School governance and funding policy in South Africa: Towards social justice and equity in education policy

Equity and redress, in and through education, are fundamental commitments of the new South African democratic government that ensued in 1994 after a brutal and protracted history of colonial and apartheid segregation and oppression denied the majority black population the fundamental right to equitable and quality education. A raft of ambitious and far-reaching policies were put in place to achieve these laudable goals. Yet more than 26 years after the ending of colonial and apartheid rule, the South African education system, and society in general, remain, far from equal – made apparent by the current COVID-19 pandemic. In this paper we take a critical (re)look at South African education governance and funding policies, considering why the South African Schools Acts (SASA) and the National Norms and Standards for School Funding (NNSSF), first promulgated in 1997 and 1998 and subsequently amended, have not delivered as expected on the promises of equity, redistribution and redress. The paper advances conceptual flaws, operational failures and implementation naivety as to why these promises have not been realised, advocating for an alternative social justice model for school governance and funding.


Introduction
We start by reconsidering the commitment of the post-apartheid education system to address equity and redress in and through the education system and its financing. Emerging from the long shadow of apartheid, the transformation of education governance and financing policy was key to achieving the creation of an education system which realised the full potential of all children, equipping them with the skills, knowledge and disposition to contribute to the social, political and economic development of South Africa. We critically reflect on this education governance policy trajectory since 1994, focusing on the key research question which drives this paper: How have education governance policies in post-apartheid South Africa conceptualised and institutionalised equity and redress in funding and governance and with what effect? In reviewing the education policy trajectory, we argue that the policies have failed to deliver on the commitments to equity and redress as promised since the ending of apartheid in 1994. Based on this review, we argue for an alternative social justice redistributive model of school governance and funding derived from what one of the writers originally called the Fifth Option in education governance (Sayed, 2003). We begin with a brief description of the methodology on which the analysis is based followed by a review of the literature and outline of the conceptual framework. We then critically examine the key policies showing how they fall short in terms of providing equity and redress in and through education. We conclude by mapping out key policy agendas integral to the development of a socially just education governance and financing agenda for schools in South Africa.

Methodology
This paper is based on the view that policy construction and development reflects a particular social, economic and political context in which differing social forces seek to make and remake the world (Rizvi & Lingard, 2010). Through a detailed engagement with the content, and structure of South African education governance policies, we analyse the social, economic, and political beliefs, values and practice that have informed their construction. We specifically pay attention to the two landmark governance policies in South Africa as described below, the SASA and the NNSSF first promulgated in 1997 and 1998 respectively and subsequently amended. The detailed critical analysis of these written policy texts is based on critical discourse analysis (CDA), examining how inequality is reproduced through policy within a given historic, social and political context. CDA aims to go beyond a merely descriptive process to examine the ideologies and power reflected in policy texts. In addition to the discursive analysis of the education governance policies, we also review relevant empirical data from national and provincial education management information systems (EMIS), national and provincial treasuries, educator salary data (Persal), and the SNAP Survey of Ordinary Schools.
Further, we draw upon reflections from our own separate professional engagement in the education governance policy formation processes to address the issues of equity and redress in and through policy. We draw upon the conceptual framework of Nancy Fraser (1997Fraser ( , 2009) outlined below, to understand how social justice, particularly in relation to redistribution and representation, was framed and achieved though post-apartheid South African governance and funding policies. Collectively, this data offer an opportunity to deconstruct the beliefs, assumptions, values and socio-political dynamics that have informed education governance policy development.

Literature Review and Conceptual Framework
Different approaches to policy and policy analysis exist (Rizvi, 2006;Sayed & Ahmed, 2015) each tied to different educational approaches. Following Rizvi (2006), we argue that policy is not just factual but normative and contested. Rather than viewing policy texts as neutral, they are inextricably shaped by and reflect contexts like the current neo-liberal discourses on education. Policy can be understood as a value-laden process which is discursive in nature (Rizvi & Lingard, 2010). We draw from these broader conceptualisations of policy in considering how approach shapes policy outcomes.
