Journal ArticlesPublicationsUncategorized

Decentralisation and recentralisation in South Africa’s local government: case studies of two municipalities in Limpopo

By Thokozani Chilenga-Butao


Democratic decentralisation was introduced in South Africa during the transition to democracy (1990-1994). It followed a long trajectory of centralisation and decentralisation processes that took place during apartheid. This paper argues that in order to more adequately understand the prospects for decentralisation to achieve its intended outcomes in South African local government, one has to understand some of the complexities and political dynamics present in this sphere of government. In so doing, it shows that the intended outcomes of decentralisation are far from the realities of local government on the ground, specifically municipalities. Case studies of two Limpopo mining town municipalities, Lephalale and Mogalakwena, are used to demonstrate some of these complexities and political dynamics. The Mogalakwena case study will show that, despite the codification of recentralisation in the South African constitution, regional and political party elites misuse the policy to politically interfere in municipalities. The effects of this are that service delivery slows down and local government is subjected to localised national and provincial political battles. The Lephalale case study shows how the layers of decentralisation between apartheid and democracy have led to this municipality being dependent on private and parastatal mining companies for the provision of and access to public goods and services.


Governance in South Africa is structured around three spheres – national, provincial and local. Local government, particularly with respect to municipalities, is structured to be a close interface between the state and society, mandated to deliver basic services to their respective communities.

This paper will demonstrate the complexities of democratic decentralisation in South Africa and a more recent trend of recentralisation, and show that local government, as a product of decentralisation, has been unable to meet the intended outcomes of this form of governance and state organisation. These complexities include the misuse of legislation that sets the parameters for power and authority amongst the spheres of government, and the historical legacies of municipalities. The aim of this paper is to provide a nuanced reflection of decentralisation in the South African context and to use empirical data to argue that decentralisation and recentralisation must be improved on constantly in order to fulfil the functions of local government. This paper lays bare some of the administrative complexities in municipalities in South Africa, and thus complicates the scholarship on decentralisation and recentralisation.

Decentralisation is described as ‘any act in which a central government formally cedes powers to actors and institutions at lower levels in a political-administrative and territorial hierarchy’ (Ribot 2002:v). The most common actors in decentralisation are subnational governments – such as provincial and local governments in South Africa. There are three types of decentralisation: administrative, fiscal and political. Administrative decentralisation is the process of delegating ‘central government authority and responsibility’ to subnational governments (Cheema and Rondinelli 2007:6-7). Fiscal decentralisation describes the legislative and practical arrangements for funding subnational governments, including the ability of subnational governments to raise their own revenue (Cheema and Rondinelli 2007:7). And political decentralisation describes ‘increasing citizen participation in selecting political representatives and making public policy’ (Cheema and Rondinelli 2007:7) in subnational governments. Decentralisation has been at the core of the making of the spheres of government in South Africa and empowering local government. The constitutional and legislative structure of decentralisation is codified in the Constitution of the Republic of South Africa, 1996, the Municipal Structures Act, 1998, and the Intergovernmental Relations Framework Act, 2005, amongst others. Importantly for this paper, Schedule 4 Part B1 of the Constitution, together with S83 – S842 of the Municipal Structures Act, outline local government’s responsibilities, which include providing basic public goods and services. This legal framework structures the relationship between national government and local municipalities, and is intended to create the conditions for the delivery of public goods and services in a localised way, tailored to local needs.

But the vision of decentralisation has not been translated into concrete reality in municipalities. Rather, what has occurred are attempts to rein in local government power through forms of recentralisation. Recentralisation is the process of national governments rescinding the devolved powers of subnational governments when they are unable to provide public goods and services to their localities. There are a number of reasons that subnational governments are unable to meet the service provision needs in their localities, including lack of fiscal autonomy (Dickovick 2011), lack of capacity and lack of proper regulation (Erk 2014:540-2, Olowu and Wunsch 2003:6).

This paper will demonstrate that decentralisation is not a straightforward process, owing in great measure to the historical legacy of apartheid’s historical experiments in decentralisation. Additionally recentralisation either becomes an option for municipalities or is used as a tool of political interference in other municipalities. Using two case studies of mining town municipalities in Limpopo, Mogalakwena and Lephalale, this paper will challenge the officially held views of the intended outcomes of decentralisation by showing how municipalities have had to contend with the complexities of decentralisation in their specific localities.

Decentralisation should be understood from two perspectives. Firstly, local government obviously has a long history that predates the democratic period. Secondly, the political-administrative interface has limited the ability of municipalities to carry out their duties and effectively provide the goods and services listed in Schedule 4, Part B of the Constitution. The interaction between the politics and the administration and the history of local government has also complicated decentralisation. In principle, ‘decentralisation’, refers to the act of a national government devolving administrative, fiscal and political power to subnational governments (Siddle and Koelble 2012, Olowu and Wunsch 2003, Ribot 2000). Subnational governments can be located in regions or provinces – the intermediate level between a national government and a local government – or located in more local and district areas. Crook and Manor (2000:11-12) describe the advantages of democratic decentralisation as being transparency, openness and accountability. Similarly, Siddle and Koelble (2012) summarise the objectives of decentralisation as the promotion of democracy, legitimacy, public participation, developmentalism, responsiveness, efficiency, and competition; reducing corruption; improving communications; and defusing conflicts, usually racial or ethnic conflicts (Siddle and Koelble 2012:35- 38).

