et al. 2010).
In spite of the depth of the unemployment problem in South Africa, many economists have sought to explain it as an abnormality rather than engage with the dynamics that (re)produce it. This started in the early 2000s, with Bhorat emphasising ‘the simultaneous existence of a skilled labour shortage and unskilled labour surplus’ (2004: 976) to argue that skills development would be the key to reducing unemployment. This argument was profoundly flawed for, as pointed out by Makgetla and Van Meelis (2003): ‘Even if more jobs were created for skilled than for unskilled people, it does not follow that increasing skill levels would in turn generate more jobs.’2 In the following years, drawing on dual labour market theories (which posit that there are two distinct labour markets, a formal and an informal), some economists have argued that ‘insider’ formal workers were forcing many ‘outsiders’ to either remain unemployed or to work informally (Kingdon and Knight 2007). Informal employment in South Africa is very low, however, especially by comparison with other African or middle-income countries, and much of its growth has been related to the informalisation of work rather than to new opportunities in the informal economy. Why can’t informal activities ‘soak up’ more of the excess labour in South Africa? Valodia (2013) suggests that:
Unlike most developing countries where small-scale, informal producers are able to capture a significant proportion of domestic consumption, the South African economy is dominated by large-scale, monopolistic producers with reach deep into the consumption basket of the South Africans of all income classes. Even in the most remote, rural and low-income communities, the basic consumption basket is dominated by goods produced in the formal economy, with very little – if any – capacity for local, informal producers to capture a sizeable proportion of local demand.
It is furthermore important to discard the notion that unemployment is high because informal wage employment is not captured by labour statistics; if anything, informal sector (especially self) employment is exaggerated by the Quarterly Labour Force Survey (Pons-Vignon, forthcoming).3 Following Pollin et al. (2006), our contention is that South Africa’s unemployment is a product of its structural economic features, with the three immediate reasons accounting for the very high unemployment rates being historically high unemployment, sluggish growth, and declining labour intensity of growth. Growth has overall been sluggish in South Africa since the late 1960s, with the exception of the period 2003 to 2007. Mohamed (2010) argues however that economic growth, during the latter period, far from being associated with long-term investment in the real economy, was linked to increased debt-driven consumption and speculation in financial and real estate markets and much of the growth in services employment was related to the outsourcing of low-pay services from manufacturing, in particular cleaning and security (Tregenna 2008). While the official unemployment rate had slowly declined during the 2000s, unemployment went up again after the economy recorded a million job losses in the wake of the global crisis. Last but not least, the continued dominance of sectors associated with the minerals-energy complex means that the most dynamic sectors are capital- rather than labour-intensive, with limited capacity to increase employment significantly even when they grow (Ashman et al. 2011). These structural features suggest that it will be necessary to change fundamentally the economic structure in order to tackle unemployment and poverty.
Largely ignoring these structural dynamics, debates about unemployment in South Africa have been dominated by calls for more labour market flexibility as the sustained unemployment crisis means that the national focus is primarily on job creation, often couched in ‘any job is better than no job’ terms. However, Bhorat and Cheadle (2007) have shown that the South African labour market was not rigid at all when compared to that of other countries:
Classified as an upper-middle income country, the comparisons across the regulatory indices are surprising. In the first instance, it is evident that South Africa’s measures of labour regulation compare quite favourably with those found in the rest of the world. In almost all of the individual regulatory sub-indices, South Africa yields a level of regulation that is lower than both the mean for upper-middle income countries, and for the sample of countries as a whole. For example, in the case of alternative employment contracts – the legislative regime governing part-time work, contractual employment and so on – South Africa yields an extremely low measure of labour regulation.
It is furthermore evident that if the roots of unemployment are structural, they cannot be reduced to a mere ‘frictional’ dimension related to a neoclassical understanding of the labour market as the place where supply and demand for labour meet. The focus on an imaginary ‘rigid’ labour market (and elusive ‘overpaid’ unskilled workers) is therefore little more than a diversion from a serious engagement with unemployment. It is all the more so that the South African labour market is in fact extremely flexible (and probably too flexible). Employers can do pretty much whatever they please in practice.
Casualisation: When no job is better than many jobs
The quality of jobs in South Africa has declined dramatically over the past twenty years. The implication is that working poverty, which was a structural feature of segregation and apartheid (Wolpe 1972), has all but disappeared. Many South Africans work long hours, but for miserable pay and in insecure, often hazardous conditions. Is any job really better than no job?
Non-standard forms of employment are increasingly common throughout the South African labour market, in line with the global restructuring of work which has led, through a great diversification of employment arrangements, to widespread precariousness. Contrary to what is often assumed, this is not restricted to the ‘margins’ of the labour market, but is increasingly a feature of its core (Chang 2009). In South Africa, restructuring started ahead of the transition to democracy and has since become a wideranging phenomenon in sectors as diverse as healthcare, mining and forestry (see Pons-Vignon and Anseeuw 2009; and Von Holdt and Webster 2005 for a broad range of case studies). In mining in 2008, one out of three workers was employed by a contractor or a sub-contractor (Bezuidenhout 2008), a figure which has probably increased in the wake of the 2012 violence across platinum and other mines. The forms taken by work-restructuring have been varied, and include the growth of third-party employers such as labour brokers and contractors, alongside a sharp rise in casualisation, documented in a vast range of sectoral case studies but poorly captured by labour force surveys (for a methodological discussion, see Sender and Pontara, 2010). Casualisation, for instance, entails work arrangements such as homework (Godfrey et al. 2005) or the hiring of gangs of workers by the day to perform certain tasks. With a few exceptions, for instance in transport (Barrett 2003) or metals and engineering, trade unions have not been able to counter employer strategies and prevent casualisation.
Labour casualisation has entailed a marked deterioration in levels of pay and security. In terms of pay, this is visible in the consistently low wages received by workers covered by sectoral determinations (see section 2), two-thirds of whom were classified as ‘poor’ in 2007 – with an increase in the number of poor workers in certain sectors since the adoption of a determination (