Education policy since 1994 has spread access more evenly than learning, and gains follow school capacity

Introduction

The South African Schools Act of 1996 told the state to fund public schools fairly across the system so that past inequalities in provision could be corrected (Republic of South Africa, 1996). Some sixteen years later the National Development Plan asked for something much plainer, which was that every child in Grade 3 can read and write (National Planning Commission, 2012). Those two statements more or less bracket this paper. In between there was real redistribution of money and rules, and on the best recent evidence there is also a learning gap that is still very wide (Böhmer and Wills, 2025).

The argument here is that basic education policy since 1994 has been better at spreading access and resources than at spreading learning, and that whatever benefit it has delivered has tended to follow school capacity, meaning the ability of a governing body, a principal or a district to read a policy and act on it, more than it has followed need (Christie, 2021; Aina and du Plessis, 2023; Tshabalala and Nthontho, 2024). I should say early that the evidence is mixed. One of the ten studies uses national datasets (Böhmer and Wills, 2025), but most are small qualitative studies of a handful of schools, so I treat their findings as pointers and not as proof. The paper looks first at what the policy architecture promised, then at how policy reaches schools, who gains from the funding rules, what the learning data show, and whose voice counts in all of this.

What the policy architecture promised

The design of the 1996 Act is easier to admire than to apply. Section 34 puts the duty to fund public schools on the state and ties it to redress. Section 36 then asks every governing body to make reasonable efforts, within its means, to add to what the state provides, and section 39 lets a majority of parents at a meeting vote for fees, with exemptions for parents who cannot pay (Republic of South Africa, 1996). Put those three provisions side by side and a tension appears. The state is meant to equalise, and at the same time schools with more means are invited to top up. Govender (2022) reads this as a product of the policymaking process, arguing that although consultation in the 1990s was wide, experts and powerful lobby groups carried a lot of weight in shaping what came out.

Govender’s method is worth a sentence in its own right. She calls it critical policy historiography, which really just means taking a long historical view of school-funding policy and asking whose interests a policy ended up serving. She critiques the 1996 Act as the foundational moment of funding policy in the democratic era, looks at the actors who shaped it, and then reviews the changes made since. Building on the work of Salim Vally, she argues for a policymaking framework that strengthens the influence of poor and marginalised communities on policy outcomes (Govender, 2022). That reading is persuasive for funding. What one case study of one policy area cannot tell us is whether the same elite effect shows up in curriculum or teacher policy, though some of the later sources hint that it might (Tshabalala and Nthontho, 2024).

The National Development Plan moves the language from rules to results. In its summary form it asks for at least two years of preschool for all children, for all children in Grade 3 to be able to read and write, for better quality in underperforming schools, and for an education accountability chain running from the state to the classroom (National Planning Commission, 2012). Those are sensible aims. A chain of accountability from the state downwards is also a rather top-down picture, and it sits a little awkwardly beside Govender’s (2022) case for giving poor communities more say. The two could be made to work together, but the Plan does not say how, and so the question of who actually holds whom to account stays open.

How policy reaches the school

Policy does not arrive at a school as a finished thing. It arrives as a circular, a workshop, a new form, and somebody has to make it work on a Monday morning. Tshabalala and Nthontho (2024) interviewed primary school principals about exactly this, using educational policy implementation theory, and found that principals are key actors in the process. They also report a run of difficulties: principals have little voice when policies are developed, resources are short, stakeholders want different things, support and training are thin, and some people resist change (Tshabalala and Nthontho, 2024). None of that sounds exotic. It does suggest that a good deal goes missing between the policy text and the principal’s office, and that the missing part is mostly support (Christie, 2021).

