The National Skills Fund (NSF) was established in terms of the Skills Development Act to finance skills development priorities of national significance. Its role differs from that of a sector education and training authority (SETA), which primarily addresses skills requirements within a particular sector. The NSF operates across sectors and can fund national priorities, including artisan development, occupational programmes, workplace-integrated learning, bursaries, unemployed training, worker education, rural development, small businesses and cooperatives, and research and capacity within the post-school education and training system.

Its principal source of funding is the Skills Development Levy paid by employers. For most liable employers, that levy is 1% of payroll, with 20% of levy revenue allocated to the NSF. The Fund also earns investment income and receives other income permitted within the legislative framework.

The governance model is interesting because the NSF is not a conventional public entity governed by an independent board of directors. It operates within the national Department of Higher Education and Training. The minister of higher education and training is the executive authority, while the director-general of the department is the NSF’s accounting authority and carries statutory accounting responsibility for the financial management and proper use of its resources. The NSF administration manages the actual funding processes, contracting, projects, monitoring, and payments.

The National Skills Authority (NSA) occupies a separate position. It does not govern the NSF, act as its board, or approve individual allocations. The Skills Development Act does, however, require it to advise the minister on the strategic framework and criteria for allocating NSF money. It also has wider responsibilities for monitoring national skills strategy and investigating matters arising from the application of the Act. Beyond that, the Auditor-General audits the NSF and Parliament exercises oversight.

Uncomfortable Question

On paper, that appears to be a substantial accountability chain. But there is an uncomfortable question sitting inside the governance model: what protections does the NSF have against a corrupt minister?

This is not an allegation against the current minister or any previous minister. It is a test of the architecture itself. Governance should be designed on the assumption that somebody occupying a position of authority could eventually attempt to abuse it, and its strength lies in how difficult it makes that abuse.

The NSF has no independent governing board with fiduciary responsibility sitting between the political executive and a fund receiving billions of rand annually. The minister is the executive authority, the accounting authority is the director-general of the minister’s own department, and the NSF administration operates within that department. The NSA advises, monitors, and can investigate, but it does not govern the fund or exercise fiduciary authority over individual allocations.

There are important protections. The director-general is bound by the Public Finance Management Act; funding decisions remain subject to legislation, internal controls, and administrative law; the Auditor-General provides external audit: Parliament provides oversight and unlawful decisions can be challenged through the courts. The vulnerability is that several of these protections operate within the same departmental hierarchy or identify wrongdoing after decisions have been taken and money has potentially moved.

That deserves particular attention, because NSF funding is not confined to open competitive calls. Its funding architecture also provides for targeted or closed calls, programmatic funding arrangements, and unsolicited proposals. There can be entirely legitimate reasons for each of these mechanisms, but greater discretion requires greater scrutiny.

If a future minister were corrupt or politically compromised, could influence be exerted over funding priorities or down the departmental hierarchy? Could preferred organisations repeatedly gain access to funding? Could a series of individually compliant decisions cumulatively direct substantial amounts towards favoured constituencies? Most importantly, who within the present structure has both the independence and authority to recognise that pattern and stop it before the money leaves the fund?

Piggy Bank

An NSF receiving billions annually should never be capable of becoming anybody’s personal piggy bank. The governance test is whether the controls make that practically impossible, rather than whether abuse can eventually be detected through an audit, parliamentary inquiry, forensic investigation, or court process. With approximately R72 billion having flowed through the NSF, the absence of an independent governing board deserves serious examination.

It also makes the next question considerably more important: what did South Africa actually get for the money?

The NSF exists because South Africa needs to develop the human capability required by its economy. Training is the intervention and economic participation is the intended result, which means we need to be much more demanding about what constitutes success.

If 10 000 unemployed people enter training, we have 10 000 enrolments. If 7 000 complete, 5 000 qualify, 3 000 enter employment, and 2 500 remain employed two years later, we have five very different measures of performance. Yet skills development is repeatedly reported through the numbers at the beginning of that chain: money allocated, programmes funded, and learners enrolled.

Consider artisan development. We need to know how many apprentices entered, completed their workplace experience, attempted their trade tests, and passed. We should then know how many indeed worked as artisans, in which trades and provinces, what each qualified artisan ultimately cost, and whether the intervention reduced the occupational shortage it was intended to address. The same methodology should apply to every major NSF programme.

But R72 billion of expenditure later, since inception, we must ask, what did South Africa get from the NSF?

Since the Skills Development Levy system began in 2000, approximately R72 billion in total revenue has flowed through the NSF. We know broadly where much of the money went, including artisan development, learnerships, training for unemployed people, bursaries and scholarships, Technical and Vocational Education and Training (TVET) and community colleges, workplace-integrated learning, worker education, small businesses and cooperatives, infrastructure, training equipment, research and institutional capacity. In 2021/22, approximately R3.3 billion also went towards a funding shortfall in the National Student Financial Aid Scheme (NSFAS).

Those are descriptions of expenditure. They do not tell us what changed because the money was spent, and after 25 years that is surely the more important question.

What Did Each Successful Outcome Cost?

Suppose two programmes each receive R100 million. Programme A enrols 5 000 people and produces 1 000 qualified people who enter sustained employment. Programme B enrols 10 000 people but produces only 500 people who enter sustained employment. Measured by enrolments, programme B appears twice as successful. Measured by sustained employment, programme A produced twice the result.