This paper argues that in order to more adequately understand the prospects for decentralisation to achieve its intended outcomes in South African local government, one has to understand some of the complexities and political dynamics present in this sphere of government. There is a great disparity between the nature and structure of this: there is a world that separates South Africa’s de jure approach to decentralisation and the reality of how decentralisation operates within South Africa’s local government. This disparity is sometimes caused by the lack of resources within local government to fund local sphere responsibilities, despite the fiscal decentralisation that allows them to collect their own revenues for this purpose. Sometimes this disparity arises from a lack of skills and capacity to run municipalities through proper planning, deficient policy implementation and inadequate responsiveness to local challenges (Smoke 2003:10, Dickovick 2011, Erk 2014:542-3, Koelble and Siddle 2014:1126-28). But the disparity is also caused by the historical legacies that are found within local government, as well as more immediate political causes. The connection between these historical legacies and political causes is underexplored.

Scholarship often understands decentralisation as technocratic or administrative reforms intended to make government more efficient and effective (Smoke 2003, Ribot 2002, Crook and Manor 2000, Manor 1999). This is underpinned by theories that argue that distribution in society is achieved best by local governments that are closest to communities and therefore have the ability to directly and appropriately provide for the needs of local constituencies, oversee equity through proper distribution of local resources, ensure better service provision as  a consequence of competitive behaviour between private-sector companies, non-governmental organisations and local administrations, and allow participation and democratisation as a result of local decision making (Ribot 2002:8-10). Smoke (2003) agrees that decentralisation has the ability to generate ‘improved efficiency’, ‘improved governance’, ‘improved equity’, ‘improved development and poverty reduction’ (Smoke 2003:9-10). Moreover, according to developmental state theories, decentralised local governments are instrumental in the development of local communities and the state as a whole (see Saloojee 2012, Bardhan and Mookherjee 2006).

Research into the Lephalale and Mogalakwena local municipalities explored in this article was conducted between 2015 and 2017, as part of a local government research project at the Public Affairs Research Institute (PARI). This article recognises that apart from the findings by the researchers and authors, the two municipalities offer further opportunity for analysis from a different vantage point, considering the common thread of decentralisation that runs through the histories and current issues in both municipalities. This article uses primary data, such as interviews with municipal officials, as well as the reports that emanated from this research into Lephalale and Mogalakwena, and draws further conclusions about the dynamics of decentralisation and recentralisation in South Africa.

The Mogalakwena case study will show that despite the codification of recentralisation in the South African Constitution, regional and political party elites misuse legislated levers of recentralisation to interfere politically in municipalities. The effects of this are that service delivery slows down and local government is subjected to national and provincial political battles that become localised. On the other hand, the Lephalale case study will show how the layers of decentralisation between apartheid and democracy have led to this municipality being dependent on private and parastatal mining companies for the provision of and access to public goods and services. This transgresses the responsibilities of local government and requires a strategic recentralisation by national government to create a more sustainable approach towards the provision and access of public goods and services in the area.

Compromise and continuity

The long history of decentralisation in South Africa can be characterised as one of compromise between centralising and decentralising tendencies during the apartheid years, and then the continuation of many of these processes and structures following the onset of democracy. Prior to democracy in South Africa, moments and forms of decentralisation can be delineated into three main eras: the Union of South Africa phase (1910– 1948); the period of grand apartheid (1948–1982); and the reform and eventual collapse of apartheid (1982–1994) (Wittenberg 2006:330-5).

During the first period, there was a central national government that governed four provinces at the intermediate level. At the local level, there were  municipalities.  These  three  spheres  covered  the  governance  of whites only and were, chiefly, confined to the urban areas – the cities and towns. In rural areas, there was a system that comprised ‘privately owned farms – under white ownership – and a number of reserves where land was communally owned’ (Wittenberg 2006:330). Indeed, a parallel system of local government was conceived to deal with so-called ‘native administration’, with the Department of Native Affairs at the administrative head of a system of native commissioners, magistrates and tribal authorities (Wittenberg 2006:331). A graphic of this system, from Wittenberg’s work, follows.

Wittenberg (2006:331) describes this system as ‘dualistic’ because it was divided along racial lines: a form of weak white minority control of the administration, alongside central government control of the majority. Describing this, he points out that the ‘state was unitary rather than federal, but the existence of well-organised regional interests ensured that there was always some element of decentralisation’ (Wittenberg 2006:330).

During the second period, grand apartheid (1948-1982), forms of decentralisation intensified as administration and governance were more forcefully aligned to the ideology of racial separation at the core of apartheid (Wittenberg 2006:331). Processes of centralisation and decentralisation took place simultaneously. First, new forms of centralised control were introduced ‘particularly over the lives of black South Africans, through oppressive and repressive measures’, such as ‘control over movement, access to the labour market and access to accommodation’ in urban areas (Witttenberg 2006:332).

Yet the apartheid state also undertook a ‘remarkable, if warped, attempt at decentralisation’ where some administrative power (Wittenberg 2006:332), and ‘political decentralisation’ (Beinart and Dubow 1995:16), were ceded towards the bantustans. In addition, in 1961, Urban Bantu Councils were introduced to supposedly give black people in ‘white South Africa’ a measure of local government representation, followed in 1977 by Community Councils (Atkinson 1984:1). These structures were promoted by the apartheid government as the realisation of its slogan ‘that each race, within prescribed geographical spaces, was authorised to govern itself’ (Stanton 2009:58). But Stanton notes that, in reality, they amounted to ‘the apartheid government renouncing national government responsibility for providing infrastructure and service delivery to African people’ (2009:58). These forms of decentralisation are represented in another of Wittenberg’s images.