To advance the debate, we draw on the work of Fraser (1997Fraser ( , 2009, which for us refers to the idea of social justice as distributional justice, reallocating resources accessed solely by the privileged to the historically disadvantaged. Moreover, we understand the concept as suggesting that while policy as the authoritative allocation of values seeks to achieve the "sameness" of all, it also affirms and recognises the marginalised and subjugated identities of the oppressed. To this end, we analyse post-apartheid governance policies in relation to: • Redistribution, which concerns equity in the distribution of education resources and outcomes for different groups, particularly the marginalised and disadvantaged. • Recognition, which concerns respect for and affirmation of diversity in education processes and content. • Representation, which concerns the participation, particularly of the excluded, in education governance at all levels of the system in policy formulation and implementation. Key to our framework is engaging with the idea of democracy and citizen participation in education policy. In South Africa, this is understood in relation to the democratisation of access to education policy outlined in the National Education Policy Act of 1996 in the context of a new democratic state. However, a stipulation of this policy regards the nature of devolved governance and the dynamic that was established between the centre and the provinces, or decentralised states. i The provisions of the Constitution, designed to accommodate a semi-federalist, decentralised state, locates fiscal control centrally, within an authority responsible for disbursing funds to the devolved sites, but with a limited monitoring and oversight role (Republic of South Africa, 1996). In other words, the national state funds, but does not directly deliver or provide education and as such, basic education is regarded a concurrent power. In this model of democracy, the associative bond between the centre and the sites is created through a process of juridification (Sayed, 2003). This involves the central Ministry of Education promulgating norms that are allowed in legislation. Accordingly, legislation facilitates the monitoring and regulation of internal state actors. As such, juridification is characterised by complex lower dynamics between the centre and provinces fraught with conflict from the political arena to the judiciary. In the semi-federal devolutionary impulses, the state, as shown below, reflects the key defining moments of educational restructuring in the international context, namely choice and participation. ii

Contextualising school governance and funding policy change in post-apartheid education: Moving towards equity and redress
In this section we provide a brief historicised synopsis of the genesis of SASA and the associated NNSSF. The policies are to be read in conjunction because of how and how much different schools are resourced and financed. The governance and funding of South African schools after the end of apartheid can be traced as far back as the period immediately pre-dating its demise when the National Party administration created Model B and C (Grant Lewis & Motala, 2004;Sayed, 2003). These proposals, igniting a strong commitment to education devolution, became core to the education negotiation strategy of the apartheid government during the transition discussion in the Convention for a Democratic South Africa (CODESA). At the same time, Sayed (2003Sayed ( , 2016 points to the devolutionary impulses of the African National Congress (ANC) and allies born from it, the experiences of people's education and grassroots involvement in the anti-apartheid struggle. In a sense, the negotiated settlement between 1990-1994 resulted in, notwithstanding diverse ideological and political orientation, an agreed commitment to education decentralisation and devolution. The shared consensus for devolution was accompanied by an approach to transition in which the idea of stability, particularly stability in education, arose as an overarching concern. Intense debates took place about decentralisation and selfmanaging schools, with a view evolving that localised control and governance through decentralised school funding would lead to equity, efficiency and redress.
The devolutionary commitment to education governance and concomitant concern for education stability were arguably the foundational approach of the key governance and funding policies in post-apartheid South Africa, namely, the Schools Act enacted in and amended in 2003, and the NNSSF, enacted in 1998and amended in 2006 In order to further engage and enfranchise local communities, the SASA (Republic of South Africa, 1996) extended considerable autonomy, including conveying the status of juristic persons, to newly created school governing bodies (SGBs). These were established in every school and made the school principal, along with elected representatives, the most important stakeholders. SASA is responsible for including admission, language policy and, perhaps most vital, the power to appoint both educators and non-educator staff in those schools where fees are charged. Sayed (2009), Mestry (2012) and Van Dyk and White (2019) note that the SASA has created two types of SGBs. The first are Section 20 schools with limited power over funding, including how they spend the resources transferred to them as schools, and consequently limited in their management abilities and mandate. In contrast, Section 21 SGB schools applying for additional powers are those that have fairly full control of their funds, including the funds raised by school fees, and can appoint educators, procure and pay for services, and determine the school extracurricular activities. SASA, in both its original and amended form, has established a twotier SGB structure with those serving poorer schools having limited management control and autonomy.
Notwithstanding the differences between these two types of SGBs and schools, SASA's policy logic privileges parental control over schools as it suggests they are heavily invested in these arrangements. The devolution of education authority to school communities by the SASA is premised on the idea that those parents or guardians are primary beneficiaries and so in the case of fee charging, similar to paying a user charge, should have the power to manage schools as they see fit, subject to certain regulations.
