South Africa’s education system has no shortage of policy ambition. It has strategies, data systems, institutional mandates, and a growing national agenda for digital transformation. Much of the system, however, is still fragmented, writes Mark Burke in this latest opinion article.

Learner records sit in different databases. Schools, colleges, universities, funding agencies, and qualification authorities collect overlapping information through separate processes. Provinces work with uneven capacity and uneven systems. Some digital tools are improving, but mostly within institutional silos. That is the setting in which Digital Public Infrastructure, or DPI, becomes relevant. The question is whether South Africa can use the current push for digital identity, data exchange, and digital payments to make education services more coherent from school entry to post-school transitions and into work.

This is a practical question. People move through the education system over years and across institutions. Their information needs to be created once, updated carefully, and used lawfully in many contexts. Traditional e-government projects tend to digitise what already exists inside departments. DPI starts with shared functions that can support many services across the government, including identity, authentication, data exchange, verification, and payments.

In education, those building blocks can reduce the need to repeatedly complete forms entering the same information, make credentials easier to verify, enable records to move across transitions (such as from school to university), and support services designed around the learner’s actual path rather than bureaucratic boundaries of institutions. Used well, DPI offers a way to connect systems.

There is also a reason for caution. Education technology has repeatedly been sold as a quick fix to better outcomes. It is not. UNESCO’s 2023 Global Education Monitoring Report rightly argues that technology should be treated as a means rather than an end, and that decisions should be judged by whether they are appropriate, equitable, evidence-based, and sustainable. That caution is especially relevant in South Africa, where digital reform runs into persistent inequality, uneven access to infrastructure, and varied administrative capacity. A DPI approach is useful only if it addresses those conditions directly. 

Education as a connected public service?

Education is often discussed as a sector, but from a service perspective, it is a chain of linked events. A child is registered, enrolled, assigned to a school, marked present or absent, assessed, supported, transferred, examined, certified, and then perhaps admitted to college, university, or training. Some learners need nutrition, transport, disability support, or psychosocial services. Some need bursaries. Some need their qualifications verified years later.

These are different transactions, handled by different institutions, but they belong to the same person over their lifetime. A system that treats each event as administratively separate places the burden of connection on learners, families, educators, and education officials. A system that builds shared digital foundations can reduce that burden, provided the underlying rules and data are trustworthy.

International experience suggests that this is not an abstract possibility. Estonia’s X-Road architecture shows the value of secure interoperability between registries without collapsing everything into one database. Mexico’s digitisation of professional licensing demonstrates how standardised credentials and linked public infrastructure can reduce long waiting times and repetitive administrative steps. India’s DIKSHA (Digital Infrastructure for Knowledge Sharing) platform shows that a large and diverse system can build shared infrastructure for teachers, learners, and digital content while still accommodating local differences.

These cases differ in political setting, capacity, and scale, but they do point toward a common lesson. Education services improve when governments invest in reusable public infrastructure that underpins sector applications, rather than commissioning one disconnected platform after another.

A few principles recur across those experiences. One is to design services around life events and transitions rather than departmental mandates and policies. Another is the significance of a durable identity and authentication layer that allows learners and educators to be legible across institutions and over time. A third is interoperability based on open standards, shared definitions, and modular architecture. These technical choices shape whether a learner can move through the system without repeatedly having to provide administrative details and whether institutions can trust each other’s records.

Open standards deserve emphasis because they affect cost and flexibility. South Africa has had many public digital projects where institutions become dependent on software that is hard to adapt, expensive to maintain, and difficult to connect to other systems. In education, that risk is multiplied because the sector spans many institutions, many vendors, and very long technology lifecycles. A standards-based approach does not necessarily eliminate vendors. It does reduce the cost of change. It also makes the reuse of resources more plausible. If identity, credential, and exchange layers are built on common standards, new services can be added without forcing every institution to migrate at once. That is particularly important in a fiscally constrained environment, where reform has to proceed incrementally and where procurement mistakes have long-term consequences.

