that way the controversy around the design of the new CMA map in Western Cape. Locally, there were many people in favour of a unique CMA for the whole province which would have gathered the dominant commercial farmers from the Breede with the powerful city of Cape Town within the same organisation. The DWA opposed such a possibility, even though this solution made perfect sense, considering the provincial water system (interview with a consultant who has extensively worked for DWA, especially in support of the establishment of BreedeOverberg CMA and Inkomati CMA, Marseilles World Water Forum, March 2012). Finally, the DWA head office succeeded in having BreedeOverberg merge with Gouritz CMA, and the Berg, where Cape Town is located, merge with Olifants-Doorn CMA. By splitting these two strong players into two CMAs, and knowing that Cape Town covets water from the Breede, the DWA adopted a divide and conquer approach and ensured that it would impose itself as a natural referee between these now competing interests. 4.3.3. Sharing or competing over responsibilities? What is also at stake with the creation of CMAs is the sharing of responsibilities and therefore of institutional power over the regulation of water resources. Indeed, although the CMAs will remain formally placed under the authority of DWA, once fully delegated to them, the functions of CMAs will be far from insignificant. CMAs’ competences could extend to the ‘‘ability to authorise, license and regulate water use’’ in their WMA once a CMA is declared the ‘‘responsible authority’’ (Official policy document, ‘‘Implementation of Catchment Management in South Africa: the National Policy – First edition’’, 15 March 2001: http://www.dwaf.gov.za/CM/ wua.htm). This corresponds with core competences of DWA which cover policy development, the formulation of the National Water Resource Strategy with which CMAs’ catchment resource strategy has to abide, and the specification of resource-directed measures. The eventuality of CMAs disputing the responsibilities of DWA is not hypothetical as even core competences of the strategic planning division, such as the so-called ‘‘operating rules’’ prerogatives, can sometimes be exercised by some advanced CMAs, as in Inkomati. For instance, Inkomati CMA has built a sophisticated information model that helps with the monitoring of a more refined ecological reserve. Because of that, Inkomati CMA managed to obtain a function that is usually performed by DWA’s planning division (interview with Inkomati CMA acting CEO, Nelspruit February 2011). Another illustration of CMAs’ consolidating power can be found with the BreedeOverberg CMA. Prospects for future transfer of water from the Breede to Cape Town in a bid to cover the needs of the expanding urban area had met with the resistance of the BreedeOverberg CMA which does not want ‘‘its’’ water to be

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

confiscated. The CMA sought a new audit of water availability from DWA head office as the data about water users and the level of the ecological reserve used by the Strategic Planning unit in Pretoria were not up to date. A feasibility study has now been commissioned in order to find alternative transfers from another WMA and cater for the needs of Cape Town. In any event, this project has disappeared from the list of priority projects for the DWA. It shows that more and more, DWA has to consult and sometimes accommodate its strategic plan with local interests expressed by established CMAs (ibid.). The uncertainty surrounding the actual task divisions between DWA and the CMAs is sharpened in the case of DWA regional offices (DWA RO), especially seeing that ‘‘Once all relevant catchment management functions have been assigned or delegated to the CMA, the primary role of the DWA RO would be to ensure cooperation between CMAs and audit the implementation of catchment management by the CMA.’’ (official policy document, ‘‘Implementation of Catchment Management in South Africa: the National Policy – First edition’’, 15 March 2001: http://www.dwaf.gov.za/CM/ wua.htm) Yet, CMAs are supposed to report directly to the head office which is also directly involved in the auditing of CMAs (in terms of their compliance with national policies, NWRS, RDMs and CMSs). Therefore, the shrinking of DWA RO staff appears inevitable. This event was already envisioned and most of its staff were supposed to join the CMA. However, owing to the perceived uncertainty around CMAs’ viability, most of the DWA RO staff were dissuaded from taking a chance with their careers and therefore staff transfer to CMAs has not happened yet (interview with BreedeOverberg CMA manager, Worcester August 2010). Thus, we see that the defiance of DWA towards CMAs has largely controlled the fate of CMAs. Indeed, the inability to attract the right level of expertise was evoked as one major challenge of CMAs’ establishment, alongside financial limitations (internal question paper no. 14, replying to parliamentary question no. 1575, 21 May 2010).

