Introduction

South Africa is a country with substantial diversity across “racial”,Footnote 1 ethnic, economic, social, cultural, political, linguistic, geographic, and administrative lines, where small spatial concentrations of wealth coexist with vast expanses marred by extreme poverty, inequality, and unemployment. After the country’s transition to democracy from apartheid in 1994, the 1996 South African Constitution enshrined justiciable socio-economic rights, created relatively decentralized multi-level government structures comprising national, provincial, and local government spheres within a unitary state, and crafted a system of intergovernmental relations.

Provincial governments were endowed by the 1996 Constitution with an extremely restricted list of exclusive functions and limited own revenue raising powers (South Africa, 1996), largely reliant on intergovernmental grants from national government. While policies in respect to concurrent functions such as basic education and health are largely set at national government level, provinces play mainly an implementation function, subject to norms and standards set centrally.

This chapter aims to assess the extent to which political, economic, linguistic, and geographic diversity has influenced the design and ongoing implementation of the system of intergovernmental financial relations (IGFR) in relation to the nine South African provincial governments. By reviewing the post-apartheid evolution of the institutional trajectory of provincial governments in South Africa, the chapter also analyzes the extent to which “fiscal federalism”—or IGFR as it is more commonly referred to in South Africa–has accommodated diversity in practice.

Section 2 of this chapter analyzes the genesis and evolution of the nine South African provincial governments. Section 3 examines the role of provincial governments in the IGFR system. Section 4 considers the extent to which provincial diversity influences intergovernmental revenue sharing. Section 5 discusses re-centralization pressures since the adoption of the 1996 Constitution. The devolution of functions from the national to the local government sphere (e.g., housing and public transport to large metropolitan municipalities) also influences provincial-local intergovernmental relations. Section 6 compares the provincial experience of re-centralization to the experience of devolution in local government. Section 7 concludes by reflecting on the gradual hollowing out of provincial government in South Africa. Provincial powers which were already limited in the Constitution have been further attenuated over the last two decades by the re-centralization of functions (e.g., social security, training and vocational colleges to national government and the proposed National Health Insurance system which will centralize current provincial health functions). This is exacerbated by centralized wage bargaining, hence provincial governments have little control over their largest cost driver, introducing budgetary rigidities. Given that provincial governments have few of their own revenue sources and are largely reliant on intergovernmental grants, the escalation in the provincial wage bill has tended to outstrip increases in intergovernmental grants and crowded out other provincial non-wage expenditures crucial for service delivery. This has been exacerbated by national pressures for fiscal consolidation in the post-pandemic recovery period (Sachs et al., 2022).

The highly centralized application of intergovernmental relations has limited scope for provincial experimentation and innovation. Pervasive state capture and corruption over the last decade have resulted in greater and more complex regulatory environments for provincial governments which further limit innovation and create stark trade-offs between probity and service delivery.

The Emergence of Provincial Governments and the Evolution of the Intergovernmental Fiscal System in South Africa

During the apartheid regime, “White” South Africa was characterized by spatial segregation along the lines of the four designated “racial” groups—Whites, Africans, “Coloreds”, and Indians—within its four provinces (the Cape Province, the Transvaal, the Orange Free State, and Natal). There were ten black homelands called Bantustans established within the borders of South Africa for specific “tribes”, e.g., Ciskei and Transkei for the Xhosa, Qwa Qwa for the Basotho, and KwaZulu for the Zulu. There were also four so-called “independent” states (i.e., Transkei, Ciskei, Venda, and Bophuthatswana). These served as reservoirs of cheap labor, requiring black people to obtain “passes” to enter “White” South Africa and seek work in farms and mines under the migrant labor system, as part of the apartheid policy of “influx control” (Wilson & Ramphele, 1989).

A cornerstone of the apartheid regime’s ideology was “separate development”, which rhetorically embodied the ideal of separate but equal development for all “races”, but in practice resulted in a racial hierarchy of economic opportunity, social mobility, and access to government services along racial lines (Posel, 2001).

The South African Constitution of 1996 absorbed all these homelands and the so-called “independent” states (see Fig. 1) into the Republic of South Africa and established nine new provinces: Western Cape, Northern Cape, Eastern Cape, Free State, KwaZulu-Natal, Mpumalanga, Limpopo, North West, and Gauteng (see Fig. 2). Provincial government boundaries were “not designed to constitute linguistically or culturally homogeneous entities” since this would have been perceived by the African National Congress party (ANC) as entrenching “one of the most pernicious evils of apartheid, a state designed along ethnic lines” (Murray & Simeon, 2010, p. 238).

Fig. 1
A map of South Africa's provinces and homelands before 1996. Cape takes most of the area. It is located towards the southern section. The central section is for Orange Free State with Transvaal situated in the northern part. The homelands Venda, Gazankulu, and Lebowa are located in the north.

(Note From “The nine provinces of South Africa”, by M.C. Alexander [2018], [https://southafrica-info.com/land/nine-provinces-south-africa/])

South Africa's provinces and “homelands” before 1996

Fig. 2
A map of South Africa with its nine provinces. The northern section has Limpopo, Gauteng, and Mpumalanga. The central section has North West, Free State, and Kwazulu-Natal. The southern and south western sections have Northern Cape, Western Cape, and Eastern Cape.

(Note From “The nine provinces of South Africa”, by M.C. Alexander [2018], [https://southafrica-info.com/land/nine-provinces-south-africa/])

The nine provinces of South Africa after 1996

Gauteng, the economic heartland of South Africa, produced about 35% of the country’s GDP in 2016, housed about 25% of the country’s population in 2017, has a population density of 785 people per square kilometer, but comprises only 1.7% of the country’s land area. By contrast, the Northern Cape is the largest province, comprising about 31% of the country’s land area, but accommodated only 2.1% of South Africa’s population in 2017, with a low population density of only three persons per square kilometer (Alexander, 2018).

