Skip to content

Putting the economy in its place: Geographical economics in South Africa

Literature › Works

Krugell, Waldo F. (2011). Putting the economy in its place: Geographical economics in South Africa. Scientific Contributions, Series H: Inaugural Address Nr. 242. North-West University (Potchefstroom Campus).
Type report
Scope Post-1994 South African economy
Method synthesis review
Period studied 1990–2010
Regions South Africa, Gauteng (Johannesburg, East Rand, Pretoria), Cape Town, Durban, Port Elizabeth, 354 magisterial districts
Themes demography urbanisation, inequality poverty unemployment, transition post apartheid, labour markets wages, bantustans removals influx, methods data sources, comparative theory
Citation key krugell2011

Argument

The address summarises the author's earlier co-authored studies on how local economic growth unfolded after South Africa's 1994 re-entry into the world economy. It argues that apartheid planning added a layer of spatial inequality on top of first-nature geography (coast, gold, Johannesburg). Convergence tests (beta, sigma, kernel density, Markov chain) find at best slow convergence and, on current dynamics, persistent or widening gaps between districts. Cities, above all the Gauteng agglomeration, act as engines of growth through thick labour markets and firm-level agglomeration economies. The policy conclusion is that unbalanced growth can still be inclusive if investment goes into human capital, infrastructure and urban planning.

Evidence and method

Regional Economic Explorer (REX) database compiled by Global Insight (sub-national data, mainly 1996-2004, for 354 magisterial districts); PIMSS data for 1990-2000; census data; a national survey of day labourers; the 2007 World Bank Enterprise Surveys for firms in the four major cities. It mostly summarises the author's earlier papers (Naudé and Krugell 2003, 2006; Bosker and Krugell 2008; Krugell and Rankin 2009; Blaauw and Krugell 2010).

Findings

  • Economic output is highly concentrated spatially: a handful of places produce most of national output. (p. 8)

    In 2010 the top 15 out of 245 places produced approximately 70 per cent of output

  • Local growth since 1996 was very uneven across regions. (p. 8)

    Thus over the period 1996 to 2010 only 67 regions grew at rates faster than 3 per cent per annum.

  • Apartheid added racial spatial inequality on top of natural geography. (p. 8)

    In effect it added a further layer of inequality on top of the spatial inequality due to the country’s first-nature geography.

  • Apartheid spatial planning imposed efficiency costs on the urban economy. (p. 8)

    resulted in a spatial economy characterised by inefficient land use, excessive transport costs, and under-investment in transport infrastructure, telecommunications and electric power.

  • Sigma convergence across districts fell in 1990-96 but dispersion rose slightly after 1996, which the author links to globalisation. (p. 12)

    The standard deviation of log of real per capita income for all cities and towns declined from 0.6147 in 1990 to 0.5258 in 1996, after which it increased again slightly to 0.5466 in 2000.

  • Kernel density estimates show no convergence of district incomes between 1996 and 2004. (p. 13)

    In 2004 this probability increased slightly to 51 per cent, imply no evidence of convergence.

  • Markov chain analysis projects a heavily diverging steady-state distribution of district incomes. (p. 14)

    98 per cent of the regions earn less than 0.36 times the national level of GDP per capita and only 2 per cent of the regions more than that

  • Movement toward the steady state is slow and mobility is low, so the pattern is persistent. (p. 15)

    while the movement towards the steady state takes quite some time (a half-life of 58 years)

  • There is a large rural-urban income gap. (p. 15)

    the total average annual income in rural areas in 2000 was R 18,506 compared to R 51,107 urban areas

  • The rank-size coefficient for South African cities is below Zipf's value of 1, which the author reads as cities that are too small. (p. 18)

    estimated q, and found this to be = 0.75 in the South African case.

  • Day labour is widespread and concentrated in a large number of places, mostly involving black African men. (p. 19)

    There are nearly 1,000 places in South Africa where a minimum of 45,000 mostly black African men stand and wait to be picked up for day labour

  • Firm-level data give tentative support to agglomeration economies in Gauteng. (p. 22)

    tentative firm-level evidence of agglomeration economies in the land-locked Gauteng province that contains Johannesburg, the East Rand and Pretoria.