That changes how we should evaluate billions of rand of public investment. For unemployed learners, we should know whether they found employment or became self-employed, whether their income increased, whether they were still economically active 12 or 24 months later, and whether there was genuine demand for the occupation in which they were trained. Without longitudinal tracking, we know who entered training, but we know far less about whether the intervention changed their economic position.

And the same test must be applied to institutions.

The NSF has invested and committed more than R8 billion in the public TVET college sector through infrastructure, equipment, apprenticeships, learnerships, occupational programmes, workplace-integrated learning, and institutional capacity. Some of that may have created significant long-term value, but the investment still has to be measured against what it produced.

If a college received workshops and equipment, did its training capacity increase? Were more artisans produced? Did throughput and trade-test pass rates improve? Is the equipment still operational and being used, and what did each additional successful graduate cost? A building and a piece of equipment are assets, but neither becomes a skills outcome until somebody uses them to become economically productive.

NSFAS Shortfall

The same questions apply to the approximately R3.3 billion NSF contribution towards the 2021 NSFAS funding shortfall, associated with 48 919 students. We should know how many completed, what they studied, how many entered employment, how many qualified in occupations identified as national priorities, what the completion cost per student was, and what economic return the investment generated.

This is important because every funding decision carries an opportunity cost. R3.3 billion allocated there could not simultaneously finance artisan development, workplace training, unemployed youth programmes, or emerging occupational requirements elsewhere. That does not make the NSFAS allocation wrong, but it does make demonstrating its return essential.

And we must also ask whether the money followed skills shortages.

South Africa generates enormous quantities of labour-market information. SETAs produce sector skills plans, government identifies occupations in high demand, employers report vacancies and shortages, universities and colleges produce enrolment and completion data, and Statistics South Africa measures the labour market.

If South Africa identifies a shortage of a particular occupational skill, we should therefore be able to measure its scale, intervene, and then measure it again. Did the shortage reduce? By how much? If it did not, why did we continue funding the same intervention?

This becomes even more important as artificial intelligence changes occupations, automation transforms manufacturing, mining, and logistics, renewable energy creates new technical requirements, electric mobility changes automotive skills, and cybersecurity demand expands. The NSF cannot spend billions solving yesterday’s skills shortages. It needs to anticipate tomorrow’s, which requires a continuous cycle of research, prediction, investment, training, employment, and measurement.

Failure Should Produce Knowledge

Not every NSF programme will succeed, nor should we expect it to. A developmental fund should be prepared to take calculated risks. Demand can be overestimated, employers may not absorb learners, technology can change, and providers can underperform.

Failure itself is therefore not the greatest concern. Failure without learning is. When a project performs poorly, we should know whether the provider was subsequently funded again, whether the qualification was redesigned, whether the original occupational assumptions were reconsidered, and whether future allocations changed as a result. A R50 million intervention that fails is expensive. One that fails, teaches us nothing, and is repeated is considerably worse.

The NSF should ultimately be capable of demonstrating a complete line of sight from the employer levy being collected to a national skills priority being identified, a project approved, a provider contracted, a learner recruited and trained, a qualification achieved, employment obtained, and that employment sustained.

South Africa already holds much of the information required across the NSF, SETAs, Quality Council for Trades and Occupations, training providers, NSFAS, Statistics SA, employers, and other government systems. The challenge is connecting it lawfully and intelligently.

A national skills system through which approximately R72 billion has flowed should have a measurement architecture capable of showing, by occupation and province, the money invested, enrolments, completion and qualification rates, employment after six, 12, and 24 months, cost per successful outcome, employer demand, occupational shortages and provider performance. Then we could stop measuring success primarily by how much money was spent and start making better decisions about where the next rand should go.

The Real Audit is an Outcomes Audit

The Auditor-General's role remains indispensable. Money must be accounted for, procurement must comply with the law, financial statements must be reliable, and irregular expenditure must be identified. But a completely clean financial audit would still not tell us whether the NSF succeeded.

A project can be lawful, correctly procured, and accurately accounted for while producing little economic value. The NSF therefore requires another form of audit alongside the financial one: an outcomes audit that asks whether the investment actually worked.

And the NSA must keep a tighter control over the NSF within the boundaries of its mandate.

The NSA and the NSF are closely connected, but they perform very different functions. The NSF is the funding mechanism through which billions of rand are directed towards national skills-development priorities, while the NSA provides an important layer of strategic advice and oversight around that system.

The NSA does not govern the NSF, manage its money, or approve individual funding allocations. Under the Skills Development Act, however, it advises the minister of higher education and training on the strategic framework and criteria for allocating money from the NSF, while also monitoring implementation of the national skills strategy.

In practical terms, the NSF deploys the money, while the NSA should be interrogating whether the framework guiding those investments reflects South Africa’s actual and emerging skills requirements.

This makes the NSA particularly important to business, because organised business holds five voting seats on the authority and therefore has a formal route through which employer and labour-market intelligence can influence the strategic direction within which NSF funding decisions are made.

Dr Ivor Blumenthal is a specialist in labour law, skills development, organisational strategy and workplace governance, with senior leadership experience including Regional HR Manager, CEO of the Services SETA, EU representative for South Africa, Vice President of WorldSkills and Johannesburg City Councillor. He was also a former member of the National Skills Authority. His expertise spans employment relations, regulatory compliance, human capital development and institutional reform, underpinned by legal, psychological and commercial expertise, with a longstanding passion for radio and journalism.