In February 1990, the ban on the ANC and other political parties was lifted by the National Party (NP) apartheid government, and Mr. Nelson Mandela, the leader of the ANC, was released after 27 years of imprisonment. This fundamental shift was followed by the Groote Schuur Minute, sealing the NP and the ANC’s commitment to peace and a negotiated settlement in May 1990. The ANC renounced the armed struggle in August 1990, and the National Peace Accord was signed in September 1991. These events collectively signaled the beginning of a multi-party negotiation process, which included the Inkatha Freedom Party (IFP) based in KwaZulu-Natal, and the homeland administrations. The subsequent Convention for a Democratic South Africa (CODESA I) talks started in December 1991. These talks culminated in the NP’s capitulation to ANC demands for an interim government, general constitutional principles for a non-racial, non-sexist, democratic South Africa in February 1992, and a Whites only referendum held by the NP, in which the reform received overwhelming support from the white electorate. In an atmosphere of extreme tension, fear, infighting, and mistrust, the CODESA II talks forged agreements at the Multi-party Negotiating Forum on a constitutional assembly, an interim government, the release of political prisoners, hostels which housed migrant laborers,Footnote 2 dangerous weapons, and mass action. Finally, on November 18, 1993, the interim Constitution of South Africa was ratified paving the way for democratic elections, with the proviso that the two newly elected Houses of Parliament would sit as a Constituent Assembly to draft a new final Constitution (Humphries & Shubane, 1993; Maasdorp, 1993).

At that stage, the decentralization debate centered around regionalism, which sparked intense political controversy on an array of disparate models and generated a vibrant academic literature. The NP’s brand of highly decentralized federalism with entrenched regional autonomy (including regional taxation powers) was conceived as an instrument to curtail the future ANC majority government’s powers. Given the racial and spatial divides of the apartheid past, the ANC, wary of attempts to dilute its ability to govern the country from a strong center, to curtail redistribution and the potential for political territorial mobilization on the basis of race or ethnicity, punted a unitary state with strong local government, but weak regional elements. The IFP, based in Natal with mainly Zulu support, endorsed a confederationalist approach which would have amounted to virtual secession (Humphries & Shubane, 1993; Maasdorp, 1993; Schlemmer, 1994; Wittenberg & McIntosh, 1993).

The ANC’s policy positions on regionalism, which would later transmute into the provinces, were outlined in two policy documents: Ten Proposed Regions for a United South Africa (African National Congress, 1992), which proposed a regional demarcation similar to the developmental regions used by the Development Bank of South Africa, and the ANC Regional Policy (African National Congress, 1993). The latter policy document advocated that the specifics of regional boundaries should be left to the Constituent Assembly, but it did delineate proposals on the powers and functions of regions, which later influenced Schedules 4 and 5 of the 1996 Constitution. It also outlined technical arguments against the devolution of taxes, such as personal and company income taxes and value-added tax to regional governments, and advocated instead a form of sharing of nationally collected tax revenue by a proposed permanent Advisory Commission on Fiscal Decentralization, which would later become the Financial and Fiscal Commission created by the 1993 interim Constitution.

Given apartheid history of racial discrimination and exclusion in public services, including in education and health systems, racially segmented labor and financial markets and spatial segregation of residential suburbs, liberation movements—such as the dominant ANC—favored a single, centralized unitary, non-racial state with all citizens entitled to the same level of public services and the same economic opportunities. Simeon and Murray contend that while the 1996 Constitution gives full expression to cultural differences and diversity in the private sphere (e.g., religious and linguistic differences in the education system and eleven official languages), “such differences are to be recognized and institutionalized as little as possible in the political sphere” (2009, p. 546). Aware of how the apartheid government had exploited ethnic and tribal rivalries as part of a divide and rule strategy, ANC leaders were concerned that a political system which legitimated and entrenched such differences might intensify these conflicts, undermining their vision of a united, non-racist, non-tribal South Africa (Simeon & Murray, 2009).

The creation of provincial governments in South Africa was a concession by the ANC to the NP apartheid regime during the negotiated transition to democracy. The ANC favored a strong national government and strong local government, while the NP saw provincial governments as a vehicle for protection of White minority rights and privileges. The creation of the provinces therefore played an important peace building function, because they permitted the NP to achieve a majority in the Western Cape and the Inkatha Freedom Party (IFP) to achieve a majority in KwaZulu-Natal. The National Party also had an interest in protecting the linguistic rights of White Afrikaners; the Western Cape has the largest proportion of Afrikaans speakers, while KwaZulu-Natal, the stronghold of the IFP, hosts the large proportion of isiZulu speakers. After the first democratic elections in 1994, the NP gained control of the Western Cape and the IFP of KwaZulu-Natal (De Villiers, 2007; Lodge, 2005). By 2023, eight out of nine provinces were controlled by the African National Congress, and only the Western Cape was controlled by an opposition party, the Democratic Alliance (DA).

South African fiscal conditions at the dawn of the new democratic dispensation in 1994 presented a rather bleak picture. Aggregate fiscal discipline was weak. Increasing national government expenditure in the dying days of the apartheid era, coupled with weak revenue collection, created burgeoning debt, raising concerns about fiscal sustainability, and the looming specter of a debt trap (Abedian et al., 1995; Ajam & Aron, 2007; Folscher & Cole, 2006). In keeping with the apartheid philosophy, a highly centralized budget process allocated resources along racial lines, with substantially less per capita being spent on health, basic education, and other services for black South Africans in comparison to their white counterparts.

The Role of Provincial Governments in the South African Intergovernmental Fiscal System

Besides confronting the immediate macroeconomic and equity challenges of the democratic transition in 1994, the incoming ANC government was also charged with the long-term project of implementing the fiscal and socio-economic rights provisions of the newly adopted 1996 Constitution. Unlike the constitutions of many other countries, the South African Constitution gives substantial direction on the type of fiscal institutions that should support decentralized, democratic accountability (e.g., public procurement, generally accepted accounting practice, and intergovernmental revenue sharing).