Curriculum policy gives a second window on the same problem. Teane (2023) describes the Curriculum and Assessment Policy Statement as the latest of a long run of policies introduced since apartheid ended, and notes that teachers have complained of lacking the skills and resources to put it properly in place. Her study followed teachers from 22 secondary schools who were trained by a university to perform the Life Sciences practical work that the policy prescribes, and she reports that the partnership gave them the skills and was followed by better learner performance (Teane, 2023). That is encouraging, but I would hold it loosely. It is one partnership in one subject, it rests on teachers’ reports, and it depended on a university stepping in, which is not something the department can count on for every subject in every district (Tshabalala and Nthontho, 2024).

Christie (2021) asks what counts as normal schooling in a system whose inequalities are already well known. Drawing on Foucault’s ideas of governmentality and dispositif, she argues that the pandemic exposed the structural inequalities in schooling and also the tangled governmental processes and logics that produce and normalise them. She then questions how far government capacity can actually meet its own stated aim of equal provisioning of schooling for all, and she uses the supply of water to schools as her illustration (Christie, 2021). I find the water case persuasive precisely because it is so dull. Nobody needs a theory of curriculum to see that a school without water is not being equally provisioned. If policy cannot reliably reach that level, the finer promises about reading and assessment seem to rest on shaky ground (Teane, 2023).

Who benefits from the funding rules

Start with schools at the top of the quintile scale. Aina and du Plessis (2023) studied financial decision-making in fee-paying public primary schools through a qualitative multiple case study, using interviews and document analysis, and found that governing body members at Quintile 5 schools were educated professionals who, in line with the Act, strongly shaped how the school’s money was managed. They set this against earlier literature suggesting that stakeholders at non-fee schools in township and rural areas have little influence over financial decisions, which sits badly with what the Act intends (Aina and du Plessis, 2023). That contrast is useful. The same statute gives both kinds of governing body the same formal powers, and the people who actually sit on them appear to be very different (Republic of South Africa, 1996).

The non-fee side of the picture comes from Zwane and Mestry (2025). They begin from earlier findings that SGB members in many non-fee schools lack the financial expertise to solve practical problems, and then used a phenomenological design to ask stakeholders how school management teams (SMTs) and governing bodies work together on school finances. Their finding is that many non-fee schools have made little tangible effort to promote that collaboration, and that SMTs as a structure are not permitted to attend governing body meetings, because the Act does not provide for them there, even though principals sit on the body and other SMT members may be elected as teacher representatives (Zwane and Mestry, 2025). So the people with day-to-day knowledge of how the school runs are partly outside the room where its budget is discussed, which seems an odd result for a policy meant to redress (Republic of South Africa, 1996).

The statute itself leans towards capacity. Under section 21 a governing body may apply for extra functions, such as maintaining property or buying textbooks and equipment, and the head of department may refuse only where the body lacks the capacity to perform the function effectively (Republic of South Africa, 1996). In plain terms, the more a governing body can already do, the more it is allowed to do. Add section 36’s expectation that schools top up state funding from their own means, and the structure seems to hand the biggest freedoms to those best placed to use them (Aina and du Plessis, 2023; Govender, 2022). That is basically what I mean when I say benefits follow capacity.

Two cautions are fair here. These two studies use different designs and neither compares the two kinds of school directly, so putting them side by side is my own comparison and not theirs. And Aina and du Plessis (2023) report strong professional influence over finances, which says nothing yet about whether the money was spent well, since strong professionals can steer it badly as easily as well. Even so, the contrast is consistent with Govender’s (2022) argument that poor and marginalised communities need more influence over policy outcomes, and I did not find anything in the sources that pushes against it (Zwane and Mestry, 2025).

Early years, reading and the learning gap

The clearest national evidence on learning comes from Böhmer and Wills (2025), who compared Grade 4 PIRLS data from 2016 and 2021. Average reading achievement fell by 31 points, from 320 to 288, which they put at roughly half to six-tenths of a year of learning, and the share of very low achievers, below 200 points, doubled from 13.4 to 26.5 per cent. Scores in the wealthiest 20 per cent of schools did not rise, while they fell significantly in the poorest 70 per cent (Böhmer and Wills, 2025). The pandemic is obviously part of this. Still, the authors point out that South Africa already had the highest socio-economic inequality in reading of all participating countries before it hit, which fits Christie’s (2021) argument that the pandemic exposed a pattern that was there already.