Since the original SASA in 1997, the state has redefined governance policies. For instance, since 2002, SGBs have not been consulted about issues around staff placement. Such arrangements are now provided for by the 2002 Education Laws Amendment Act (ELAA) and fall under the jurisdiction of the provinces (The Presidency, Republic of South Africa, 2002). Further, since the 1996 SASA, the act has been amended, as not all schools charge school fees. Other amendments include changes to the ELAA, which extend the authority and responsibility of the principal.
The NNSSF, a conjoint policy with SASA, originally introduced in October 1998 (Department of Education [DoE], 1998) and subsequently amended in 2006, also required a change to SASA. It provides guidelines for how provinces should fund schools, directing them to spend at least 60% of the budget on poorer schools and providing for fee exemption eligibility in fee charging schools based on income. Key to the financing approach of this policy is a determination of the poverty ranking of each school based on income levels, dependency ratios and literacy rates in the surrounding community (the five-tier quintile system). The NNSSF originally allowed all schools to charge school fees. An amended NNSSF (DoE, 2006) was developed to address some of the effects of this policy and increase access to at least some schools which were classified fee-free, mainly in quintiles 1 to 2. iii This was amended in 2010 to include quintile 3 schools in the no-fee school category, encompassing approximately 60% of all schools in South Africa. These schools are allocated a larger disbursement by the Provincial Education Department to compensate for the funds lost as a result of the new no-fee policy. Schools that charge fees in quintiles 4 and 5 but enrol learners who are fee exempt are liable for a block grant-based allocation up to the maximum allocation of a learner's costs in a no-fee school (Department of Basic Education [DBE], Republic of South Africa, 2015). Although no-fee schools are identified at provincial level, guidance on determining this status is nationally uniform and based on three poverty indicators: income, unemployment rate and level of education of the community.
The amendments to both NNSSF and SASA signalled a significant policy shift. The government policy of schools charging fees was no longer viable as originally proposed in SASA 1996, and thus for many schools that original ideathat the "governing body of a public school must take all reasonable measures within its means to supplement the resources supplied by the state in order to improve the quality of education provided by the school to all learners at the school" (Republic of South Africa, 1996:24)no longer held true. The revised NNSSF also stipulated stricter rules around facilitating access to fee-charging schools for children without sufficient means to pay those fees. Exemptions were to be granted according to a codified set of criteria and procedures. iv The logic of fee charging, parental choice and devolved school governance represents a particular insertion of the private into the public (Sayed & Motala, 2012). The private is inserted into the public as central to how the state provides education and how the middle class secures the local school site as a form of semi-private schooling without having to leave the public sector. The education governance system in South Africa was reconfigured to ensure that the private is an integral component of the state provision of education to assure quality and equity, retaining the middle class as advocates of reform and accountability.

Findings and Discussion
Unravelling the Education Governance and Financing Policy Logic of Equity and Redress in SASA and NNSSF As sketched above, South Africa has adhered to a model of school governance and funding since the end of apartheid in 1994, which has sought to introduce a neo-liberal prescription of user charges in education on one hand (though amended in 2006 to render schools in quintiles 1 to 3 no-fee schools, as discussed later), and an ultra-hyper-individualist "parent as consumer of education" approach to school governance, with strong school choice and market orientation in allocating places tempered by the idea of soft zoning, on the other. We argue that this was more than a simplistic realisation of neoliberalism as some researchers and scholars have suggested in the past. It was premised on a most particular conception of equity and redress buttressed by the best principles of participatory democracy using Fraser (1997) and Fraser and Honneth's (2003) theory of social justice. Fraser's (1997) conception of social justice argues for equity along two axes: representation and redistribution. In governance and funding policies, representation was core to the logic of devolutionary impulses of the SASA. Holding SGBs responsible for school governance and for key education decisions enabled parents, the majority members of SGBs, to hold power of school decision making, albeit within some limits. Recognising the differential capacity of SGBs, SASA in its implementation of the devolved school governance models delineated two categories of schools: those deemed Section 21 and those deemed Section 20. Mestry (2018) notes that Section 21 schools, which in South Africa are primarily wealthy quintile 4 and 5 schools, have greater power to determine their own affairs, a power which includes exclusive management of their finances, whether coming from government or from their own school fees or donations (Mestry, 2018). In contrast, Section 20 schools may have democratic SGBs, but they are constrained by very limited powers. They are, in the main, dependent on education officials and school leadership for many of the core decisions surrounding education and schooling. v The principle of fee charging appears on the surface to be a classic World Bank user-charge model (Ahmed & Sayed, 2009). However, in South Africa, this somewhat neo-liberal governance prescription is turned on its head as the user charge is deployed as a mechanism to effect redress and equity in and through education in several ways. Firstly, the quintile ranking of all schools in South Africa, which accompanied the NNSSF in 1998, determined that those schools in wealthier quintiles (mainly 4 and 5) with higher fees would receive less financing from government for their operating expenses, including payment of telephones and other running costs. In doing this, the state was able to effectively argue a form of redress for equity purposes in that resources were being disbursed disproportionately in favour of poorer schools serving marginalised learners.