The adoption of the 1996 Constitution precipitated a complete restructuring of public policy and reconfiguration of the South African public sector into a unitary state with three distinct, but inter-related spheres of government: national government, nine provincial governments, and 257 municipalities in the local sphere. The emerging IGFR system was predicated on the parameters for expenditure assignment, revenue assignment, intergovernmental grants, and borrowing powers as set out in Chapter 13 of the Constitution.

Revenue-raising powers, in terms of the Constitution, remain highly centralized in the national government. The most significant and productive taxes, such as value added tax, personal income tax, and corporate income tax, are thus reserved for the national government, and are collected by a single entity, the South African Revenue Service (SARS). Provincial governments, by contrast, have few of their own revenue sources. Own revenue sources refer to revenue instrument for which provincial governments have substantial control (in the sense that they can set the revenue base and/or the rate in their own province). Section 228 of the Constitution assigns province revenue sources which are not very buoyant or high yielding, namely “taxes, levies, and duties other than income tax, value added tax, general sales tax, rates on property, and customs duties”. Provincial owned revenue sources include gambling taxes, hospital patient fees, and motor vehicle license fees. Income raised within provinces as “own revenue” (mainly from car licenses and hospital fees) amounts to less than 5% of the overall provincial budget (South Africa. National Treasury, 2023). The primary administrative rationale for this centralization is that tax collection is easier to administer at a national level due to economies of scale and the mobility of tax bases across provinces, and that the duplication associated with a more decentralized system is avoided.

Significant expenditure responsibilities have, however, been decentralized to the provincial governments, notably basic education, health, agriculture, and provincial roads. Consequently, there is a substantial vertical fiscal imbalance between the significant expenditure mandates of provincial governments and their restricted own fiscal resources. Provincial governments are accordingly highly dependent on intergovernmental fiscal transfers from the national government (Ajam, 2019).

The functional competences devolved to provincial governments may be concurrent (shared responsibility of national, provincial, and/or local governments) as outlined in Schedule 4 of the Constitution, or exclusive (sole responsibility and discretion of the provincial, municipal, or national government) as outlined in Schedule 5 of the Constitution. Under concurrent functions, such as primary and secondary education and health services, the national government sets policy, but implementation is largely the responsibility of the provinces. The separation of policy, national financing, and implementation (at provincial level) has sometimes resulted in unfunded or partially funded mandates, together with complex coordination problems. Unfunded or partially funded mandates occur when functions are devolved to provinces or municipalities without commensurate funding sources. All residual functions that are not specifically enumerated in Schedules 4 and 5 remain exclusively at the national level (for example, matters relating to foreign affairs, defense, trade policy). Local government’s concurrent and exclusive competences are detailed in part B of Schedules 4 and 5.

Provincial legislation with regard to the exclusive Schedule 5 functions takes precedence over national legislation, except when national legislation is necessary to establish national norms and standards, to maintain economic unity, to protect the common market in respect of the mobility of goods, services, capital, and labor, or to promote economic activities across provincial borders (section 146 of the Constitution). Thus, provinces do not only have a limited degree of fiscal and political autonomy, but this is further weighed against the broader national interest.

Given that provinces have significant expenditure responsibilities and comparatively small own revenues, section 214(1a) of the Constitution confers on provincial governments the right to an “equitable share” of nationally collected revenue. In addition to the provincial Equitable Share grant, which is unconditional, section 214(1c) also permits national government to extend to a provincial government, from its national share of revenue, conditional allocations which are earmarked for specific purposes.

The process by which tax revenue is collected nationally by the South African Revenue Service, pooled together with the proceeds of debt finance and subsequently divided among national and subnational governments, is referred to as revenue sharing. The process is formalized annually in an annual Division of Revenue Act (DoRA) tabled in Parliament. The Fiscal and Financial Commission (FFC), an independent body established in terms of section 198 of the interim Constitution and section 120 of the final 1996 Constitution, has the constitutional mandate to make recommendations to Parliament on equitable allocations to national, provincial, and local government from nationally collected revenues.

The Constitution lists a number of factors in sections 214(2) (a) to (j) to be taken into account in determining a province’s Equitable Share allocation. These factors include the need to ensure that they are able to deliver basic services and perform their mandates, economic disparities within and across provinces, differing fiscal capacities and efficiency, and the developmental needs of provinces. Section 220 of the Constitution formally establishes the FFC to give impartial advice to Parliament on the equitable sharing of revenue. The Constitution thus acknowledges that the revenue-sharing process has both administrative and political dimensions. The technical administrative process requires that the national government annually consults provincial governments, organized local government (the South African Local Government Association), and the FFC. The political process of revenue-sharing culminates annually in the DoRA passed by the National Council of Provinces and the National Assembly of Parliament, together with the national budget.

The unconditional nature of the Equitable Share intergovernmental grant means that the Equitable share is in effect a substitute for provincial governments’ own revenue and thereby strengthens the fiscal discretion and integrity of the provincial government as a sphere of government (rather than merely administrative extensions of the national government). The national government cannot instruct provincial governments directly on how to spend their Equitable Share intergovernmental grant, but provincial governments are obliged to adhere to any minimum norms and national standards of public service delivery set by the national government. Furthermore, the conditions of service of provincial civil servants (such as salaries and benefits) are negotiated centrally via collective bargaining in the national sphere. Given that provincial government services such as Health and Basic Education are very labor-intensive, and that personnel budgets constitute the biggest part of provincial government expenditure, in practice, provincial governments have limited fiscal authority.

Provincial governments also have limited borrowing powers, conferred by section 230 of the Constitution. Section 230 imposes the fiscal rule that provincial operating budgets be balanced, permitting debt financing for bridging purposes only within a particular financial year. This effectively rules out borrowing to fund current expenditure in operating budgets. However, provincial capital borrowing is permitted. Under the interim Constitution, the Borrowing Powers of Provincial Government Act, Act 48 of 1996 (South Africa, 1996) was promulgated, which established a Loan Coordinating Committee with identical membership to the Budget Council, to coordinate provincial borrowing. No regulations were ever issued in terms of this Act, and the Budget Council agreed in 1997 that there would be no borrowing until a framework compliant with the final Constitution was finalized (National Treasury, 2001).