5. Discussion The fact that the Minister of Water Affairs publicly announced in a media release in mid-March 2012 that a new institutional momentum would be granted to the remaining 9 CMAs so that they could become operational as soon as possible, does not change our statement about DWA’s reluctance shown towards CMAs. Indeed, before the ANC elective conference in Mangaung in December 2012, the Minister was acting under the political pressure of delivering on the promises made in terms of transforming water resource management in a post-apartheid context. Therefore, it seems that the coexistence of CMAs with a centralised approach to water resources management with all its challenges will remain acute for the years to come. Yet it seems that there is little awareness or at least recognition of the existence of these competing scales of management. Indeed, no real plan has been considered in order to reconcile these two management approaches and build bridges across the scales. Inasmuch as the White Paper on a National Water Policy for South Africa 1997 (point 6.6.3) acknowledged that: ‘‘Inter-basin transfers will have to meet special planning requirements and implementation procedures, which must involve agencies from both the donor and recipient catchments’’, it seems that in practice no such coordination mechanism has been put in place. Another good illustration of that lack of consideration for coordination across governance scales can be found in the decision to withdraw the DWA representative from the governing board of BreedeOverberg CMA. The nomination of such a representative sitting on the governing board was only temporary, and meant to accompany the CMA installation process. However, since his

9

mandate was not considered legitimate by other stakeholders’ representatives, his presence was not retained once the CMA gained more autonomy from the DWA regional office (interviews, Worcester November 2010). How then will IBTs be treated in the new CMAs? The first option is to treat transferred water just as an input/output equivalent to runoff (for incoming IBTs) and losses (for outcoming IBTs). This option is definitely the simplest one. It could avoid institutional complexity, because, as Gupta and Van der Zaag (2008: 37) have stated, ‘‘In fact a doubling of spatial scale may result in a quadrupling of institutional complexity [. . .] Interconnecting two river basins would then give rise to a proliferation of coalitions and/or conflicts.’’ Yet, it will deprive some CMAs – for instance the Limpopo CMA – of the control of the major part of their incoming water resources which originate from the Vaal water system. Major actors, such as the Rand Water Board, which operates in three new CMAs, will clearly not be satisfied by this option: this Water Board, which supplies 11 million users with bulk water, obtains water from the Orange, the Tugela and Inkomati CMAs and its waste water is returned to the Limpopo and Olifants CMAs. A lack of coordination between CMAs would potentially threaten its supply. Other actors, such as Metropolitan Councils and commercial farmer associations, also now only reluctantly consider not having any control over water transferred out of the CMAs they belong to. Indeed, while we have emphasised the reluctance of the DWA shown towards CMAs, arguing that a strong political commitment will be necessary to impose CMAs in the South African water institutional landscape, we also note that other actors or interest groups – commercial farmers being one of them – are strong supporters of a decentralised organisation which they perceive to be a good way to secure for themselves a better say in water allocation and licensing matters (Waalewijn et al., 2005; Bourblanc, 2012). The second option is to devise new institutional arrangements. The ‘‘hydrosocial cycle’’ (Linton, 2011; Swyngedouw, 2009) is an appropriate analyticalframework that can be used to help manage the South African waterscape. It requires acknowledging that for any ‘‘rescaling’’ process:  the South African waterscape is highly modified and corresponds exactly to a ‘‘hybrid’’, i.e. it is part natural, part social. This necessitates taking into account both ‘‘natural’’ and ‘‘artificial’’ flows of water as part as the same ‘‘hydrosocial cycle’’. From this perspective, one should not divide Gauteng into different CMAs, but rather consider it as one hydrosocial cycle, using water from upper Lesotho, the upper Tugela valley, and connecting it, through its effluents, to the Crocodile and Marico catchments;  the shape of the South African waterscape is the result of political struggles, rooted in the history of colonialism, segregation and apartheid, but also in a professed ‘‘hydraulic mission’’ accompanying the state-building process (Molle et al., 2009). This past has conferred a prominent role on the Department of Water Affairs that cannot be ignored if one wants to understand the chance of success of a given reform or new institutional arrangement. Yet, as much as the concept of the hydrosocial cycle opposes natural determinism with regard to environmental issues, the same should hold as well for socio-political matters: recognising the legacy of the past, the structural effects of vested interests and/or institutional inertia does not imply that no change or reform is possible. Considering the degree of inequality associated with the apartheid water policy and the extent to which local voices have been marginalised, change is especially desirable in the South African context. Democratic reform has to enable marginalised local interests to voice their concerns and needs. Gupta