Finally, during the terminal phase of apartheid (1982-1994), the minority government attempted to ‘create bridging structures that cross-cut the existing administrative systems’ according to a principal of so-called ‘power sharing’ between different races. Nine development regions and multiracial Regional Services Councils were introduced by the Botha administration at the provincial and district levels respectively. These reforms sought to pacify resistance to the apartheid government and temper demands for its end (Stanton 2009:62, Wittenberg 2006:334). Through the Black Local Authorities Act, 102 of 1982, the Black Local Authorities (BLAs) were established in urban areas outside of the bantustans: these were intended to be ‘a channel for residents to voice their needs’ (Stanton 2009:62). But as Stanton notes, they represented a weak form of decentralisation because they were not funded by government but through local levies, and they were led by ‘local residents … [who were] … approved of by national government’ (Stanton 2009:62), and these structures were the explicit target of anti-apartheid resistance. Thus, during apartheid, decentralisation was used both purposefully and opportunistically as an instrument of separation and containment, rather than of nation-building. Therefore, Prud’homme (2003) summarises the pre-democratic state as comprising both ‘genuine decentralisation for the white minority but centralisation for the black majority’ (Prud’homme 2003:18).

After apartheid ended, the country entered a phase of democratic decentralisation during the transition to democracy. At this time, ‘political liberalisation and democratisation’ became linked to decentralisation (Olowu and Wunsch 2003:38). The prospects for decentralisation were heightened during the negotiated transition to democracy in the early 1990s. Both the Congress for a Democratic South Africa (CODESA) and the multi-party negotiation process (MPNP) took place against the backdrop of violence and tension. Where the structure of the state and decentralisation were concerned, the negotiations became heated because the National Party (NP) and the Inkatha Freedom Party (IFP) preferred federalism over decentralisation (Humphries and Rapoo 1994:4-5), as opposed to the present system, in which the national government has ceded some administrative, fiscal and political powers to provinces, yet still exercises concurrent powers with the provincial and local spheres. In this system, ultimate authority rests with the Constitution, and not with regional or provincial autonomy. The African National Congress (ANC)

preferred the unitary state option (Simeon and Murray 2009:538) and settled on a form of decentralisation for the local sphere in order to allow greater participation and to encourage more effective service delivery (Koelble and Siddle 2012:70, Steytler and Mettler 2001:94, Friedman and Humphries 1993:10). Bardhan and Mookherjee note that South Africa’s democratic decentralisation resembled a ‘transition from one form of decentralisation to another’. Previously there was simply a ‘different form of decentralisation [which] was used as an instrument of central control and racial division by the apartheid regime’ (2006:30).

Decentralisation was eventually chosen over federalism during the transition to democracy (Humphries and Rapoo 1994, Steytler and Mettler 2001, Simeon and Murray 2001). The main reasons for the choice of a democratic ‘decentralised unitary state’ (Humphries and Rapoo 1994:9) were social cohesion (Simeon and Murray 2001), allowing a ‘common citizenship’ (Friedman 1993:10); and most importantly, as a conduit for ‘access to the state and to economic resources’ (Friedman 1993:10). Under the mandate of the Municipal Demarcation Act, 1998, the Municipal Demarcation Board began drawing municipal boundaries in local government.

Early on, the ANC government also took political decentralisation seriously by encouraging participatory democracy in local governments in developmental and local policy making. In his seminal work on participatory democracy in Brazil, India and South Africa, Heller (2012) notes that the developmental policy process of integrated development plans (IDPs), which involved participatory democracy, began with enthusiasm which later waned. This was a consequence of the ANC’s ‘electoral hegemony’, which disincentivised the need for the party to maintain participation in local government (Heller 2012:651). As such, the original impetus to encourage participation in South Africa’s local government disappeared as government outsourced its work to consultants who ‘crowded out community structures’ (Heller 2012:655).

Why has so little changed?

The historical legacies of local government and its political dynamics have meant that the implementation of democratic decentralisation has been complicated. Theories about the realities of democratic decentralisation suggest that recentralisation is the imposition of national authority that has failed to empower and capacitate local government to

perform its functions. This is only partly true in South Africa’s local government. Recentralisation is used for political motives, but it can also serve as a restorative measure (Chilenga-Butao 2019) for municipalities that are unable to meet their mandates and perform their functions.

Recentralisation occurs when decentralisation reforms are not planned or executed carefully, wholly or continuously. Olowu and Wunsch describe this gap as that between the ‘idealised’ and ‘actual’ results of ‘decentralisation reforms’ (2003:6). The idealised process promotes a linear process of decentralisation which involves devolution of powers, authority and resources; a greater sense of accountability in local governments; and more participation and better service provision. But the actual process of decentralisation is often complicated by various factors, such as:

incomplete statutory reform … resources are retained or recaptured by central actors … resources of localities are consumed paying for salaries of officials … local councils are ineffective because of low levels of education, poor organisation, infrequent meetings, internal division, and executive dominance … local institutions … designed to maintain central control; and/or … local elites dominate local governance from behind the scenes. (Olowu and Wunsch 2003:6)

These patterns can also be summarised as ‘elite capture’, ‘capacity constraints’ and ‘financial constraints’ (Siddle and Koelble 2012:50-52). Faced with this, local governments require more oversight, and are more dependent on national government.