The Budget Review in 2001 announced: “Provincial borrowing powers for capital projects will also be phased in over the Medium Term Expenditure Framework. The Budget Council is expected to approve a framework later this year” (National Treasury, 2001). It was expected that when the 1996 Constitution came into force, new legislation for provincial borrowing powers would be enacted. This had not happened by 2022 and provincial borrowing to date has been limited, which is probably prudent, given the negligible provincial own revenues and poor expenditure controls and financial management practice in many provinces. However, the Gauteng Provincial government did access loan finance of R1 billion (one billion rands) in 2011 to contribute to the construction of the Gautrain light rail infrastructure project (Gauteng Department of Finance, 2011).

Municipalities are endowed with more substantial fiscal capacity, being entitled by the Constitution to impose rates on property and surcharges on fees for services provided by or on behalf of the municipality (for example, for electricity or sewerage). Municipalities are also allowed to borrow, subject to similar restrictions as the provincial government.

In order to balance provincial and national priorities, provincial governments are given a degree of discretion in their delivery of concurrent functions within the parameters set by the national legislative frameworks, provided they conform with minimum national norms and standards of service delivery, and do not conflict with the national economic unity, national security, or minimum national standards, in terms of sections 42(2) and 100 of the Constitution. As observed earlier, the national government in general formulates policies which are then implemented by provinces and municipalities.

This concurrent, overlapping nature of the functional assignment of competences of provinces (and municipalities) necessitates effective intergovernmental relations. Accordingly, section 41 of Constitution places great emphasis on “cooperative government” as the fundamental norm to underpin intergovernmental relations: “[…] organs of state have the obligation to respect the constitutional status, institutions, powers and functions of government in the other spheres” and to “co-operate with one another in mutual trust and good faith” by:

  1. i.

    fostering friendly relations;

  2. ii.

    assisting and supporting one another;

  3. iii.

    informing one another of, and consulting one another on, matters of common interest;

  4. iv.

    coordinating their actions and legislation with one another;

  5. v.

    adhering to agreed procedures; and

  6. vi.

    avoiding legal proceedings against one another.

The Intergovernmental Relations Framework Act of 2005 (South Africa, 1996) outlines procedures to deal with intergovernmental disputes, with the aim to resolve potential conflicts and pre-empt litigation between spheres of government.

In terms of the Constitution, the national government plays an important role in monitoring the compliance of provincial governments with national minimum norms and standards, supporting provincial governments in the discharge of their mandates and intervention (in terms of section 100 of the Constitution), in the event that a provincial government is unable or unwilling to fulfill its executive obligations. This should be read in with section 125(3) of the Constitution, which requires that the national government “by legislative and other measures must assist provinces to develop the administrative capacity required for the effective exercise of their powers and performance of their functions”.

For the national government to play its supervisory and support role effectively, it needs to ensure that clear, well-specified service norms, and standards are in place, that the relevant national sector departments have adequate monitoring and evaluation systems to assess compliance with national norms, as well as the capability to build capacity in subnational governments as required.

National norms and standards are typically enshrined in sector legislation for concurrent functions and regulations, issued in terms of the relevant pieces of legislation, for example, the South African Schools Act of 1996 (Act 84 of 1996), and the National Health Act of 2003 (Act 61 of 2003). These norms and standards are therefore not unilaterally introduced by the national government but are the product of consultation with provincial governments in intergovernmental forums, such as the MinMecs,Footnote 3 and in the National Council of Provinces legislative process. The absence or poor specification of national norms and standards undermines national supervision and support effectiveness, since effective monitoring and evaluation depend on those pre-defined norms and standards. Where norms and standards do exist, sometimes there is non-compliance due to provincial governments’ lack of funds or implementation capacity.

A hotly contested case in point is national school infrastructure norms, which became the subject of a court case in 2012 between an NGO, Equal Education, and the national Minister of Basic Education, Ms. Angie Motshekga, (Equal Education, 2012). On March 2, 2012, Equal Education filed an affidavit in the Eastern Cape’s Bhisho High Court against Motshekga, and twelve other respondents, including nine provincial Ministers of Education, the Minister of Finance, and the Eastern Cape Provincial Government. The NGO sought emergency relief for two schools, Mwezeni Senior Primary School and Mkanzini Junior Secondary School in the Eastern Cape, where classes had to be taught in mud classrooms or corrugated iron shacks. They also sought an order that would instruct Motshekga to prescribe minimum norms and standards for school infrastructure (Equal Education, 2012). This was granted by the Bhisho High Court on July 11, 2013. On November 29, 2013, the Minister of Basic Education approved infrastructure norms and standards for the construction of classrooms, access to electricity, water, sanitation, libraries, perimeter security, school safety, sport and facilities, electronic connectivity. The regulations stipulated three, seven, ten, and seventeen year targets, which the provincial Education departments must meet in order to eradicate school infrastructure backlogs by 2030. According to the Provincial Norms and Standards Reports for seven of the nine provinces available on the national Department of Basic Education website, in 2021, provinces such as Limpopo and Mpumalanga still fell short of the 2016 target for ensuring that schools have adequate access to water, electricity, and toilets. Other provinces like the Northern Cape, Western Cape, and Gauteng have achieved these standards but—due to severe budget constraints—still fall short of the minimum standards in respect of schools constructed from inappropriate materials such as asbestos, metal, or wood, overcrowding due to classroom and new school backlogs, poor perimeter security, and no internet access (South Africa. Department of Basic Education, 2022). Huge educational inequalities continue to exist across public sector schools within and among provinces, and between the public sector and the private education sector which, conversely, largely meets first world educational standards.