South Africa already has more building blocks than is often assumed

South Africa is not starting from a blank page. In basic education, the Department of Basic Education (DBE) already manages significant information assets. Keeping track of developments and dynamics in the basic education sector incorporates data from various sources such as the Education Management Information System (EMIS), the Learner Unit Record Information and Tracking System (LURITS), the South African School Administration and Management System (SA-SAMS), Provincial Data Warehouses, information systems developed for the Data-Driven Districts Initiative, and the National Senior Certificate Database. Taken together, these systems amount to an existing, if uneven, data backbone for more than 13-million learners. However, they are not yet a coherent service architecture.

These systems are being upgraded. The SA-SAMS modernisation programme has been shifting toward a web-enabled and cloud-based environment using OpenEMIS technology. According to DBE, the revised system is meant to improve data accuracy, consistency, real-time reporting, process efficiency, and data security. That does not solve the wider interoperability problem on its own. It does, however, show that basic education data systems are already in transition. The question is how this series of incremental upgrades can be tied to a larger national architecture for education data and service delivery.

The post-school system has similar foundations. Technical and Vocational Education and Training Management Information System (TVETMIS) is a unit-record system that stores data on colleges, campuses, programmes, subjects, staff, and students in the TVET sector. The Department of Higher Education’s and Training’s HETMIS as an integrated information system intended to bring together unit-level data across higher education, TVET, Community Education and Training (CET), private colleges, and Sector Education and Training Authority (SETA) systems under common information standards. The logic is to move away from fragmented aggregate reporting and toward comparable, reusable unit-record data that supports planning and management across the sector. This is closely aligned with the logic of DPI.

South Africa also has a qualifications infrastructure. The South African Qualifications Authority’s National Learners’ Records Database contains records related to qualifications, part qualifications, professional designations, and the structures of the National Qualifications Framework (NQF). It is not yet a seamless lifelong learner record that functions across all public and private use cases. But it is relevant to any future system of portable digital credentials, streamlined verification, and learner-controlled record sharing.

The broader point is that South Africa already has many of the assets needed for a stronger education service architecture. The weakness is in fragmentation, uneven quality, inconsistent standards, and limited reuse across the learner’s lifecycle.

Why digital transformation is still necessary

Those existing assets matter because South Africa’s education outcomes remain uneven and, in some areas, persistently poor. Participation rates are high, especially up to age 15, and completion of twelve years of education has improved over time. Yet high participation has not translated into consistently strong learning outcomes.

The system still struggles with progression, dropout, overcrowded classrooms, provincial inequality, and weak performance in subjects that shape access to further study and employment. DBE’s review of progress in the sector notes continued demographic pressure, declining infrastructure spending in real terms, and major regional shifts in enrolment, especially into Gauteng. Beyond getting learners into classrooms, the question is whether the system can respond earlier, target support more effectively, and use information to improve outcomes.

The infrastructure environment reinforces that point. South Africa has broadly achieved extensive mobile broadband coverage, but meaningful connectivity remains uneven. The DBSA digital infrastructure study argues that the main constraints now are affordability, device access, digital literacy, quality of service, and the lack of reliable data on the connectivity status of public facilities, including educational institutions.

That distinction matters. An improvement agenda built on assumptions of universal school connectivity or device access will not survive contact with actual conditions. Any education DPI worth building in South Africa must work across unevenly resourced settings. It has to be hybrid by design, capable of offline or low-bandwidth use where necessary, and careful about the gap between nominal connectivity and meaningful access.

Administrative structure makes the challenge harder. Basic education is shaped nationally but administered through nine provincial departments, each with varying levels of capacity. Post-school education is distributed across universities, TVET colleges, CET institutions, SAQA, quality councils, SETAs, and funding systems such as National Student Financial Aid Scheme (NSFAS).

In this setting, a single replacement platform is unlikely to succeed. What matters more is a common language across systems, incorporating standards, identifiers, registries, exchange rules, and governance arrangements that allow institutions to remain distinct while making information portable and usable. The aim is not to do away with institutional boundaries. It is to stop those boundaries from forcing learners to start over at every administrative transition.