Secondly, redress and equity in the form of access to such wealthier schools were guaranteed to those who could not afford fees, by the fee exemption policy prescribed by the NNSSF in 1998 and further tightened in the 2006 amendments to the policy.
Thirdly, redress and equity were achieved by the principle of equalisation of expenditure. In other words, as observed by scholars such as Motala and Carel (2019) and Sayed and Motala (2012), the key funding characteristic of post-apartheid education expenditure was that it was norm-based primarily on the number of learners in a school, with built-in equity measures for distributing resources and capital expenditure. Further, the provincial "equitable share" formula ensured that the poverty of the provinces was an overall factor in tax revenue disbursement across provinces. In these ways, the key argument for the state was that SASA and the NNSSF had put into play a policy logic to achieve equity and redress while crucially allowing for school choice and control by those who could afford to pay fees. In Fraser's (2009) terms, the NNSSF laid the basis for the redistribution of resources in post-apartheid South Africa in how resources were distributed for those in wealthier schools and neighbourhoods and exemption for those unable to afford, and in the mechanisms of school choice: mobility for those who were able to do so. Equity, redress and justice are thus conceived as properties of individuals which need historical rectification so that all could be, sometime in the future, treated the same. However, justice as differential distribution stands in contrast to the notion of redress as uniformity and standardisation across the education system.
In essence, the governance of semi-private schooling is one in which middle class users of the service are able to top up on public funding beyond their tax contributions. Redress as a policy imperative thus found a natural policy alignment with the idea that the middle class could be charged fees for achieving the quality of schooling which they deemed necessary for a good educational system. vi In one sense then, the post-apartheid governance and funding formula makes possible the attainment of social justice as represented in the 27,000 democratic SGBs encouraging the participation of parents and local community members. It also facilitates redistribution through the commitment to equity and redress in the financing of schools, by using fee charging as a mechanism to direct resources where needed, and through equalisation of expenditure using nondiscrimination norm-based criteria, with funding following the school and not the learner.
Yet institutional class, race and gender bifurcation is coloured around the edges. Schools in neighbourhoods that witnessed a flight of the white and wealthy population into rich suburbs or gatedliving communities created a new generation of commuter schools in which the wealthy have exited from certain schools, essentially abandoning these to the marginalised and underclass. Additionally, in the South African context, particularly in the Western Cape, these have increasingly become the schools for migrants from other African countries.
Thus, while the system is bifurcated, there is a fluidity in school movement in South Africa, described by Chisholm (2004) as "racial hopping", in which wealthy and middle-class Black parents leave townships to send their children to previously White schools. Children in township schools have moved to formerly Indian and Coloured schools which are more affordable, and the children of domestic workers and other township children attend the suburban schools on fee exemption, the latter paying a considerable transport cost. It presents a complex picture of a mobile learner population, exercising limited choice in terms of affordability. Thus, wealth, mobility and hot knowledge allow parents, middle-class teachers and government officials to send their children to advantaged ex-Model C schools. This racial hopping reveals another paradox in the South African context: that is, racial desegregation has only occurred in wealthy and richer quintile schools and among the wealthy classes of the previously disadvantaged, as the data in Figure 1 reveal. The narrative and policy trajectory of school governance and funding in the South African context reveal an institutionally bifurcated education system, both as a consequence but also as a premise of policy in which the poor, mainly Black population, is schooled in under-resourced and dysfunctional schools, largely in former township schools (about 80%), while the wealthy have access to semi-private public schools (ex-Model C schools) that primarily serve the White population and the new Black elite. It would seem that, perhaps contrary to intent, legislation around education has facilitated access to very differently resourced schools for different socio-economic groups. The policy has effectively secured admittance to and control of what were White schools for the children of the new middle class, a situation referred to as Middle Class Fee Clustering (Ahmed & Sayed, 2009;Sayed, 2003). This has positive implications for school governance capacity in these schools.