Since their inception, there have been substantial variations in the governance and implementation capacity across provincial governments and significant differences in service delivery outcomes. Provinces like the Eastern Cape and Limpopo have struggled with severe capacity constraints, whereas the urbanized, wealthier provinces like Western Cape and Gauteng have exhibited better delivery capacity (Murray & Simeon, 2010). Provincial discretion has been limited, and some provincial governments have been characterized by a pervasive lack of accountability and high levels of maladministration, fraud, and corruption (De Villiers, 2007; Hendriks, 2017; Lodge, 2005; Naidoo, 2009). Since the Zuma administration in 2009, all three spheres of government in South Africa have been subject to widespread corruption and state capture (Swilling et al., 2017). As a result, the regulatory environment—for example, in relation to supply chain management in terms of the Public Finance Management Act 1 of 1999 and its regulations—has become so complex that it has stifled innovation without doing much to deter financial misconduct (Ajam, 2016). Provincial governments are more accountable “upward” to national government than “downward” to their provincial electorates (Murray & Simeon, 2010).

Yet there has been some degree of innovation within the provincial government sphere. An example of provincial innovation includes the provision of antiretroviral drugs for people living with HIV/AIDs, despite the AIDS denialism at national government level during the tenure of former President Thabo Mbeki between 1999 and 2008. Doctor Fareed Abdullah headed the HIV/AIDS program in the Western Cape Department of Health which had pioneered the roll-out of AZT to pregnant women to prevent mother-to-child transmission of HIV, when the national Health Minister, Nkosazana Dlamini-Zuma, and President Mbeki refused to do so (Nattrass, 2008). He acknowledges the role of provincial autonomy—however limited it may be—in achieving that outcome:

And I was faced with a simple, but difficult dilemma—follow the party line or do the right thing as a doctor. It took me three seconds to decide what to do. I had the authority at that time—public servants had authority—the authority to do things, to order a new drug, to start a new program. So we continued with the treatment program and national couldn’t stop it because we really understood the powers that provinces have. Constitutionally, a province can set their own policy and we jealously guarded the right of the province. (Huisman, 2022)

Similarly, the roll-out of Nevirapine—an antiretroviral drug—to prevent mother-to-child transmission had been delayed by the national government, confined only to two pilot sites per province. In July 2002, the Constitutional Court ruled in Minister of Health v. Treatment Action Campaign and OthersFootnote 4 that Nevirapine be provided to all pregnant women who needed it. A key consideration in the Court’s decision was that the Western Cape, Gauteng, and KwaZulu-Natal provincial governments had already begun supplying Nevirapine. In April 2004, the national government changed its policy and extended access to anti-retroviral medications to all South Africans living with HIV/AIDS (McNeil, n.d.).

Provincial Diversity and the Equitable Share Grant

As noted earlier, provincial governments are largely reliant on the “Equitable Share” intergovernmental grant. From total tax revenue and borrowings, debt service costs and a contingency reserve are top sliced, as reflected in Table 1. The balance (R1 663.5 billion in 2022–2023) is then split among the three spheres of government in the “vertical division of revenue” which is a political decision taken by Cabinet.Footnote 5 In the 2022–2023 fiscal year, 49.7% of this amount went to national government, 41.2% (R682.5 billion) to the provincial governments collectively, and 9,1% to local government, which has significant own revenue sources such as property rates and tariffs for water and electricity sales.

Table 1 Vertical division of nationally raised revenue in South Africa in billions of rands, 2019/2020 to 2024/2025

Table 1 illustrates that intergovernmental allocations to provincial governments consist of the unconditional Provincial Equitable Share (PES) grant (amounting to R560.8 billion in 2022–2023), and conditional grants which are transferred by national government departments to provincial governments. These are earmarked for specific national government objectives delineated in a framework approved for each grant by the national Parliament as part of the annual Division of Revenue Act. Conditional grants include the education infrastructure grant, the national school nutrition grant, the district health program grant, the national tertiary services grant, and the provincial roads maintenance grant. There are also small indirect transfers to provincial governments which are in-kind grants, for instance where a national government department builds school infrastructure on behalf of a provincial government or expands optometry and audiology services to schools.

The Provincial Equitable Share (PES) grant is allocated to each of the nine provinces by means of a largely demographically driven formula, the structure and underlying variables of which are transparently published each year by the South African National Treasury as part of its annual Budget Review document. As outlined in Table 2, the PES formula has generated the largest allocation of 21.4% of the R560.8 billion total provincial revenue pool to Gauteng Province in 2022–2023 (R120 billion), and the smallest allocation of 2.7% of the provincial revenue pool (R14.9 billion) to the Northern Cape Province. The PES formula used in the 2019–2020 division of revenue consisted of six components, and some of the components are based on subcomponents.

Table 2 Total intergovernmental transfers to South African provincial governments, 2023/2024
  1. 1.

    An education component (48%), based on the size of the school age population (ages five to seventeen) and the number of learners (Grade R–12) actually enrolled in public ordinary schools.

  2. 2.

    A health component (27%) based on the health risk profile of each province and its health system case load. For example, women of child-bearing age and older persons consume more health services than the population average; therefore, a province with a greater share of women of child-bearing age and of older persons would have a greater health risk profile, which is taken into account in the formula.

  3. 3.

    A basic component (16%) derived from each province’s share of the national population.

  4. 4.

    An institutional component (5%) divided equally among the provinces, which recognizes that some costs of running a provincial government are not related to the size of a province’s population or factors included in other formula components.

  5. 5.

    A poverty component (3%) based on the province’s share of poor households (in other words, people falling in the lowest 40% of household incomes in Statistic South Africa’s 2010–2011 Income and Expenditure Survey), supporting the redistributive thrust of the formula.

  6. 6.

    An economic output component (1%) based on the Gross Domestic Product by Region (GDP-R) data published by Statistics South Africa which measures the gross domestic product produced in each province (South Africa. National Treasury, 2023).