There is a governance benefit here as well. When institutions exchange data through defined standards and documented permissions, responsibility is easier to assign. Errors can be traced. Records can be corrected closer to the source. Audit trails become more credible. That matters in education because mistakes in identity, enrolment, results, or funding status are more than minor inconveniences. They can delay admission, interrupt teaching, block support, or sometimes quietly push a learner out of the system for good.

The opening created by the national DPI agenda

The broader national digital transformation context now presents an unusual opening. South Africa’s digital transformation roadmap is built around digital identity, data exchange, digital payments, and trusted digital channels. Education is integral to that agenda. The roadmap explicitly points to the use of verified credential wallets for educational certificates, more automated school enrolment, and better-targeted bursary allocation. MyMzansi’s Initiative 1 is direct in describing a verified document-sharing and wallet system that should include education certificates and other forms of verified attainment. Education can build on national DPI foundations, provided the sector is clear about its own requirements early enough to influence the shape of those foundations.

Education should therefore engage the wider DPI agenda with focus. Digital identity in education is about secure recognition of learners, educators, and institutions across time and across service contexts. Data exchange is about internal administrative efficiency and allowing records to move between schools, between schooling and post-school systems, and between education and funding authorities without repeated manual re-entry of identity information. In education, digital payments involve bursaries, allowances, fees, and other parts of the learner’s path. If the sector does not clearly define its own service architecture, it could end up adapting belatedly to a national stack built with other priorities in mind.

What a coherent education architecture would look like

A practical response would be to build a coherent national education data architecture anchored by an interoperability platform. This should not be a single mega-system. It needs a federated layer that connects existing systems through shared standards, governed interfaces, and clear rules about what counts as authoritative data. In that sense, it would be less a database than an exchange and trust layer through which schools, provinces, colleges, universities, SAQA, NSFAS, and other actors could use existing data assets more coherently.

The first requirement is a National Education Interoperability Framework (the public health sector adopted such a framework for digital intolerability in 2022). South Africa has fragments of standardisation in different parts of the system, but not a sector-wide framework spanning schooling, TVET, higher education, quality assurance, and student funding. Such a framework would need to define data standards, identifiers, exchange protocols, semantic models, metadata rules, audit requirements, and access conditions. It should be federated so that provinces and institutions retain their systems but operate under common rules that enable interoperability. Without that layer, local digital improvements will continue to produce useful islands rather than a connected public service.

The second requirement is to stabilise and govern a set of core registries. At minimum, South Africa needs a learner registry, educator registry, institution registry, qualification and credential registry, curriculum registry, and an assessment and support registry. Parts of these already exist in distributed form. The task is to make them authoritative, linkable, and governed. A learner registry would support longitudinal tracking across transitions. An educator registry would improve human resource planning. An institution registry would strengthen geospatial planning and service mapping. A qualification registry would support digital verification. An assessment and support registry would make it easier to move from annual reporting to earlier intervention. These are foundational since, without reliable registries, more advanced service design rests on weak data.

The third requirement is a modular transition path. South Africa should avoid trying to replace SA-SAMS, LURITS, TVETMIS, the NLRD, and post-school systems with one new platform. A more credible approach is to upgrade existing systems, enable interoperability, improve data quality, and create an exchange layer that allows services to function across them (unless, of course, the underlying technology is completely outdated). UNESCO’s work on reimagining EMIS is useful here. A modern EMIS should be integrated, modular, timely, and useful for management and learning support byond annual statistical reporting. 

Where public value would become visible

This architecture only matters if it improves real services. One obvious domain is classroom and school management. Better integration between learner attendance, assessment, curriculum coverage, school administration, and district support could help identify risk earlier and support action before problems become entrenched. Data should not move only upward for compliance. It should return to schools and districts in forms that improve decisions about learners and teaching. The logic already visible in the Data Driven Districts model is useful here, as its focus is on the system responding sooner to those at risk of falling behind or dropping out.