Three Reasons Why Equity and Redress have not been Fully Realised
In this section, we turn our attention to the reasons why the policy commitment to equity and redress in South African school education has not been fully realised in the anticipated ways. There are three sets of arguments advanced to explain this policy gap, involving the disjuncture between intention and effect and outcome. They relate to the way the The second set of reasons has to do with how the South African polity replicates a large federal system with strong and unequal provinces. While they do not have tax appropriation and spend functions, provinces nonetheless have control over the operationalisation of schools.
The third set of reasons has to do with the failure to conceive of the implementation realities of a set of policies which are ambitious in scope, far reaching in scale and spread over a vastly unequal 27,000-strong school system with almost 12 million learners in a society in which the legacy of a colonial and apartheid past runs deep. The forward mapping of the policy makers rubs against the bottom-up institutional realities of schools located in marginalised, disadvantaged, spatially segmented and unequal neighbourhoods.

Conceptualisation of equity and redress: Equity, rectification and equalisation, but not redistribution
The first set of reasons has to do with the conceptual unpinning of the notion of equity and redress in educational governance and funding policy. Equity and redress, key framing assumptions of the policies, are conceptualised as equalisation of expenditure, which means spending less on operating expenses of some public schools that charge fees (approximately 40% of all schools after the introduction of no-fee schools in 2006). The notion of equity in policy is not redistributive in nature: it does not fundamentally redistribute human resources, physical infrastructure or teacher salaries across the system. It merely uses a weak notion of redistribution as rectification in which it seeks to right the racially skewed unequal education expenditure under apartheid (equalisation). In theory, cutting back on the marginal cost of some public schools suggests that such resources become available to poorer public schools. However, the level of these resources and how they are distributed require further investigation. Notwithstanding efforts to level the resource playing field, there are still considerable discrepancies in terms of how well qualified teachers are distributed, the capacity of SGBs to effectively oversee and administer schools and manage various aspects of non-personnel related funding. Ultimately, these differences are manifested institutionally, in outcomes for schools which closely correspond to levels of affluence within the community and for individual learners affected by household poverty and wealth inequality (Ahmed & Sayed, 2009;Motala, 2009).
More fundamentally, the argument that equalisation of expenditure and reducing operating cost resource allocation to wealthier schools provide equity is tempered if private contributions pending are factored into the equation, as the following tables shows. The data below reveal that the loss of income by wealthier, semi-private schools is compensated by private input from school fees, resulting in a lower learner-educator ratio (LER), a more expansive curriculum offering, better educational facilities, and better supportto name but a few advantages accruing to wealthy parents from the current systemwhich translate into better learning outcomes and school performance. Table 1 above shows that while state expenditure is fairly equitable across public schooling, and thereby the provision of racial equity as well, private contributions distort funding levels (Motala, 2009;Sayed & Motala, 2012). Particularly in wealthier provinces, funds provided by stakeholders, mainly parents, distinguish schools within the public sector. Thus, while equity in education rectifies some form of inequity at the margins, for obvious political and policy reasons it leaves unchanged the privilege that wealthier schools have accrued as a result of the colonial and apartheid legacy. This rectification approach to equity in education is arguably a hallmark of the post-apartheid transformation strategy which, as many commentators have argued, fails to provide for strong forms of social justice in which active forms of redistribution and justice are pursued.