The variables on which the PES formula is based reflect the marked differences in the spatial distribution of economic activity and poverty across the nine South African provinces, with population size broadly proxying for the demand for public services provided by provincial governments. The education component is a weighted average of Statistics South Africa’s 2021 mid-year population estimates of the school age population in each province (50% weight) and school enrolment data from the national Department of Basic Education’s Learner Unit Record Information and Tracking System (LURITS) (weighted 50%). KwaZulu-Natal Province, for example, has the highest school aged population in 2021 (3.033 million learners) but a school enrolment of only 2.891 million learners which suggests that some learners are not attending school or have migrated to other provinces. By contrast, Limpopo Province has a school in enrolment of 1.798 million learners which exceeds the Statistics South Africa estimate of 1.703 million children between the ages of 5 and 17 years (South Africa. National Treasury, 2023).

The variables in the health component of the PES formula attempt to capture the material differences in the health care needs and demands across provinces. As noted earlier, the health component of the PES comprises a health care system case load subcomponent (weighted 25%) and a risk adjusted subcomponent derived from an estimate of each province’s proportion of the population without medical insurance and hence reliant on public health services, based on the 2019 General Household Survey (weighted 75%). The health care system case load output subcomponent is based on two variables: each province’s share of the average number of visits to primary health care clinics in 2019–2020 and 2020–2021 drawn from the District Health Information System, and the average of each province’s share of total patient-day equivalents at public hospitals in 2018–2019 and 2019–2020. In the risk adjusted component, a risk adjustment index is applied to the percentage of each province’s population not covered by medical insurance. The risk adjusted index itself consists of a weighted average of five variables: age and sex, total fertility rate (to capture the additional costs of health services to pregnant women), premature mortality (to proxy for a higher burden of disease), sparsity (to capture the higher per capita costs of delivery in regions which are remote or with low population density), and a multiple deprivation index (based on variables such as highest education levels, the state of the living environment, and poverty based on the ownership of assets or household goods). These capture the vast differences in the social determinants of health across provinces (South Africa. National Treasury, 2023).

Recentralization in the Aftermath of the 1996 Constitution

The future of provincial governments has been shrouded in a substantial degree of prolonged policy uncertainty. In 2007, a policy process to review the provincial and local government system was initiated by the then Department of Provincial and Local Government (currently known as the Department of Local Government and Traditional Affairs), which was to culminate in a White Paper on Provinces and a revision of the Local Government White Paper (South Africa. Department of Provincial and Local Government, 2007). Fifteen years later, there has still been no resolution to this policy process, exacerbating the policy vacuum.

While the future of provincial governments as a distinct and autonomous sphere of government remains unresolved, several individual sector-oriented decisions—in relation to social development for instance—have been taken over time by the national government de facto attenuating the role of provincial governments. Presidential Proclamation R7 of 1996 had, in terms of the 1993 interim Constitution, assigned the administration of social assistance, including the payment of social grants, to provincial governments. On September 6, 2004, the Constitutional Court in Mashavha v. President of the Republic of South Africa and OthersFootnote 6 ruled that Proclamation R7 was invalid, and that social assistance was not a matter which could be regulated effectively by provincial legislation since minimum norms and standards across the country were required. This judgment paved the way for the promulgation of the South African Social Security Agency Act of 2004 (South Africa, 2004) which centralized the social security function and established the South African Social Security Agency to manage the payment of social grants such as the old age pension, child support grant, foster care grants, and disability grants.

In 2009, a decision was made that Further Education and Training (FET) Colleges (currently known as Training and Vocational Education Colleges), which had formerly been administered by the nine provincial education departments, be made a national competence. The FET Colleges Amendment Act of 2012 moved this function from the Members of the Executive Council for Education of the provinces (i.e., provincial Ministers) and the provincial Heads of Education Departments to the national Minister of Higher Education and Training and the Director-General of that Department (South Africa, 2012). After complex negotiations with the nine provincial education departments, the National Treasury, and the nine provincial treasuries, the funds allocated to FET colleges were transferred to the national Department of Higher Education and Training. The transfer of the function took effect on March 31, 2015.

In the 2015–2016 fiscal year, the port health function was also moved from the nine provincial health departments to the national Department of Health (South Africa. National Treasury, 2015). R380.4 million was also shifted from the PES grant to the national Department of Health. The Financial and Fiscal Commission was consulted both in respect of the FET colleges and port health function shifts and both the basic education and health sector MinMecs and the Budget Council approved. The Budget Council is an intergovernmental forum comprising the national Minister of Finance and the provincial counterparts, plus the nine Members of the Executive Council (MECs) for Finance.

An unsuccessful attempt to re-centralize relates to the provincial roads function. The Administrative Adjudication of Road Traffic Offices (AARTO) Amendment Act of 2019 aimed to create a single national traffic system of road management, shifting from the judicial enforcement of traffic laws through criminal law to a compulsory system of traffic law administrative enforcement through administrative tribunals, administrative fines, and a demerit points system. This would remove the enforcement of all road and traffic laws from provincial and local governments to the national government. In Organization Undoing Tax Abuse v Minister of Transport and Others,Footnote 7 the Gauteng High Court found on January 13, 2022 that AARTO and AARTO amendment Acts were unconstitutional since they “unlawfully intrude upon the exclusive executive and legislative competence of the local and provincial governments, respectively” [para 45, p. 20]. This demonstrates that the South African courts have, in some cases, tried to preserve the autonomy of provincial governments and municipalities, albeit limited, in exercising their powers and functions and prevent unwarranted intrusion from the national sphere. It remains to be seen whether the national government will appeal this finding.