A second domain is learning and teaching support materials. This is one of the clearest cases where digital infrastructure could improve service delivery, but also one of the clearest cases where politics will matter. GTAC’s review found fragmented provincial procurement models, poor inventory data, weak use of available digital modules, and heavy reliance on manual spreadsheets. The Limpopo textbook crisis remains a reminder of what happens when procurement, data, communication, and accountability fail together. A stronger digital architecture could support common catalogues, better stock visibility, clearer distribution tracking, and more rational planning across print and digital materials. It could also support shared repositories of approved content and more consistent metadata for learning materials. But such a shift would alter procurement practices and market positions, in which case resistance would be expected.

A third domain is specialised learner support. If education data can be exchanged lawfully and with appropriate controls, the state has a better chance of identifying learners who need targeted support, whether due to absenteeism, learning difficulties, disability-related needs, or other barriers to participation. That requires more than analytics. It requires clear rules about purpose limitation, access rights, audit trails, and consent where appropriate. The point is not unrestricted data sharing. It is better governed, purpose-specific sharing that reduces the burden on families and front-line staff while making earlier support more feasible.

A fourth domain that spans the full education lifecycle is a National Learner Record and Credential System. South Africa’s digital identity roadmap already anticipates the use of verified credential wallets. Over time, every learner should be able to maintain a portable, verifiable record of achievement across schools, TVET, universities, occupational training, and shorter-form credentials, where appropriate. That would simplify transitions, reduce fraud, improve recognition of prior learning, and ease movement between education and employment.

Student funding is another part of the picture. Although the supporting systems differ, the service problem is familiar. For instance, students and families submit duplicate information, institutions verify pieces of it, and delays or inconsistencies carry real costs. NSFAS has itself framed digital transformation as necessary to improve the student experience and reduce administrative delays. A better education interoperability layer would not solve all funding problems, which are partly political and fiscal, but it could reduce duplication between identity checks, enrolment confirmation, academic status, and eligibility processes. This is significant because it could influence whether students remain in the system, register on time, and move through post-school education with some predictability. A coherent education architecture should therefore be designed with the funding journey in mind, in addition to the learning journey.

The politics will matter as much as the architecture

The case for a more coherent education data architecture is strong. The harder question is political. Any serious shift toward interoperable education services will affect procurement, administrative discretion, and institutional power.

Provincial systems may resist common standards if they appear to narrow room for local control. Publishers and content providers may resist moves toward shared repositories or more transparent digital ecosystems. Teacher unions may worry that richer data systems will be used for surveillance rather than support. Administrative staff may reasonably be concerned about what automation means for their roles. These are issues to be considered from the outset.

That is why the transition should begin with a small number of visible, high-value use cases rather than with a promise to redesign the entire system at once. Good candidates include digital credential verification, simpler learner transfer processes, more reliable enrolment data, better Learning and Teaching Support Materials (LTSM) inventory management, and quality information flows between education institutions and student funding systems. These are practical problems with direct effects on users. They also allow government to test standards, privacy controls, interfaces, and audit arrangements in real settings. If early uses reduce paperwork, improve accuracy, and shorten delays, they will create space for broader reform. 

Trust, in this area, depends heavily on privacy and accountability. UNESCO notes that few countries legally guarantee the privacy of educational data, and that many education technology products collect far more information than learners and families expect. South Africa has a broader data protection framework, but education-specific rules on access, role-based permissions, data minimisation, retention, auditability, and redress will still need to be made more concrete if interoperable systems are to be trusted. 

Towards digitally-enabled education services in South Africa

South Africa can use digital public infrastructure to improve public education. The country has enough existing systems, registries, and policy momentum to make that ambition credible. But the test is not whether education becomes more digital. It is whether services become more coherent, more usable, and more responsive to what learners, teachers, families, and institutions actually need.

That depends on basic things such as standards that hold, registries that can be trusted, systems that connect, and governance that makes the use of data lawful and understandable. If those pieces of the puzzle are put in place, the wider national DPI agenda could help education move from fragmented administration toward a more connected and integrated public service. If they are not, it will add another layer of digital complexity to a system that already has too much of it.

Written by Mark Burke, a researcher and advisor with expertise in digital governance, and a focus on public-sector digital transformation. His research interests are digital identity, privacy, and citizenship in the digitalisation of public services.