Recentralisation occurs, secondly, when local governments without capacity ‘become dependent on patronage networks to manage day-to-day affairs, and subsequently witness the erosion of their ability to balance off the centre in intergovernmental relations’ (Erk 2014:540). This usually occurs when the number of local government units multiply, causing dependence on patronage networks for service delivery, due to the lack of skills and experience required to manage complex tasks (Erk 2014:540). The Constitution makes provision for recentralisation, described as ‘interventions’ in S100 (Constitution of the Republic of South Africa 1996 S100 (1):51) and S139 (Constitution of the Republic of South Africa 1996 S139 (1):67). Intervention legislation allows national government to take over the administration of a province or its departments, and for a provincial government to take over the administration of a municipality. However, there are what I have called elsewhere ‘legislative gaps’ in the regulations for interventions, between their articulation in the Constitution, and their implementation in subnational governments (Chilenga-Butao 2019:127, also Rampedi and Ledger 2019).

Furthermore, the initiation and implementation of interventions are sometimes politically motivated. A restorative measure that is supposed to assist provincial and local governments in carrying out their tasks, chief among these being service delivery, can thus become a punitive manoeuvre used to discipline subnational government officials.

Manor offers poignant insight into why South Africa’s local government may not fulfil its intentions ‘despite genuine promise’ (Manor 2001:1). The source of the gap between the intentions and realities of decentralisation in local government is the White Paper on Local Government 1998, which fails to provide important implementation details for successful democratic decentralisation. These prerequisites are significant ‘financial resources’; ‘substantial powers’ for local government; and, accountability between bureaucrats and elected representatives, as well as between elected representatives and the electorate (Manor 2001:4). Local government resources are usually controlled by the national and provincial spheres (Manor 2001:4); local government is expected to execute ‘highly complex tasks’ (Manor 2001:7) with few resources; and, elected officials are unlikely to be accountable because they are beholden to bureaucrats and they cannot always make decisions in the interests of large electorates (Manor 2001:14-15).

Moreover, Siddle and Koelble suggest that the local government executive structure, one of the key dimensions of decentralisation, is the target of contest and interference by political elites. Politics directly interferes with municipal work through ‘suspensions and side-lining of officials who may not be in favour with the party or faction in power at the time … undue pressure being placed on officials, creating tensions between the dictates of good governance and party-political demands … appointments may become politically motivated’ (Siddle and Koelble 2012:102). The connection between interventions and the local government executive structure to which Siddle and Koelble refer is that provincial and political party elites have used interventions to interfere with executive positions in local municipalities. Therefore, decentralisation struggles to meet its intentions in local government and is prone to recentralisation due to a complex system overladen with overwhelming political dynamics.

The position of mining towns in Limpopo

Both Lephalale and Mogalakwena are municipalities covering mining areas located in Limpopo. They were both subject to private economic interests through mining activities, and were also subject to apartheid laws that ‘imposed’ the apartheid-ideological Lebowa bantustan on sections of their localities, while carrying out forced removals of black African residents in the area (Phadi and Pearson 2017:1).

Lephalale includes the town formerly known as Ellisras, which had grown to become a mining centre in the late 1950s after the parastatal, the Iron and Steel Corporation (ISCOR), embarked on coal mining (Phadi and Pearson 2017:2). Amidst the mining activities that took place, a peculiar type of parastatal-led decentralisation took shape in Lephalale which saw apartheid’s racially-segregated decentralisation converge with the powers of local governance effectively held by ISCOR. Mine officials effectively ran the town for decades, and resisted attempts to municipalise the locality under both apartheid and democracy. In 1989, ISCOR was privatised, and in 2006 its Kumba branch became Kumba Iron Ore and Exxaro Limited (Phadi and Pearson 2017:7). At the same time, Eskom unveiled its plan to build the Medupi Power station in Lephalale, to avert power shortages in South Africa (Phadi and Pearson 2017:7). As such, many of the local residents of Lephalale have come to rely on the mines and their operations, and other related businesses in the municipality, for their livelihoods. As will be argued, Lephalale Municipality itself remains reliant on Exxaro and Eskom for the delivery of key services, which significantly hampers its effectiveness.

Like Lephalale, Mogalakwena is a municipality that depends on mining as its main source of economic activity. It includes 189 villages that were part of the Lebowa bantustan, and the town now known as Mokopane (formerly Potgietersrus) (Phadi and Pearson 2017:7). Mogalakwena’s large-scale mining operations have a much more recent history than Lephalale: Johannesburg Consolidated Investments began mining the platinum deposits in the early 1990s, which was subsequently bought by Anglo-American; other mining companies have also sought to establish mines in the area since, including Ivanhoe Platinum (Phadi and Pearson 2017:8). The municipality has endured highly fractious political contests in recent years which has drawn in competing political elites and repeatedly disrupted service delivery. Below is a summary of key statistics relevant for Lephalale and Mogalakwena, that demonstrate the need for effective service delivery in these mining towns.

Mogalakwena – elite political interference

An examination of the first case study, Mogalakwena, surfaces the argument that decentralisation has not met its intended outcomes because political elites interfere in the administrative decentralisation of the municipality. One episode of elite political interference occurred through a S139 intervention which sought effectively to by-pass the administrative leadership of the municipality for factional reasons. This is a unique and underexplored issue in the field of local government today.