The National Health Insurance (NHI) Bill of 2019 is still currently being deliberated in Parliament (South Africa, 2019). This aims to establish an NHI Fund which will pool funds from general taxes complemented by mandatory payroll and surcharge taxes into a single fund which is publicly financed and publicly managed. The objective of the NHI is to maximize income and risk cross-subsidization, increase population coverage, and provide health services free at the point of care. The NHI Fund would be the single purchaser of personal health services at the relevant level of care for the South African population through various contracting arrangements from accredited public and private facilities, which would provide the health services, the so-called “purchaser-provider split”. Implementing the NHI and the purchaser-provider split would require a substantial re-configuration of the public health system, inter alia, in relation to primary health care, district hospital services (level 1 services), regional hospital (secondary services), tertiary and specialized hospital services, and emergency medical services. This would have far-reaching consequences for provincial health departments.

For example, section 32(2) of the Bill makes provision for the national Minister of Health to amend the National Health Act of 2003 “for the purposes of centralizing the funding of health care services”. Section 32(a) permits the national Minister of Health to re-delegate to the provincial health departments the provider functions as “managing agents, for the purposes of the provision of health care services” after the purchaser-provider split, for which the NHI fund will contract with the provincial governments. The Minister of Health may also designate provincial tertiary and regional hospitals as “autonomous legal entities accountable to the national Minister through regulation”.

Section 32(3) requires that provincial health department functions be amended to comply with the provisions of the NHI Bill and its transitional arrangements outlined in section 57 of the Bill, “without derogating from the Constitution or any other law”. Section 49(2)(a)(i) of the NHI Bill envisages the “shifting of funds from the provincial equitable share and conditional grants into the (NHI) Fund”.

The constitutionality of the NHI Bill implications for provincial governments is likely to be contentious in the Western Cape Province, which is governed by an opposition party, the Democratic Alliance party. The principles of co-operative government and intergovernmental relations enshrined in section 41 of the South African Constitution enjoin the three spheres of government “not to assume any power or function except those conferred on them in terms of the Constitution”, and to “exercise their powers and perform their functions in a manner that does not encroach on the geographical, functional, or institutional integrity of government in another sphere”.

The Constitutional Court will have to judge whether the reduction in the powers of provincial governments inherent in the NHI Bill is warranted by its potential to promote universal health service coverage and the progressive realization of the constitutional right to health. What is clear is that the fiscal and policy discretion of provincial governments will be greatly reduced in the health domain, tilting the scales more and more toward imposing uniformity, rather than accommodating provincial diversity and innovation.

Devolution and Local Government in South Africa

Although the prime focus of this article is on the provincial governments, it is interesting to reflect on the experience of the devolution of functions to the local government sphere, in particular to large metropolitan city governments. Devolution of functions to local government, despite its strong constitutional framework and the ANC’s notional support for strong local government, has in many instances been bedeviled by prolonged delays, uncertainty, and a lack of urgency on the part of the national government.

The national government has dragged its feet in delegating functional areas which are crucial to urban governance of cities, such as housing and public transport. This legislation delegation process has been characterized as “slow and incomplete”, often unaccompanied by the devolution of adequate resources, and has been “blamed for diluting, even stunting, urban autonomy in South Africa” (South African Cities Network, 2022: 77).

One of the reasons for the reluctance to devolve function to the local sphere is that a large swathe of municipalities are dysfunctional. However, it could be argued that this calls for a more nuanced and differentiated approach to devolution rather than “the tendency to recentralize power as a cure for municipal malfunction” (South African Cities Network, 2022: 95).

Section 11(2) of the National Land Transport Act of 2009 (NLTA) empowers the national Minister of Transport to assign any of the transport functions set out in section 11(1)(a) of the NLTA to a municipality to achieve the objectives of the Constitution and the 2009 Act itself, subject to sections 99 and 156(4) of the Constitution and sections 9 and 10 of the Municipal Systems Act. A Cabinet member, in terms of section 99 of the Constitution, has the discretion to assign, by agreement, any executive power or function that is to be exercised or performed in terms of an Act of Parliament to a municipality. Section 156(4) of the Constitution obliges a Minister to assign to a municipality the administration of a Schedule 4A function that necessarily relates to local government if the identified municipality has the capacity to administer the function and the function would most effectively be administered locally. Schedule 4A of the Constitution lists functional areas of concurrent national and provincial legislative competence, including public transport. Section 11(4) of the NLTA also permits a municipality to request the Minister of Transport to assign to it a national function in the NLTA if that municipality has an acceptable integrated transport plan.

In March 2022, Cabinet approved the White Paper on National Rail Policy (WPNRP) which acknowledges that some progress has been made toward the devolution of urban rail (South Africa. Department of Transport, 2022). The cities of eThekwini and Cape Town, for instance, have already conducted feasibility studies on the devolution of passenger rail and have established transport authorities to assume the management authority for urban rail, while the Passenger Rail Agency of South Africa (PRASA), a national government business enterprise with ownership control under the national Department of Transport, delivers the actual services. The WPNRP notes that this arrangement appears to be “work in progress” and that “their separate bus and rail system maps do not show intermodal interchanges with each other” (South Africa. Department of Transport, 2022, p. 84). Acknowledging that cities like Cape Town, Ekurhuleni, eThekwini, Johannesburg, and Tshwane have “already demonstrated capacity to take on the general management of entities the size of the commuter railways in their jurisdictions”, the WPNRP envisages the development and approval of a Devolution Strategy for Commuter Rail to guide the assignment of the commuter rail function to the municipal sphere of government, aligned with the Integrated Urban Development Framework as part of the next National Land Transport Strategic Framework to take effect in 2023 (South Africa. Department of Transport, 2022, p. 84). A Public Transport Funding Model would also be developed to sustainably finance passenger and commuter rail.