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

10

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

Second Stage Reconciliation Strategy

Vaal River System: Reconciliation Strategy Study

Existing Transfer Schemes Proposed Transfer Schemes

Mokolo Catchment

Crocodile Catchment

Vaal Catchment

Vaal Second Stage Reconciliation Strategy Report_v20.doc

A-1

2009/11/06

Fig. 4. Map of the Integrated Vaal River System after Vaal River System: Large Bulk Water Supply Reconciliation Strategy (DWA, 2009).

and van der Zaag (2008) rightly question the compatibility between IBTs and values underlying IWRM, particularly the attention that has to be paid to the local problématique. The Steering Committee of the Large Bulk Water Supply Reconciliation Strategy of the Vaal River System (2005–2009) provides an illustrating example. Indeed, the composition of this Steering Committee indicates that the Department is negotiating with interest representation (not especially linked to the catchment boundaries) in which half of the 53 members come from DWA central office, 8 from municipalities, 6 from water user utilities such as Rand Water, and it also includes one representative each from Eskom, the Chambers of Mines and the Chambers of Business (DWA, 2009: 1). Thus, when the DWA makes room for local interest representation in some of its decision arenas, it still secures most of the seats for itself. Integrating local views in water management does not have to copy the blueprint of a decentralised organisation in every case, though. A CMA is not a guarantee for addressing pressing political issues (Bourblanc, 2012). Thinking away from the search for the perfect scale and/or organisation, advocates of the problemshed approach concentrate instead on ‘‘issue networks’’. However, talking about ‘‘network’’ in an environment with such rivalries between powerful actors does not seem adequate. Ostrom’s notion of ‘‘institutional arrangements’’ (2005) does not fit better in our case as it relies on a rational choice and individualistic postulate that does not correspond to a society like South Africa’s that is strongly divided along ethnic lines. Surely enough, there is no task-specific governance level for the South African waterscape, therefore, new arrangements will have to emerge in a multi-level governance context (Hooghe and Marks, 2003). In some cases where a CMA is already a reality, these new arrangements might

espouse the contours of a decentralised organisation, in others it will not (see Fig. 4).

6. Conclusion Our objective in this paper has been to demonstrate the challenges that water reform faces in South Africa with regard to (irrelevant) spatial scales and (competing) levels of governance. We discussed the rescaling process of water resource management from both a territorial and an organisational perspective. By depicting the South African hydro-social cycle rather than its hydrological cycle, we have highlighted the paradox of retaining the river basin scale as a meaningful territory for management purposes and we particularly emphasised the problems that setting up an organisation at the river basin level inevitably raises. Considering the South African hydro-social cycle, we realise how the legacy of the past has shaped the water resource itself, i.e. in its physical/ hydrological dimension, rendering the river basin less meaningful than elsewhere. Through the hydro-social cycle concept, we also highlighted the dominant power position acquired by DWA over time in the management of water resources. Moreover, we realised that finding the right scale or territory to manage water resources might be a vain search in the South African context, considering its complexity. Yet, as much as the hydro-social cycle concept opposes natural determinism with regard to environmental issues, the same should hold as well for socio-political matters: recognising the legacy of the past, the structural effects that vested interests exert on the South African waterscape and/or institutional inertia does not imply that no change or reform is possible. Considering