At the time that this research was conducted, Mogalakwena’s municipal manager, Willy Kekana, was being challenged by the provincial department of Cooperative Governance, Human Settlements and Traditional Affairs (CoGHSTA) through a S139 intervention (Phadi and Pearson 2015:19). Kekana had been appointed some years earlier on the instruction of the provincial leadership of the ANC (Phadi and Pearson 2015:8), which oversees the deployment of senior management positions in local government (Phadi and Pearson 2015:8). Yet Kekana was opposed by a faction of the Waterberg Regional Executive of the ANC, and from 2011, conflict ensued between Kekana and proxies of the region in the municipal council, including the mayor, and senior managers of the administration. Regional politicians claimed that Kekana’s contract was illegal – technically true, as he had been granted a five year contract in violation of the Municipal Systems Act (see Pearson in this issue). Much of the contestation that ensued thus emerged over what Siddle and Koelble describe as politically motivated appointments (2012:102).

There were several failed attempts by a faction aligned to the mayor in council to remove Kekana through court processes. But when the ANC Provincial Executive was removed in 2013 and a new provincial government constituted, the Limpopo Department of CoGHSTA stepped in and attempted to take the municipality under administration, citing Kekana’s invalid contract and ‘institutional instability’ (Phadi and Pearson 2015:43) as the reasons for a S139 intervention. The intervention was interdicted by a ruling of the North Gauteng High Court in a case launched by a local social movement and Afriforum against the Department. When the then MEC of CoGHSTA tried to push through a plan outlining how the S139 intervention would proceed (Phadi and Pearson 2015:43), Kekana and his supporters in Council acquired another urgent interdict. In his judgment, judge Neil Tuchten found that:

The seriousness of this step cannot be overstated. With a stroke of his pen, the MEC attempted, in favour of a functionary of the MEC’s own choosing, to circumvent the carefully constructed network of constitutional and other statutory powers which led to the vesting in the municipal manager by the democratically elected representatives of the community served by the municipality, of the municipal manager’s powers to administer the funds of the municipality. The functionary selected by the province, declared the MEC, would not be accountable to the council of the municipality and ultimately the voters within the municipality but effectively to the MEC. (Mogalakwena Local Municipality v. PEC Limpopo, MEC CoGHSTA, Minister CoGTA, NCOP, DH Makobe, 35248/2014)

The judgment found that CoGHSTA was manipulating intervention processes to remove a municipal manager on political rather than administrative grounds. Indeed, the ‘institutional instability’ cited by the MEC as grounds for the intervention had in fact arisen as a result of other failed political attempts to remove Kekana, which had split the municipal Council and brought basic administration to a halt (see Phadi and Pearson, 2015:18-19).

The municipal manager is the accounting officer and final approver of decisions in a municipality, and thus the destabilisation of their position – and other positions that become factionalised during political battles – leads to instability over service delivery decisions (Phadi and Pearson 2015:74). As one Mogalakwena official remarked, ‘To recover from that situation you need six years’ (Phadi and Pearson 2015:74). Service delivery in local government is debilitated by political battles that take place within municipalities between provincial and regional political party elites. The misuse of S139 interventions as a political tool to deal with opponents ends up corrupting the very intentions of decentralised local government – participation, transparency and accountability.  These political dynamics are absent in many mainstream debates on decentralisation, despite lying at the root of some of the worst failings in service delivery in South African municipalities today.

Lephalale: mining administration to municipalisation

During the Union of South Africa period Ellisras was an agricultural centre as well as a ‘railway and road rest-stop’ (Phadi and Pearson 2017:1). As with much of South Africa during the apartheid years, the black communities in the area were ‘forcibly removed’ to villages 40 to 70 kms away from the whites-only town (Phadi and Pearson 2017:1). Industrial decentralisation also shaped the economic and political patterns of the town and villages, as labour was drawn from the Lebowa bantustan to work in the mines, echoing the observations of Legassick and Wolpe (1976).

The onset of mining was announced when the parastatal ISCOR ‘acquired the property rights to six farms in the area in 1957’ – and ‘vast tracts of land’ it purchased from local farmers (Phadi and Pearson 2017:1). ISCOR came to develop a housing suburb of Onverwacht and developed a quasi- municipal structure to provide services to this community and the neighbouring Ellisras (Phadi and Pearson 2017:1-2). The mine thus undertook many of the functions of local government, providing public services such as housing, water provision, and other basic infrastructure. With significant economic autonomy ISCOR created its own governance structure, as noted by the former Lephalale town planning manager, where ‘“the mine manager is [the equivalent of] the municipal manager, the resident engineer is [the equivalent of] the head of the infrastructure department … so they actually did what the municipality did … If I was a mine manager, because it’s a private town, it belongs to me – I can make my own rules”’ (Phadi and Pearson 2017:2, quoting Dries de Ridder, interviewed September 17, 2015.). Processes of decentralisation in this and other industrial towns (see Sparks 2012) were thus shaped by the inextricable link between local governance and the economic interests in the town.

