What kind of industrial policy is required to reverse deindustrialisation in South Africa? Bhaso Ndzendze and Tshilidzi Marwala recently opined that South Africa’s industrial policy needs to adjust to the “realities” of the Fourth Industrial Revolution (4IR) by using artificial intelligence and big data to identify markets. This, they suggest, will create and support jobs in “smart” factories (based on the “Internet of Things”) and combat unemployment, poverty and inequality.

This is wishful thinking. It imagines that South Africa, like the mythical African country of Wakanda in the superhero film Black Panther, can become rich enough to solve poverty and inequality through immensely powerful technology. Not only do they argue that 4IR offers meaningful economic salvation for South Africa, but they imply it is the only way to save us: we need to adopt “world-class manufacturing” or be condemned to replicate the failures of the past.

This is science fiction. Ndzendze and Marwala correctly observe that South Africa has historically relied on the advantages afforded by natural resources and climate, rather than on “acquired skills and expertise”. They correctly note also that protectionist regimes encouraged “copying for the domestic market instead of innovating at a world-class level”. But this does not mean that the solution is to base industrial policy only on the acquisition of skills and “advanced production methods” suitable for “world-class” manufacturing, where “world class” means the 4IR, smart factories and the Internet of Things.

Producing efficiently and profitably does not require producing at the most “advanced” technological level. There is always a choice of technique and a range of production possibilities. It is economically illiterate to imply there is only one competitive way to produce for global or domestic markets. Indeed, we have argued that an important failure of successive South African industrial policies – and one which Ndzendze and Marwala do not identify – is its “adapt or die” approach to encouraging ever more capital-intensive production.

Clothing production in South Africa has long illustrated the co-existence of different types of technologies (and associated wage levels). Relatively high-wage firms exist alongside lower-wage firms, generally competing in different product lines.

Manufacturing operations can be relatively labour-intensive, meaning that they employ more people per unit of capital than relatively capital-intensive operations. Labour-intensive operations have lower wages and labour productivity (because more people are employed and so the ratio of output to employment is relatively low). Relatively capital-intensive operations have fewer workers and can pay higher wages. The challenge for a high unemployment “labour surplus” country like South Africa is to design an industrial and labour policy regime that accommodates both capital- and labour-intensive operations. Yes, we should support 4IR type “smart” factories, but these are not the only factories we should support – and we certainly should not be consigning labour-intensive production to the dustbin of history, or worst still actively destroying it.

We have argued elsewhere that South Africa’s industrial and labour policy regime is based on the myth that the government needs to destroy labour-intensive manufacturing in order to “force” South African manufacturing “up the value chain” and into the wonderful Wakanda-like world of high-wage, high-productivity “cutting edge” production. Our policymakers appear to have assumed that the efficiency gains of this strategy would be so vast that employment would grow, despite fewer jobs being created for every additional unit of capital. This, despite its radical-sounding Wakanda-esque imagery, proved to be just another version of discredited trickle-down economics. Instead of boosting South Africa on to some sort of “high road” to development, the main achievement of our industrial policies in manufacturing was to demonise labour-intensive firms and drive them out of business.

The clothing industry is a case in point. Clothing production in South Africa has long illustrated the co-existence of different types of technologies (and associated wage levels). Relatively high-wage firms exist alongside lower-wage firms, generally competing in different product lines.

We have visited clothing firms producing for the fashion industry and niche markets (for example, school blazers for private schools) that are relatively capital-intensive. They have computer-aided design, automated delivery systems and extremely sophisticated machinery – some purchased with financial support from the Department of Technology and Industry. They pay wages set by the sectoral bargaining council. They are rarely competing with the low wage, more labour-intensive operations nearby producing, in competition with China, basic items (chinos, T-shirts etc) on simple sewing machines. Indeed, the higher-wage firms benefit from the presence of labour-intensive operations because if they want to hire more workers, they can poach trained workers from the low-wage factories.

Workers in the low-wage factories aspire to those better jobs, but they certainly prefer the low-wage job to no job. Yet their employers get no support from the Department of Technology and Industry and are harassed for paying below the legal minimum. Many have left the industry, and most of the remaining labour-intensive operations have become worker cooperatives in a last-ditch attempt to save jobs. 

In our recent book “Inclusive Dualism: Labour-Intensive Development, Decent Work and Surplus Labour in Southern Africa”, we argued for a pluralist (or “dualist”) industrial and labour policy regime. Such a regime should accommodate and support firms to participate in 4IR, and to build the necessary smart factories. At the same time, it should accommodate and support the labour-intensive firms that compete in the domestic market against imports and — with appropriate policy support — could take advantage of rising wages in China and resume production for export markets. Only this dualist strategy can help begin to make a dent in unemployment. This requires accepting that we are not Wakanda, and that we need to work with the economy and the labour force we have. DM