The department of trade, industry & competition’s Industrial Development Strategy (IDS) arrives at a moment when few would dispute the need for industrial renewal. Economic growth has stagnated for close to two decades. Manufacturing’s share of GDP has declined from more than 20% in the early democratic period to about 13%. Fixed investment remains stubbornly low. Infrastructure failures continue to undermine competitiveness. Unemployment remains among the highest in the world. The IDS deserves credit for recognising these realities. Unlike some earlier policy documents, it avoids grandiose claims and acknowledges the mixed outcomes of the sectoral master plans that have dominated industrial policy to date. Yet reading the document leaves an uncomfortable impression. The diagnosis is often sound. The theory of change is not. There is a certain irony in the title itself. For generations of development economists IDS referred to the Institute of Development Studies at Sussex University. From the 1960s onwards Sussex scholars such as Hans Singer and Christopher Freeman helped transform thinking about development, technological change and industrialisation.Later generations extended these ideas through work on innovation systems, technological capabilities and institutional learning. Their central insight was straightforward. Countries don’t become prosperous because they possess resources, build industrial parks or designate priority sectors. They become prosperous because firms, workers and institutions acquired capabilities. The Sussex IDS left a significant imprint on South African development thinking, not least through figures such as former president Thabo Mbeki. By that standard, South Africa’s IDS feels curiously detached from the modern literature on industrial transformation. The department’s IDS is organised around decarbonisation, diversification and digitalisation. None of this is controversial. The difficulty is that these are trends affecting virtually every economy on earth. The IDS does not explain how South Africa will build the productive capabilities needed to respond to those trends. Throughout the document one encounters sectors, incentives, infrastructure projects, financing mechanisms and co-ordination structures. What is lacking is the concept of capability. This matters because industrialisation is not primarily an investment problem. It is a learning problem. The successful industrialisers of East Asia did not simply attract capital. They created firms capable of absorbing technology, improving production processes, developing products, training workers and competing internationally. The critical question was never how much investment entered the economy. It was what domestic firms were capable of doing with that investment. Import technology, use and adapt these well, then outperform the original providers. For nearly three decades our country has also lived with a second assumption that scientific excellence will somehow translate into industrial innovation. Considerable public resources have been invested in universities, the National Research Foundation, centres of excellence, research chairs and major scientific infrastructure. The results are impressive in many respects. South Africa enjoys internationally recognised strengths in astronomy, astrophysics, infectious diseases, virology and biodiversity science. But scientific excellence and industrial transformation are quite different phenomena. The difficulty is not merely transferring basic research from laboratory to market. The difficulty is that industrial innovation depends on engineering capability, production systems, standards, testing, procurement, management practices, supplier networks and learning within firms. The new strategy seems to recognise the problem but never fully articulates it. Innovation is mentioned repeatedly almost as a nice-to-have. AI, technology parks, commercialisation platforms and research partnerships are invoked as signs of modernity rather than integrated into a coherent strategy for industrial transformation. The treatment of AI is particularly revealing. It appears in the document because no strategy written in 2026 can avoid mentioning it. Yet the implications are scarcely explored. Peril or promise? There is no sustained discussion of how AI might transform mining, logistics, manufacturing, agriculture, customs administration or public service delivery. There is little consideration of data infrastructure, computational capacity or the skills required to deploy these technologies effectively. And don’t mention the destruction of jobs. The silence is striking because AI is not another sector. It is a general-purpose technology. Countries that successfully deploy it will enjoy productivity gains across multiple sectors. Those that fail to do so will find themselves trailing further behind. These omissions become even more significant when viewed through a political economy lens. The strategy acknowledges concentration, barriers to entry and declining competitiveness. Yet it largely avoids the deeper question of why South Africa has struggled to achieve structural transformation despite 30 years of industrial policy. There is little discussion of financialisation, the persistence of the minerals-energy complex, weak competition in key sectors, or the incentives that shape private investment decisions. The document seeks to improve the performance of the existing industrial order rather than interrogate its underlying structure. Perhaps the most important omission of all concerns education and skills. South Africa today is not a low-wage economy. Compared with rival emerging countries labour costs are high. The country cannot realistically compete on the basis of cheap labour. Nor should it aspire to do so. Its future competitiveness must therefore rest on productivity, and productivity gains ultimately depend on capability. Thousands of our domestic firms operate below international productivity benchmarks. Significant gains could be realised within five years through better maintenance systems, improved logistics, lean manufacturing, digitalisation, quality management and workforce upgrading. None of these requires scientific breakthroughs. The binding constraint is not scientific discovery but the availability of competent workers, technicians, engineers and managers. Industrial policy should therefore be judged not only by the investments it attracts but by whether firms become more productive, more technologically capable and more competitive over time. The critical challenge is not merely to build new industries. It is to raise the performance of existing ones. Our multinational corporations do this every day. Evidence that capability matters can be found in firms such as MTN, Nando’s, Sappi and Bidfood, all of which compete successfully in demanding international markets. Their success owes less to protection than to organisational capability, quality systems, managerial competence and continuous learning. Capability begins in schools, yet 80% of grade 4 pupils are unable to read for meaning, a constraint that disadvantages them through their remaining years of school. No industrial incentive scheme can resolve this. Digitalisation, advanced manufacturing, financial services, AI and knowledge-intensive exports all require talent that is capable of continuous learning, communicating effectively and adapting to technological change. These capabilities cannot be manufactured by job seekers, but are built over years. Our industrial challenges begin long before students enter universities, TVET colleges or workplaces. The weakest link in this chain is arguably the TVET system, which is projected to double in size by 2035. Currently employers report shortages of artisans, technicians and technologists, while many TVET lecturers themselves have limited experience of modern industrial environments. This results in qualifications without capability. Industrial renewal requires TVET colleges that are deeply embedded in workplaces rather than isolated from them. A good starting point would be to recognise that most TVET instructors are former school teachers without industry experience. They should therefore be incentivised to work in industry for three to six months over five years; in reverse, experienced industry personnel should be encouraged to teach or co-teach in TVET colleges. Dysfunctional schooling and TVET are not only social policy concerns. They are industrial policy concerns, for which the East Asian (and European) experiences are instructive. Schooling systems were not viewed as supporting industrialisation, but as central to industrialisation. Literacy, numeracy, technical competence and workplace discipline formed the foundations on which later technological capabilities were built. Arguably the most significant weakness of the IDS is that it treats skills as an enabling factor rather than as the foundation of industrial development itself. The result is a document that is thoughtful, measured and, in many respects, sensible. But it remains a strategy of co-ordination rather than transformation. It assumes that improved infrastructure, targeted incentives and better alignment across government will unlock industrial growth. Those interventions are necessary, but not sufficient. Kahn is a research fellow at Stellenbosch University and visiting professor at the University of Johannesburg.