It is evident that LERs are significantly more favourable in wealthier schools (Table 2). In looking at state-paid employees, the disparity in learner educator ratios is not particularly large and in most cases quite pro-poor. Once SGB-paid educators are taken into consideration, however, a pattern of much larger class sizes in poorer schools becomes apparent. Not only do SGBs in wealthier schools have the financial capacity to be flexible in arranging staffing, but they traditionally attract the best qualified candidates too. This results in funding imbalances given the higher salary levels of more highly qualified teachers and the fact that educator salaries comprise the single largest component of school budgets. This perpetuates patterns of inequality in the level of educator quality. Figure 2 reveals the disparity in terms of teacher qualifications between the top two quintiles and the rest. In each province, the staffing profile for schools in the most affluent neighbourhoods invariably records better qualified teachers than schools in poor areas. Gauteng has the least pro-rich gradient of teacher qualification, with quintiles 1 to 4, on average, reporting the same levels of educator qualification. The gap between average educator qualification in quintile 5 and non-fee charging schools is, quite interestingly, the lowest among the provinces. The highly qualified educators continue to be present in quintile 5 schools. The level of physical and mental health support staff for schools is also widely unequal by province (   Total  EC  12  34  0  4  10  60  FS  7  49  18  12  1  87  GP  0  512  0  0  0  512  KZN  16  86  11  26  52  191  LP  0  1  0  5  1  7  MP  5  5  7  4  15  36  NC  0  2  1  0  0  3  NW  1  13  0  7  31  52  WC  49  24  49  45  49  216  Total  90  726  86  103  159 1,164

The Gordian knot of federalism and devolution
While the national DBE (known as DoE prior to 2009) established broad-based norms and policies, the actual implementation in schools is a provincial responsibility. Moreover, aside from a few conditional grants that are ear-marked for specific funding priorities, tax resources are distributed as block grants to provinces without any form of ring fencing for education or other social sectors through the equitable share formula (ESF) calculated with the following weights: education (48%); health (27%); basic components (17%) (derived from each province's share of the national population); institutional (5%); and poverty (3%). Despite the allocations indicated in the ESF, each province is responsible for how much funding is provided to schools and this includes any subsidies provided to fee-charging schools or for independent schools. While there is arguably management merit in such an arrangement, the reality is that all provinces are not equal in the distribution of schooling resources, as the table below clarifies. Further, if equity and redress are core policy priorities, that presumes a need for a form of redistribution of education resources not only within provinces, but between provinces. What the data in the table below reveal is that while the ESF accommodates provincial variations and poverty, the funds raised by provincial taxes on a wealthy tax base far outstrip the redistributive mechanism of the poverty component of the ESF. Provinces such as the WC are able to spend more per learner, but their education budget comprises a much smaller share of their overall budget. As a result, poor province net incomes remain significantly lower than that of wealthier provinces, despite the ESF intention to redistribute taxes raised across provinces equitably (Motala & Carel, 2019). One consequence is that provincial allocations per learner vary widely, as illustrated in the Table 4 below. What is committed by the ESF and what is allocated at the provincial level differ drastically. There are wide discrepancies in per-learner allocations by province. For example, in the 2017/2018 academic year, Gauteng spent 10% more per learner than the EC. Moreover, as is clear from the ratio of fee-paying to non-fee-paying schools, there are considerable and historic inequalities at the provincial level. The data in Table 5 illustrate great variation in the socio-economic status of provinces. Limpopo and the EC have 98% and 96% non-fee charging schools, respectively; whereas the WC and Gauteng have 56% and 48% fee-charging schools, respectively. In terms of learner population, only 3% of learners in Limpopo are in quintile 5 schools, whereas 34.2% of learners in the WC attend quintile 5 schools.

Inadequacy of redress and equity mechanisms in governance and funding policy
Arguably, failure of the governance and funding policies to achieve equity and redress has to do with the mechanism of policy and its implementation. For example, one unsurprising scholarly consensus is that the quintile ranking on which much of the policy rests, based on proxy measures of community poverty in which a school is located, is a blunt and poorly targeted mechanism for achieving equity, requiring review (Ahmed & Sayed, 2009;Hatch, Buckner & Omoeva, 2017;Sayed & Soudien, 2005; Van Dyk & White, 2019). More fundamentally, even where there is a more refined measure of poverty and inequality, this policy only operates on the margin: its equity effects are limited to marginal operating costs (Sayed, 2016). More importantly, the implementing mechanisms of the policy, in particular the charging of fees, were devolved to SGBs and parents. As such, the decision to provide fee exemption and the communication of such exemption are left to selfinterested middle-class parents and schools. The reality, as Veriava (2005) points out, is that many SGBs will not provide such exemptions unless they are compelled to do so. For example, in the 2018 General Household Survey (GHS) well under 1% of learners report receiving a fee exemption (0.4% of learners). If we limit the sample to public, feecharging schools, 2.1% receive fee exemptions and 1.0% receive bursaries (Statistics South Africa, 2019). Similarly, Fiske and Ladd, in their 2004 review, noted that just 2.5% of families with children in primary school and 3.7% of families with children in high school received fee exemptions. These figures illustrate the fact that it is not in the interest of self-serving SGBs to grant fee exemption as the policy intends.