Dagada (2022) goes inside classrooms in the Vhembe district of Limpopo. She worked with twelve participants from four schools and records that Foundation Phase teachers faced overcrowded classrooms, a lack of resources, too little support from school management teams and parents, and poor knowledge of reading instruction, and that teachers were unable to administer the assessment tasks the department prescribes (Dagada, 2022). Her suggested remedy is workshops. That seems a modest answer to a big problem, but it matches what Tshabalala and Nthontho (2024) found about thin support and training. Taken together, the two studies hint that the Grade 3 reading aim in the Plan sits on a base of teacher knowledge that policy has not yet supplied (National Planning Commission, 2012).

The Plan also wants two years of preschool for every child, and Matjokana (2023) shows what the early end of the system looks like in rural areas. She interviewed three education officials, two early childhood development (ECD) managers and four practitioners, and found that unregistered rural centres lack the basic resources needed to give children foundational knowledge, so that children move into Grade R with limited foundations. Her recommendation is to upskill practitioners in the registration processes so that centres can reach quality programmes, infrastructure, water and sanitation (Matjokana, 2023). Registration, in other words, looks like the doorway to support, and the centres least able to walk through it appear to be the ones that most need what is behind it. That is a capacity problem again, now at the very start of schooling (National Planning Commission, 2012).

One caution about all this. PIRLS tests Grade 4 reading (Böhmer and Wills, 2025), and the qualitative studies come from a few schools in a few places (Dagada, 2022; Matjokana, 2023), so none of them measures the whole system, and reading outcomes depend on a lot besides policy. What the studies do give is a consistent pattern. Gaps appear to open early, in the early years and the Foundation Phase, and they seem to widen when something shocks the system. Whether policy since 1994 has narrowed them at all is a question these ten sources cannot settle, and a long-run trend study would be needed to settle it.

Voice, quality and the democratic question

Quality is the word everyone uses and few people define. Hoffman and Maarman (2024) start from the observation that South African schools lack a shared understanding of quality education, and they apply Sen’s capability approach through individual and focus group interviews with primary and high school teachers in two Cape Town schools. They conclude that limits on teachers’ voices have turned quality into a bureaucratic discourse, which suppresses evaluative reasoning and the freedoms to teach and learn in schools (Hoffman and Maarman, 2024). Two schools cannot speak for the country, but the finding rhymes with the principals who told Tshabalala and Nthontho (2024) that they had no voice when policy was being written.

This is where the democratic part of democratic transformation gets tested. If transformation means that the funding rules became formally redistributive, the record has something to show (Republic of South Africa, 1996). If it also means that teachers, principals and the communities around schools help shape what policy is and how it is judged, the picture is thinner. Govender (2022) says the 1996 funding settlement was shaped heavily by elites despite wide consultation, Hoffman and Maarman (2024) say teachers’ voices are limited, and Tshabalala and Nthontho (2024) say principals are missing from policy development. Three studies at three levels, and roughly the same message.

What follows from this? The studies point, mostly implicitly, towards fairly concrete things. Matjokana (2023) points to help with ECD registration, Teane (2023) to partnerships with universities, and Dagada (2022) to workshops on reading instruction, while Zwane and Mestry (2025) suggest that the Act’s rules on who sits on a governing body deserve a second look. None of it is dramatic. Each one aims at capacity at the point where policy lands, which fits my view that capacity, more than a new policy document, is what has been missing. Whether the department can supply that support at scale is a separate question, and the sources do not answer it.