Following the establishment of a mining administration, the apartheid government sought to bring the Ellisras-Onverwacht area under government control in 1986 through the creation of a Town Council, a move ‘met with resistance from ISCOR executives’ (Phadi and Pearson 2017:2). This followed the growth of the town following the establishment of the Matimba power station by Eskom, another parastatal that came to hold significant infrastructural power in the area. When the municipality was eventually formed, a number of ISCOR officials ‘became councillors in the municipality’ (Phadi and Pearson 2017:2), through elections held in 1986. This inclusion of company executives in the new municipal ‘council’ maintained the authority of private interests over public ones, maintaining ISCOR’s position as an autonomous municipal actor. ISCOR continued ‘administering basic aspects of service delivery to the town’ (Phadi and Pearson 2017:3). One long-standing municipal official recounted that ISCOR had ‘“already installed everything and it was already done here. So, when the municipality came into play, you can’t now create a new system and do your own water treatment works … [this] is in fact not supposed to be, but it’s how it has worked over the years”’ (Phadi and Pearson 2017:3).

Ellisras-Onverwacht was effectively governed as a company town in which mining officials wielded powers of local governance. During the transition to democracy and then even after the eventual amalgamation of the area under the Lephalale Local Municipality, which saw Ellisras, Onverwacht, the township of Marapong and the more distant villages combined under a single municipality, these historical legacies of company governance presented profound challenges to the incoming ANC local government as basic services fell firmly in the hands of private mining companies (Phadi and Pearson 2017:5). The amalgamation coincided with the unbundling and privatisation of ISCOR, and in the process Exarro came to effectively own both the land and the municipal infrastructure that the parastatal had established.

The post-apartheid state’s aim to produce an inclusive, deracialised, decentralised, democratic government was challenged by the forms of autonomy established by the mining administration in places like Lephalale. And, as was often the case, the former whites-only municipal council ‘was extremely resistant to change and often hostile to the incoming ANC leadership’ (Phadi and Pearson 2017:5). Multiple layers of private sector autonomy, apartheid decentralisation and the amalgamation process introduced through democratic decentralisation are the major factors shaping the nature and effectiveness of decentralisation in certain municipalities in South Africa’s local government today. These local historical realities are key to understanding why decentralisation in South Africa has been difficult to implement. Democratic decentralisation was undertaken in the midst of very specific, long-standing governance structures that underwent only slight adjustment through apartheid, but to a large extent continued into the democratic transition.

Negotiating for public services

The legacy of private sector interests in Lephalale has led to a situation where services that should be accessible by residents are managed by private sector interests. Municipalities thus have to negotiate these services with private sector companies. And these negotiations take place despite municipal plans to gain more control and access to these services (see Monama in this issue on the dynamics around the Lephalale Development Forum). Decentralisation goals such as efficiency, equity, development and poverty reduction are difficult to realise in a context of private control over basic goods. And thus, a targeted S139 intervention to realise these goals is possible in this scenario. As outlined earlier, the Municipal Structures Act outlines the responsibilities of municipalities. Two of these responsibilities are the provision of water and sanitation, which fall under the bulk services outlined in the Act. In addition to this, S139 stipulates that one of the conditions for an intervention in municipalities is to ‘maintain essential national standards or meet established minimum standards for the rendering of a service’ (Constitution of South Africa 1996, S139 (1) (b) (i)). Given the concentration of some services in the hands of a private mining company and a parastatal, there is more that can be done in Lephalale to ensure that public services are available to everyone in the municipality.

Lephalale municipality is located in a ‘water scarce region’, and, as Table 1 above shows, only 31.4% of households in Lephalale have piped water inside their dwellings, which means that the majority of residents in Lephalale need more adequate access to public services that could vastly improve their lives. The water that is available is provided by Exxaro and Eskom (Phadi and Pearson 2017:25). This arrangement is part of the historical legacy of water ownership, given that the water supply is sourced from the Mokolo dam built by ISCOR in the 1970s – as was the infrastructure for water purification and supply. Today, water from Mokolo is still purified in one of Exxaro’s plants and ‘resold to Eskom and the municipality’. During the apartheid municipalisation process in 1986, this arrangement was formalised, renewed in 1990, and remains in force today (Phadi and Pearson 2017:25). Water supply has increased in cost due to ‘infrastructural upgrades’ (Phadi and Pearson 2017:27), which were necessary as the size of the municipality has grown. But the purified water supply does not extend to the surrounding villages, which rely on a separate source of ‘chlorinated’ water that is ‘pumped from wells and boreholes to storage reservoirs owned and operated by the municipality’ (Phadi and Pearson 2017:29). Thus, municipalities such as Lephalale struggle to provide basic services for local residents because the means and access to those basic services do not belong to the municipality. The latter is a direct result of the origins of decentralisation in mining towns such as Lephalale.

The ideal scenario for the Lephalale municipality is to take ‘ownership’ (Phadi and Pearson 2017:29) of water service provision, supply water to Exxaro and the new Medupi power station, and thereby generate revenue for the municipality (Phadi and Pearson 2017:27). This would supply them with a more ‘sustainable’ water supply instead of relying on a private entity,3 and would align with the municipality’s service delivery mandate as a Water Service Authority4 (WSA) (Phadi and Pearson 2017:27). On the other hand, municipal officials in Lephalale must account for whether they have the capacity to take ownership of water service provision. Poor organisation and capacity constraints have been identified as hindrances to local government’s control over local resources (Olowu and Wunsch 2003, Koelble and Siddle 2012) – and indeed are cited by the general manager of the Exxaro mine in Lephalale as reasons for the company to hold onto the infrastructure (Phadi and Pearson 2017:31). In the meantime, ‘the municipality must operate in crisis mode while it waits for long-term sustainable solutions from the provincial and national governments’ (Phadi and Pearson 2017:29).