In May 2022, the Western Cape Minister of Transport and Public Works, welcomed the WPNRP’s affirmation of the imperative for devolution of public transport functions, but—in view of the current “crisis state of the rail system”—expressed concern that the three year timeframe for the development of the Devolution Strategy for Commuter Rail would further delay the process, and called on the national Minister to expedite its finalization to allow devolution to proceed (Western Cape. Department of Transport & Public Works, 2022, np). On December 13, 2022, the Mayor of Cape Town, Geordin Hill-Lewis, released a media statement noting that PRASA now only transports 3% of the passengers it was able to transport a decade ago and asserted that “the national government has allowed a national asset to fall into ruin and has no plan to repair it […] there is no hope of the national government or its state entities turning around passenger rail and they need to start planning for the handover process without delay” (South Africa. City of Cape Town, 2022, np). He also announced that he had written to the national Minister of Transport, Mr. Fikile Mbalula, on May 23, 2022 to request that an urgent working group be convened to run in parallel with the feasibility study (South Africa. City of Cape Town, 2022). It was reported that by December 20, 2022, he had not received a response from the Minister, despite a follow-up letter on December 1, 2022 (Stent, 2022). #UniteBehind, a commuter activist group, has announced its intention to initiate litigation in the new year to compel the national Minister of Transport to hasten the local control of commuter rail to municipalities or provincial governments, if concrete progress was not made (Stent, 2022).

Similar to the Treatment Action Campaign case discussed earlier, there are instances where national government is also inhibiting innovation at the local government level. A case in point relates to the attempts of the City of Cape Town to reduce its reliance on the national power utility, ESKOM, which has imposed rolling blackouts (euphemistically dubbed “loadshedding”) since 2007, by purchasing renewable energy from an independent power producer (IPP). The City applied for a determination from the national Minister of Energy in terms of section 34 of the Electricity Regulation Act of 2006 to approve the purchase of solar and wind power from the IPP in 2015 but the Minister failed to respond for two years and then announced that applications were on hold for an indefinite period. In City of Cape Town v. National Energy Regulator of South Africa and Minister of Energy,Footnote 8 the City argued that it had a constitutional right to procure energy in the manner it deemed best, and requested the court issue an order that the Minister’s consent was not required for an IPP to establish a new power plant to supply electricity to the City. Alternatively, it requested the court to declare that section 34 of the Electricity Regulation Act was unconstitutional because it impermissibly encroaches on the constitutional powers of local government. The Gauteng high Court postponed the application indefinitely and referred the matter to both of the parties since it held that the issue was an intergovernmental dispute that resolution should have first been attempted within the mechanisms of the Intergovernmental Relations Framework Act of 2005 before any of the parties could approach a court of law. Since the court did not pronounce the constitutionality of section 34 of the Electricity Regulation Act, the constitutionality of this provision may in future still be challenged.

Hollowing Out of Provincial Governments: Death by a Thousand Cuts?

It is clear that political, economic, linguistic, and geographic diversity has influenced the design and ongoing implementation of the IGFR system in relation to the nine South African provincial governments. However, the analysis of the post-apartheid institutional trajectory of provincial governments in South Africa illustrated that the very centralist, top-down IGFR system has not accommodated provincial diversity and has stifled innovation by provincial governments.

The already limited provincial constitutional powers have been further attenuated over the last two decades by the relocation of functions (e.g., social security, training and vocational colleges to national government, housing and public transport to large metropolitan governments, the proposed National Health Insurance system which will centralize current provincial health functions). This is exacerbated by provincial governments’ negligible own revenue sources and therefore their over-reliance on intergovernmental transfers such as the PES, centralized wage bargaining (which means that provincial governments have little control over their largest cost driver, introducing budgetary rigidities, given that escalation in the provincial wage bill has tended to outstrip increases in intergovernmental grants and crowded out other non-wage expenditures). The highly centralized application of intergovernmental relations in a uniform “one size fits all” manner has limited the scope for provincial experimentation and innovation. Finally, pervasive state capture and corruption have resulted in greater and more complex regulatory environments which limit innovation and create stark trade-offs between probity and performance.

The gradual erosion of provincial governments’ powers appears to be more as a result of disparate sectoral decisions rather than a deliberate policy decision, by default rather than as a deliberate design in a form of “creeping normality”. The term “creeping normality” refers to a situation where a major institutional shift, which would be objected to if it were introduced as a single step or over a short period, can become more easily accepted over time if implemented slowly and gradually. Abolishing the provincial governments as a conscious policy decision would trigger constitutional safeguards (such as the need for a two thirds majority vote in the national Parliament) and garner the opposition of provincial elites who benefit from the status quo. Incrementally denuding the provincial governments of their functions accomplishes much the same thing in a much less political visible way and with consequently less organized opposition. Murray and Simeon noted in 2009 that the reluctance of the ANC to accept a provincial system “means that the government has no clear vision of the role that provinces should play, there is no strong political commitment of leaders to develop the provincial system, and there is no mass support for provincial governments” (2009, p. 547). More than fifteen years later, very little has changed.

Whereas the eight provinces under ANC control seem to have little interest in preserving their constitutional integrity and autonomy, it is interesting that civil society organizations such as the Organization Undoing Tax Abuse (OUTA) have indicated their willingness to take legal recourse to expedite the devolution of passenger rail functions to municipalities. The climate emergency might trigger similar civil society campaigns in future. It is also conceivable that civil society groups might have increased appetite for litigation either to halt the further centralization of provincial functions or to advocate for further devolution of concurrent functions from national to provincial governments.

There is a fairly large body of work analyzing reform options for the provincial governments (e.g., De Villiers, 2007; Greffrath, 2012; Moeti & Khalo, 2007; Simeon & Murray, 2009). Making major changes (such as abolition or reduction in the number of the provincial governments) however require constitutional amendment. Whether the ANC can muster a two thirds majority to pass such an amendment becomes increasingly less likely. In the 2021 local government elections, the ANC lost several large cities to opposition coalitions. In the forthcoming 2024 general elections, it is quite likely that the ANC may lose some of their current eight provinces to opposition parties or coalition governments. This may be reinforced by political dynamics related to electoral reform, political party funding, and the rise of coalition government, which may create demands for greater accommodation of provincial diversity and further decentralization of powers and functions to provincial governments, or at very least, arresting centripetal pressures.