Conclusion

Basic education policy since 1994 has been strong on principle and weaker on delivery. The Schools Act set out equitable state funding and redress, and the National Development Plan later turned the aim into reading by Grade 3 and two years of preschool (Republic of South Africa, 1996; National Planning Commission, 2012). The recent studies reviewed here suggest that the distance between those aims and what happens in classrooms is mostly a distance in capacity, in the sense of what governing bodies, principals, teachers and ECD practitioners are equipped to do (Tshabalala and Nthontho, 2024; Christie, 2021).

On who benefits, the evidence leans towards schools that already govern well. Quintile 5 governing bodies appear to use their powers with confidence, non-fee governing bodies and their management teams seem to struggle, and the learning data show the poorest schools losing most ground (Aina and du Plessis, 2023; Zwane and Mestry, 2025; Böhmer and Wills, 2025). That pattern is consistent with a policy design that rewards capacity and with a policymaking process that gave poor communities limited influence (Govender, 2022). I would hold these conclusions with some care, because most of the studies are small and no source here tracks outcomes across the whole period. Even so, the direction of the evidence looks clear enough to suggest that the next stage of reform should be judged by what it does for capacity at the school and district level, and by whether those closest to the school have a say in it (Hoffman and Maarman, 2024).

Reference list

Aina, A. and du Plessis, A. (2023) Financial management in fee-paying public schools in South Africa: how responsible is the school governing body? South African Journal of Education, 43(4), article 2307. doi:https://doi.org/10.15700/saje.v43n4a2307

Böhmer, B. and Wills, G. (2025) COVID-19 and inequality in reading outcomes in South Africa: PIRLS 2016 and 2021. Large-scale Assessments in Education, 13(1), article 24. doi:https://doi.org/10.1186/s40536-025-00256-9

Christie, P. (2021) The pandemic as a portal for change: pushing against the limits of ‘normal schooling’ in South Africa. Education as Change, 25, article 8718. doi:https://doi.org/10.25159/1947-9417/8718

Dagada, M. (2022) Foundation Phase teachers’ challenges in teaching reading in South Africa. South African Journal of Education, 42(Supplement 1), article 2219. doi:https://doi.org/10.15700/saje.v42ns1a2219

Govender, L. (2022) Using critical policy historiography in education policy analysis: a South African case study. Education as Change, 26, article 10232. doi:https://doi.org/10.25159/1947-9417/10232

Hoffman, S. and Maarman, R. (2024) Teachers’ voices and quality education in the basic education discourse in South Africa. Perspectives in Education, 42(4), pp. 333-348. doi:https://doi.org/10.38140/pie.v42i4.8048

Matjokana, T.N. (2023) Early childhood care and education policy intentions and the realities in rural areas. Perspectives in Education, 41(2), pp. 258-274. doi:https://doi.org/10.38140/pie.v41i2.6741

Teane, F.M. (2023) Partnership as a strategy to overcome the difficulties associated with policy implementation: South African teachers’ views. South African Journal of Education, 43(2), article 2125. doi:https://doi.org/10.15700/saje.v43n2a2125

Tshabalala, G. and Nthontho, M. (2024) Primary school principals’ experiences of the implementation of educational policy change. South African Journal of Education, 44(2), article 2376. doi:https://doi.org/10.15700/saje.v44n2a2376

Zwane, T.T. and Mestry, R. (2025) (Dis)empowering school management teams of non-fee-paying schools to collaborate with school governing bodies to manage school finances effectively. Perspectives in Education, 43(4), pp. 152-169. doi:https://doi.org/10.38140/pie.v43i4.9482

Legislation and case law

National Planning Commission (2012) National Development Plan 2030: our future – make it work (executive summary). Pretoria: The Presidency. Available at: https://www.gov.za/sites/default/files/gcis_document/201409/ndp-2030-our-future-make-it-workr.pdf (Accessed: 18 October 2026).

Republic of South Africa (1996) South African Schools Act 84 of 1996. Available at: https://www.gov.za/sites/default/files/gcis_document/201409/act84of1996.pdf (Accessed: 18 October 2026).