This dilemma demonstrates that, contrary to the trickle-down promises of neoliberal economics, ‘public’ goods in Lephalale exclude many local residents in the municipality. Ironically, a municipal structure established after the democratic transition must still negotiate access and provision of water from private owners in the area who effectively inherited state assets during ISCOR’s privatisation. The arrangement in place was not challenged, and extraordinary plans are now put in place which are dependent on provincial and national action. Furthermore, the municipality finds itself in crisis mode, negotiating with Exxaro about transferring the water treatment plant to the municipality (Phadi and Pearson 2017:27); negotiating instalment payments for the plant because the national and provincial treasuries and the Department of Water and Sanitation are financing only part of the loan for the plant; and still awaiting the completion of the second phase of a two-phase water infrastructure upgrade by national government (Phadi and Pearson 2017:28).

The case for a targeted and appropriate S139 intervention could be made for Lephalale where the province, and possibly national government, can use recentralisation in an attempt to create a more sustainable solution for the provision of water in the area. This will certainly be the case where recentralisation can prove itself to be helpful for the needs of the Lephalale municipality, which is different to the situation that is presented in the Mogalakwena case. However, the intervention must be limited in two ways. Firstly, the process of recentralisation through a S139 intervention needs to be clarified and clearly stipulated by the national government, the national Department of Cooperative Government and Traditional Affairs (COGTA), and the provincial Department of CoGHSTA. Many of the restorative effects of a measure such as S139 are weakened or lost because those implementing this measure do not have a consistent way of doing so (Chilenga-Butao 2019, Ledger and Rampedi 2019).

Secondly, the S139 recentralisation must be treated as a temporary measure – which it is according to the Constitution – that will not interfere in the political-administrative interface of the municipality. The use of administrative positions in municipalities to achieve larger political aims actually has the opposite effect of destabilising the municipality and slowing down service delivery. This does not mean that the municipality was functioning perfectly prior to such elite political interference. But it does mean that service provision in local government, which is often slow, detailed and difficult, is further constrained by elite political interference.


The complexities of governance and politics in municipalities in particular, and local government in general, are important because they force us to question the assumptions about the advantages of decentralisation as outlined earlier in this article, compared to what constitutes appropriate action in local government when these advantages are not borne out by the reality within municipalities. Such thinking requires a closer reading of the facts on the ground and a delicate approach to theories and applications of decentralisation and recentralisation. This article has demonstrated that the gap between the intended outcomes of decentralisation and their realities need to be taken more seriously in order to improve the lives of people in these localities.

Through the Mogalakwena case study, we discover that recentralisation which should be a restorative measure for administrative dysfunction is used as an instrument for political interference, which effectively limits the ability of municipalities to deliver services. Of course, the recentralisation procedure and language is described as being in the best interests of the subnational governments, but the actions and effects that follow have political ramifications and divert attention away from the core function of service delivery in municipalities. The Mogalakwena case study also shows that political battles that seem isolated to the national sphere are actually ‘decentralised’ to subnational governments, and they affect officials in these subnational governments who end up working in highly politicised, polarised and debilitating work conditions. This inevitably affects the ability of local governments to provide public goods and services. Moreover, the effects of the political dynamics present in municipalities such as Mogalakwena explain how political interference by provincial and regional elites creates the disparity between the intentions and outcomes of decentralised local government, particularly the intentions of participation, accountability and transparency.

In the case of Lephalale, historical dynamics of decentralisation render current service delivery ineffective, inefficient and costly by decentralisation’s own standards. The democratic decentralisation that was intended to redress apartheid legacies and chart a participatory developmental path in municipalities such as this one, is confronted by historically entrenched patterns of unequal distribution of the means of production and infrastructural power, as well as limited means of livelihood for local residents of mining towns and poor service delivery. These historical legacies are inextricably linked to the performance of decentralisation policies in South African local government today.

Interestingly, and contrary to the case in Mogalakwena, the problems in Lephalale are not insurmountable if there is a proper application of a S139 intervention. And such a measure of recentralisation would be used to address a historical legacy that continues to prevent access to and provision of water as a basic service in that municipality. The challenge for national and provincial governments is to implement this recentralisation measure in a temporary, strategic and purposeful manner. If implemented properly, this could change services and perspectives on recentralisation.

  1. This section of the Constitution names the following responsibilities of local government, including local amenities: municipal roads; refuse removal, refuse dumps and solid waste disposal; street trading; street lighting; traffic and parking (Constitution of South Africa 1996:138-9).
  2. The Municipal Systems Act distinguishes between district and local The general scope of the district municipality is integrated development planning, providing bulk infrastructural development and services, capacity building within the district municipality and ensuring an equitable distribution of resources in the locality (Municipal Structures Act 1998 S83:57-58).
  3. To this end, the Lephalale municipality has explored the options for developing this plan, the details of which are contained in Phadi and Pearson’s report (2017:27).
  4. The Department of Water and Sanitation describes a Water Service Authority as ‘any municipality responsible for ensuring access to water service in the Act, [which] may perform the functions of a Water Service Provider, and may also form a joint venture with another water service institution to provide water In providing water services, a water services authority must prepare a water service development plan (WSDP) to ensure effective, efficient, affordable and sustainable access to water services’ (Department of Water and Sanitation nd).



Atkinson, D (1984) ‘The search for power and legitimacy in black urban areas: the role of the Urban Councils Association of South Africa’. Institute of Social and Economic Research. Development Studies working papers. Grahamstown: Rhodes University.