Devolving school governance in a spatially segmented, unequal society creates a context in which the experiences of the rich and poor across race and gender divides do not have common points of experiential intersection. Further, the policies assume benevolence on the part of middle-class actors and that activated behaviours in managing the policy at the school level would result in common good outcomes. Unsurprisingly, this has not been the case. The Cuba case, for example, demonstrates that achieving equity is not only structural but requires changing the dispositions and understanding of individuals who enter a society marked by the long shadow of privilege and oppression. Additionally, the model of governance assumes that there is indeed capacity equally distributed across 27,000 schools to manage the devolved power granted. This is not the case. The idea of representation as social justice rings hollow in SGBs with limited capacity and resources.
Notwithstanding the intention to achieve equity, in reality the no-fee school policy has not realised equity (Hatch et al., 2017;Nordstrum, 2012). At best, governance and funding policies, including no-fee school policies, could be described as pro-poor (Hatch et al., 2017) and while seeking to rectify the most egregious effects of apartheid, are not fundamentally able to redistribute for social justice purposes.

Conclusion and Discussion
We argue that achieving equity and redress in governance and funding policies is a wicked problem: more than 26 years after the end of apartheid, inequities between rich and poor and between different groups persist. The binding constraint, it is argued, has been the way in which the policies reviewed in this paper has conceptualised equity and redress and the mechanisms instituted to operationalise the achievement of such goals. It is this complex relationship between policy conceptualisation and policy implementation within the context of historical inequalities that has shaped the educational terrain.
From the review, it is evident that SGBs, as a key mechanism for equity in policy, continue to befuddle attempts to achieve equity in redistribution and representation, not only mirroring but exacerbating the societal divides of inequality and reflecting a bifurcated education system. Neither in terms of policy scope nor policy implementation have they unfolded as sites of representation and redress. Crucial to these inequities are, as argued, the differential capacities of SGBs to lead and manage school affairs, sharpened and rendered more visible during the COVID-19 period.
Further, we note that the way in which equity is addressed in its operationalisation rests on a weak quintile modelling system and a less than robust equitable share formula. We found that the policies do not tackle inequity in personnel expenditure. Equity is only applied to non-personnel expenditure, failing to tackle redress in human resources (teachers), which arguably explains much of the differentiation within the school system. As such, we found, similar to others, that the mechanisms of equity must be reconsidered.
Using Fraser's framework, we argue for a much stronger redistributive form of equity within the current fee charging and no-fee charging school governance system advocating for what is called a pro-redistributive Fifth Option, a redistributive scheme for public school governance and financing which promotes and protects the idea of a common good and ideal for education. In the model proposed below, we consider equity as active redistribution and we rethink some of the ways in which the operationalisation of SASA and the NNSSF can be altered to affect a more robust form of redress.
The core starting point of the model is to develop a clear and defensible notion of equity with social justice at its core. The notion of social justice in this paper is orientated to an understanding of representation and redistribution as integral to addressing inequality, which in the South African case would imply redistributing resources and privileges from the privileged to the less privileged. The suggestion offered in this paper begins from this premise.
Secondly, it is premised that adequate state funding is available to achieve schooling equity. The debates on funding adequacy have been well discussed over an extended period (Fiske & Ladd, 2004;Motala & Carel, 2019). Without it, financial shortfalls will undermine development in and through education. Adequate financing is needed to provide dysfunctional schools with the infrastructure and facilities for a sound learning environment, according to clear norms and standards. Such measures should address the historical infrastructural backlogs of colonisation and apartheid which to date have not been fully addressed in the South African context. In attempting to administer a differentiated and regulated return to school in the lockdown 3 phase in South Africa, the biggest stumbling block continues to be the poor infrastructure of disadvantaged schoolsinadequate toilets and running water, doors and windows that need repair, and an absence of sufficient hygiene resources such as soaps and toiletries.