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001

M. Bourblanc, D. Blanchon / Journal of Hydrology xxx (2013) xxx–xxx

the degree of inequality associated with the apartheid water policy, change is especially desirable in the South African context, and a democratic reform has to enable local interests to voice their concerns and needs. However, for this policy change to happen, building new organisations at a decentralised level might not always be required. Agreeing on and designing new institutional arrangements might suffice. Acknowledgement The authors would like to thank Jamie Linton for his comments on an earlier draft of this paper. References Allan, A.J., 1998. Watersheds and problemsheds: explaining the absence of armed conflicts in the Middle East. Middle East Review of International Affairs 2 (1). Blanchon, D., Turton, A., 2005. Les transferts massifs d’eau en Afrique du Sud. In: Lassere, F. (Ed.), Transferts massifs d’eau: outils de développement ou instruments de pouvoir ? Presses universitaires du Québec, Québec, pp. 247– 283. Botha, S.P., 1972. Message from the Honourable S.P. Botha, MP, Minister of Water Affairs. Die Siviele Ingenieur in Suid Afrika, février 1972, vol. 14, pp. 31–32. Bourblanc, M., 2012. Transforming water resources management in South Africa. ‘Catchment Management Agencies’ and the ideal of democratic development. Journal of International Development 24, 637–648. Brown, J., 2013. Can participation change the geography of water? Lessons from South Africa. Annals of the Association of American Geographers 103 (2), 271– 279. Budds, J., 2009. Contested H2O: science, policy and politics in water resources management in Chile. Geoforum 40 (2009), 418–430. Cohen, A., Davidson, S., 2011. An examination of the watershed approach: challenges, antecedents, and the transition from technical tool to governance unit. Water Alternatives 4 (1), 1–14. de Lange, M., 2004. Water policy and law review process in South Africa with a focus on the agricultural sector. In: Mollinga, P.P., Bolding, A. (Eds.), The Politics of Irrigation Reform: Contested Policy Formulation and Implementation in Asia, Africa and Latin America. Ashgate, Aldershot (Engl.), pp. 11–56. DWA – Department of Water Affairs, 2012. Minister Establishes Nine (9) Catchment Management Agencies, March 30th 2012. DWA, Pretoria. DWA – Department of Water Affairs and Forestry, 1999. Catchment Management Agencies and Water Users Associations. Water Act News, November Issue. DWA – Department of Water Affairs and Forestry, South Africa, 2006. Vaal River System: Large Bulk Water Supply Reconciliation Strategy: First Stage Reconciliation Strategy, December 2006. DWA – Department of Water Affairs and Forestry, South Africa, 2009. Vaal River System: Large Bulk Water Supply Reconciliation Strategy: Executive Summary. DWAF Report Number: P RSA C000/00/4406/09. Gorgens, A. et al., 1998. Guidelines for Catchment Management to Achieve Integrated Water Resources Management in South Africa, Pretoria. Water Research Commission (WRC Report KV 108/98). Gumede, W.M., 2005. Thabo Mbeki and the Battle for the Soul of the ANC. Zebra Press, Cape Town. Gupta, J., van der Zaag, P., 2008. Interbasin water transfers and integrated water resources management: where engineering, science and politics interlock. Physics and Chemistry of the Earth 33, 28–40. Hooghe, L., Marks, G., 2003. Unravelling the central state, but how? Types of multilevel governance. American Political Science Review 97 (2), 233–243. Kemerink, J.S., Mendez, L.E., Ahlers, R., Wester, P., Van der Zaag, P., 2012. The question of inclusion and representation in rural South Africa. Water Policy 15, 243–257.