Beinart, W and S Dubow (1995) ‘Introduction: the historiography of segregation and apartheid’, in W Beinart and S Dubow (eds) Segregation and Apartheid in Twentieth-Century South Africa. London: Routledge.

Cheema, GS and DA Rondinelli (2007) ‘From government decentralisation to decentralised governance’, in GS Cheema and DA Rondinelli (eds) Decentralising Governance: emerging concepts and practices. Cambridge: Ash Institute for Democratic Governance and Innovation, John F. Kennedy School of Government.

Chilenga-Butao, TJ (2019). ‘Constitutional crisis: case studies in the decentralised public administration of education in South Africa (2011 – 2015)’. Unpublished Doctoral Thesis. Johannesburg: University of the Witwatersrand.

Crook, R and J Manor (2000) ‘Democratic decentralisation’, The World Bank Operations Evaluation Department. Washington DC: World Bank. Available at: page.pdf (accessed February 22, 2020).

Erk, J (2014) ‘Federalism and decentralisation in sub-Saharan Africa: five patterns of evolution’, Regional and Federal Studies 24(5).

Friedman, S (1993) ‘Introduction: the nature of the divide’, in S Friedman and R Humphries (eds) Federalism and its Foes. Johannesburg: Centre for Policy Studies.

Government of South Africa. (1995) ‘White Paper on the Transformation of the Public Service’. Available at: gcis_document/201409/168380.pdf (accessed February 24, 2020).

_______ (2019) ‘Local government’. government-system/local-government (accessed September 19, 2019).

Heller, P (2012) ‘Democracy, participatory politics and development: some comparative lessons from Brazil, India and South Africa’, Polity 44(4).

Humphries, R and T Rapoo (1994) ‘Some remarks on governing the regions’, in R Humphries and T Rapoo (eds) Governing the Provinces: proceedings of two conferences convened by the Centre for Policy Studies on forming new governments and administrations in South Africa’s regions. Johannesburg: Centre for Policy Studies.

Koelble, TA and A Siddle (2014) ‘Institutional complexity and unanticipated consequences: the failure of decentralisation in South Africa’, Democratisation, 21(6).

Ledger, T and M Rampedi (2019) ‘Mind the gap: section 139 interventions in theory and practice’. PARI Report. Johannesburg: Public Affairs Research Institute (PARI).

Legassick, M and H Wolpe (1976) ‘The bantustans and capital accumulation in South Africa’, Review of African Political Economy 7.

Mamdani, M (1996) Citizen and Subject: contemporary Africa and the legacy of late colonialism. Princeton NJ: Princeton University Press.

Manor, J (1999) The Political Economy of Democratic Decentralisation.Washington DC: The International Bank for Reconstruction and Development/The World Bank.

_______ (2001) ‘Local government in South Africa: potential disaster despite genuine promise’, SLSA Paper 8. Brighton: Institute for Development Studies, University of Sussex.

Municipalities of South Africa (2020) Limpopo Municipalities. Available at: https:// (accessed June 23, 2020).

Municipal Structures Act No. 117 of 1998. Available at: default/files/gcis_document/201409/a117-980.pdf (accessed January 30, 2020).

Municipal Structures Act No. 32 of 2000. Available at: local-government-municipal-systems-act (accessed January 30, 2020).

Olowu, D and JS Wunsch (2003) Local Governance in Africa: the challenges of democratic decentralisation. London: Lynne Reiner Publishers.

Phadi, M and J Pearson (2015) ‘Northern frontiers: first draft report on Mogalakwena Local Municipality’. PARI report. Johannesburg: Public Affairs Research Institute (PARI).

_______ (2017) ‘We are building a city. The struggle for self-sufficiency in Lephalale Local Municipality’. PARI Report. Johannesburg: Public Affairs Research Institute (PARI).

Prud’homme, R (2003) ‘Fiscal decentralisation in Africa: a framework for considering reform’, Public Administration and Development 23.

Ribot, JC (2002) ‘African decentralization: local actors, powers and accountability’. Democracy, Governance and Human Rights Paper No.8. United Nations Research Institute for Social Development.

Rustomjee, Z (1991) ‘Capital flight under apartheid’, Transformation 15.

Siddle, A and TA Koelble (2012) The Failure of Decentralisation in South Africa’s Local Government: complexity and unanticipated consequences. Claremont: UCT Press.

Simeon, R and C Murray (2001) ‘Multi-level government in South Africa’, Canadian Journal of African Studies 43(3).

Smoke, P (2003) ‘Decentralisation in Africa: goals, dimensions, myths and challenges’, Public Administration and Development 23.

Sparks, S (2012) ‘Apartheid modern: South Africa‘s oil from coal project and the history of a South African company town’. PhD Thesis. University of Michigan.

Stanton, A (2009) ‘Decentralisation and municipalities in South Africa: an analysis of the mandate to deliver basic services’. PhD Thesis. University of KwaZulu- Natal, Pietermaritzburg.

Statistics South Africa (2011) Lephalale Municipality. Available at: http:// (accessed June 23, 2020).

_______ (2011) Mogalakwena Municipality. Available at: http://www.statssa. (accessed June 23, 2020).

Steytler, N and J Mettler (2001) ‘Federal arrangements as a peacemaking device during South Africa’s transition to democracy, Publius: the journal of federalism 31(4).

Wittenberg, M (2006) ‘Decentralisation in South Africa’, in P Bardhan and D Mookherjee (eds). Decentralisation and Local Government in Developing Countries. Cambridge: The MIT Press.