Thirdly, it is argued that it may be politically necessary to allow governing bodies to charge additional user fees with all the qualifications that have already been stated in SASA. The system proposed would, for example, top-slice 15% of every additional South African rand raised by governing bodies that choose to administer user fees. This should be placed into a development fund by a local cluster of schools, as in, for example, Zones of Pedagogic Influence in Mozambique, with the express purpose of effecting redress. This is referred to as the "redress fund." The possibility exists for those governing bodies that charge fees to access these funds if they can justify that the funds will be used for educational upgrading. This proposal is a form of taxing historical privileges, which is also a strong form of Fraser's notion of redistribution. vii In addition, there needs to be a proactive redress and equity strategy involving measures such as sharing teachers across schools which, as noted above, are an important source of inequity in the current system. For instance, the Personnel Administrative Measures can be revised to allow for more than the current 5% redress posts, shifting much needed personnel resources to the poorest schools.
The main implication of this analysis is that the failure of the education governance policies are not simply, as often argued, a matter of implementation failure. We argue that they are a failure of the conceptualisation of policies. This suggest, as discussed in this paper, the need to fundamentally revise and rewrite the education governance policies (SASA and NNSSF specifically). In addition, policy makers at various levels as well as practitioners need to pursue active redistribution strategies as outlined in this paper. Such changes at the level of policy and practice must be accompanied by engagement with the beliefs, views and actions of individual SGBs in the current cooperative devolved governance model, both at provincial and institutional levels. These cooperative governance changes should be accompanied by a change in the way the current soft zoning policy is implemented and consideration of global models of equity, such as the Right to Education (RTE) Act in India, which reserves a quarter of all places for disadvantaged students.
While we argue that changes have the potential to elevate equity in education, caution must be exercised as education changes on their own cannot address the deeply embedded structural dimensions of inequality in South Africa. The long march to a more egalitarian society and inclusive economic growth in South Africa also requires addressing inequities more broadly, including the spatial and segmented nature of South African society. As the COVID-19 pandemic has so clearly illustrated, education inequities mirror inequality in all areas of South African life. This has been especially relevant in schools, were access to data, digital platforms and computers determined whether teaching and learning could continue and the social divides in our society became more obvious and relevant, both between the public and private schools and within the public schooling sector. The policy pitfalls and gaps highlighted in this paper are instructive for other countries with similar social and economic challenges as they seek to develop transformative education systems to promote inclusive economic growth. i.
The notions of "state at centre" and "state at sites" are used in the descriptive sense of specifying the relationships of government set up by the Constitution which has carved South Africa into nine provinces. A fuller exploration of the notion of the state and civil society is developed in an earlier paper by Carrim and Sayed (1993). ii.
A review of notions of choice and participation in a market reveal two dimensions of the debate. The first is the advocacy, elaboration and critiques by researchers of either market or democratic egalitarian approaches to educational choice and provision. Secondly, researchers working in the empirical tradition (qualitative and quantitative) are engaged in determining issues such as how choice occurs, whether more choice results from greater deregulation through markets and whether markets result in privileging of some and greater inequity. iii.
The 2006 review of the South African Schools Act of 1996 and the subsequent amendment to the NNSSF, still requires parents in quintiles 4 and 5 schools to contribute to schools to top up school funding. Partial or full exemption from fees can be granted, on application, based on a needs assessment. In contrast, students in quintiles 1 to 3 schools are exempt from paying any such fees. iv.
Notwithstanding the amended policy, the reality is that quintile 5 schools find ways of excluding poor learners such as insisting on parents' bank statements and salary slips, which the amendments make illegal, but which are used to exclude those who cannot pay the fees. v.
In general, such schools can only spend their money through provincial education departments (PEDs). All resources must be procured through the PED, which means that they pay much more for everything. For example, if the school needs a computer, a Section 21 school will purchase directly through a retail store where a Section 20 school must follow state procurement policies, which typically result in such goods being more costly. vi.
To date, it is unclear how this idea of redress as a savings by the state on schools which charge fees operates. Personal inquiries by the author to the state have not divulged how monies the state saves are used for redress purposes. vii.
This option draws from a paper first published by Sayed in 2003 in a Journal of Education Planning and Administration (JEPA) book. viii.
All data except the Persal dataset are available to the public at the links provided in the References section below. Permission to use the Persal payroll data in the study was received from the DBE for an earlier chapter that Shireen Motala and David Carel (2019) worked on and was extended for this article as well. ix.
Published under a Creative Commons Attribution Licence.