11

Linton, J., 2010. What is Water? The History of a Modern Abstraction. UCB Press, Vancouver, 334 p. Linton, J., 2011. The hydrologic cycle and the hydrosocial cycle: bridging hydrosystems and hydropolitics, conference held at Paris Ouest Nanterre La Défense. In: Séminaire Hydrosystèmes et hydropolitiques, Nanterre, June the 7th 2011. Mackenzie, L.A., 1949. The Water Resources of the Union in Relation to the Development and Decentralisation of Industries. Address Delivered at the 32nd Annual Convention of the S.A. Federated Chamber of Industries at Johannesburg. Mazibuko, G., Pegram, G., 2006. Guide for Catchment Management Agency Cooperation with Local Government. WRC Report No. TT 271/06. Water Research Commission (WRC), Pretoria. Midgley, D.C., 1961. A water policy for the Vaal. In: Symposium on the Future of the Vaal River. Vaal River Catchment Association, Johannesburg, pp. 125–134. Molle, F., 2009. River-basin planning and management: the social life of a concept. Geoforum 40, 484–494. Molle, F., Mollinga, P., Wester, P., 2009. Hydraulic bureaucracies and the hydraulic mission: flows of water, flows of power. Water Alternatives 2 (3), 328–349. Molle, F., Wester, P., Hirsch, P., 2010. River Basin Closure: processes, implications and responses. Agricultural Water Management 97 (4), 569–577. Mollinga, P., Meinzen-Dick, R., Merrey, D., 2007. Politics, plurality and problemsheds: a strategic approach for reform of agricultural, water resources management. Development Policy Review 25 (6), 699–719. Nahrath, S., Varone, F., Gerber, J.-D., 2009. Les espaces fonctionnels: nouveau référentiel de la gestion durable des ressources?. VertigO – la revue électronique en sciences de l’environnement, vol. 9, Numéro 1, mai 2009. Norman, E.S., Bakker, K., Cook, C., 2012. Introduction to the themed section: water governance and the politics of scale. Water Alternatives 5 (1), 52–61. Ostrom, E., 2005. Understanding Institutional Diversity. Princeton University Press, Princeton. Pegram, G., Palmer, I., 2001. Guidelines for Financing Catchment Management Agencies in South Africa, Pretoria. Water Research Commission, Report No. 1044/1/01. Pollard, S., Du Toit, D., 2008. Integrated water resource management in complex systems: how the catchment management strategies seek to achieve sustainability and equity in water resources in South Africa. Water SA 34 (6), 671–680 (IWRM Special Edition). Republic of South Africa, 1998. National Water Act, Act No. 36 of 1998, Government Gazette, 26 August 1998. Government Printers, Pretoria. Swatuk, L.A., 2005. Political challenges to implementing IWRM in Southern Africa. Physics and Chemistry of the Earth 30, 872–880. Swatuk, L., 2010. The state and water resources development through the lens of history: a South African case study. Water Alternatives 3 (3), 521–536. Swyngedouw, E., 1999. Modernity and hybridity: Nature, regeneracionismo, and the production of the Spanish waterscape, 1890–1930. Annals of the Association of American Geographers 89 (3), 443–465. Swyngedouw, E., 2009. The political economy and political ecology of the hydrosocial cycle. Journal of Contemporary Water Research and Education 142, 56– 60. Venot, J.-P., Bahrati, L., Giordano, M., Molle, F., 2011. Beyond water, beyond boundaries: spaces of water management in the Krishna river basin, South India. The Geographical Journal 177 (2), 160–170. von Holdt, K., 2010. Nationalism, bureaucracy and the developmental state: the South African case. South African Review of Sociology 41 (1), 4–27. Waalewijn, P., Wester, P., van Straaten, K., 2005. Transforming river management in South Africa. Lessons from the Lower Komati River. Water International 30 (2), 184–196. Warner, J., Wester, P., Bolding, A., 2008. Going with the flow: river basins as the natural units for water management? Water Policy 10 (Suppl. 2), 121–138. World Commission on Dams, 2000. Orange River Development Project, South Africa. WCD, Cape Town, 122p. Final Report, November 2000. . Woodhouse, P., 1995. Water rights and restructuring in South Africa: a case study from Eastern Transvaal. Water Resources Development 11 (4), 527–544.

Please cite this article in press as: Bourblanc, M., Blanchon, D. The challenges of rescaling South African water resources management: Catchment Management Agencies and interbasin transfers. J. Hydrol. (2013), http://dx.doi.org/10.1016/j.jhydrol.2013.08.001