The changing spatial structure of economic growth in South Africa over fifty years¶
McCarthy, Jeff (2000). The changing spatial structure of economic growth in South Africa over fifty years.
| Type | working paper |
| Scope | Wider apartheid literature |
| Method | descriptive quantitative |
| Period studied | 1910–1994 |
| Regions | South Africa, Witwatersrand/Johannesburg, Durban, Cape Town, Port Elizabeth, bantustans/platteland |
| Themes | bantustans removals influx, demography urbanisation, manufacturing industrial policy, methods data sources, transition post apartheid, inequality poverty unemployment |
| Citation key | mccarthy2000a |
Argument¶
McCarthy argues that the spatial structure of South African economic growth has changed only 'glacially' since the early twentieth century. Mining-led growth on the Witwatersrand created a hub-and-spoke economy with Johannesburg at the centre and coastal cities on the rim. Using Gross Geographic Product (GGP) by magisterial district for 1968-1994, he shows that apartheid-era decentralisation modestly raised non-metropolitan growth shares in the 1960s-70s, after which growth reconcentrated in metros and on the coast. Past growth patterns are the best predictor of the 1990s pattern (regression R² of 61-80%), and he builds a 'most likely' scenario in which growth sticks to established metropolitan cores and corridors. He concludes that policy such as the Spatial Development Initiatives (SDIs) should follow the existing spatial pattern.
Evidence and method¶
Statistics SA/HSRC GGP data by magisterial district for 1968, 1972, 1981, 1991 and 1993/4 (1996 data not yet available); Union Statistics for Fifty Years (1960); Board, Davies & Fair's 1960 space-economy study; regression and rank-order 'acceleration' analyses; a gravity model; GIS mapping of scenarios. Figures 1 and 2 are not electronically available.
Place in the debates¶
- Effectiveness of apartheid-era decentralisation and growth-point policy in reshaping the space economy (see Homelands, separate development, decentralisation and removals). Sceptical: patterns are path-dependent and policy-induced deviations are small. The non-metro share of growth was higher in the 1960s-70s (89%, 95%) than in the 1980s-90s (84%, 85%), but the underlying hub-and-rim structure persisted.
- Metropolitan vs non-metropolitan drivers of spatial restructuring (urbanisation literature) (see Urban segregation, Group Areas and the apartheid city). Tests this and finds metros always accelerated faster, though the non-metro share of growth varied with apartheid policy.
- Policy-led vs path-dependent spatial growth (see Urban segregation, Group Areas and the apartheid city). Path dependence dominates, citing international experience such as China. The paper endorses 'sticking to the beaten track' with SDIs and rejects the previous government's growth-centre strategies.
Findings¶
-
Mining on the Witwatersrand was the main force shaping the spatial economy from 1910 to 1960 and overturned the earlier coastal bias. (p. 3)
The mining of the complex of minerals on the Witwatersrand, and its associated stimulatory impacts upon related economic activities such as manufacturing, was the single most important driving force
-
Agriculture and mining made up about half of GDP in 1911, against about 20% by 2000. (p. 2)
by 1911 approximately 50% of the country’s G.D.P. was comprised by just two sectors – agriculture and mining
-
Johannesburg overtook Cape Town in population between 1911 and 1960. (p. 3)
in 1911 Johannesburg – centre of the mining industry – had a population smaller than Cape Town’s, whereas by 1960 it had a population which was larger
-
Raw nominal GGP growth 1968-94 was largest in the core, with coast and mid-zone lower (the mid-zone covers about 70% of districts). (p. 5)
Coastal districts R 84 Billion Core districts R114 Billion Mid-zone districts R 88 Billion
-
The mid-zone (platteland and bantustans) led acceleration in the 1960s but came last thereafter, while the coast rose from last to first. (p. 5-6)
The mid-zone (including both the platteland and bantustans) did best in the 1960’s, and came last thereafter.
-
Non-metro growth as a share of metro growth was highest when apartheid decentralisation policies were most pronounced, then fell as they collapsed. (p. 7)
during the 1960’s and 1970’s, when the application of apartheid and decentralisation policies was most pronounced, the non-metropolitan areas did somewhat better than in the 1980’s and 1990’s
-
Past spatial growth strongly predicts 1990s growth, whereas population growth is a weak predictor. (p. 7)
1980’s growth with 1990’s growth = 80% of variance explained
-
Manufacturing, construction, trade, transport and finance were the best sectoral predictors of 1990s spatial growth, and mining and agriculture the weakest, which the author reads as secondary and tertiary activity leading growth. (p. 8)
The weakest predictors of the overall spatial distribution of growth are mining and quarrying, and agriculture, forestry and fishing.
-
Growth is highly concentrated: over 90% of top-performing districts lie within 120 km of four metros. (p. 14)
over 90% of those districts performing in the top league fall within a 120 km radius of the centre of just four metropolitan places
-
58% of 1990s growth occurred in a few core metropolitan districts covering under 4% of the country's area. (p. 15)
58% of new growth during the 1990s was realised within a few magisterial districts (core metropolitan in nature) which comprised less than 4% of the country’s area.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Gross Geographic Product (G.G.P.) data by magisterial district, 1968, 1972, 1981, 1991 and 1993/4 (Statistics SA / HSRC) | official statistical series | 1968, 1972, 1981, 1991, 1993/4 | Assembled into a file of G.G.P. by magisterial district and used to compute growth changes (G.G.P.C. 1-4), regional and metro/non-metro comparisons, regressions and projected scenarios. |
| Statistics SA 1995: Gross Geographic Product 1994 | official statistical series | 1993/4 | Used for the inter-sectoral correlation of sub-sector growth rates to identify 'predictor' sectors for the 1990s. |
| 1996 G.G.P. data (Statistics SA/HSRC), not yet available | official statistical series | 1996 | Not used; the author notes it is pending and expected not to alter conclusions materially. |
| Union Statistics for Fifty Years (South Africa, 1960), Bureau of Census and Statistics | official statistical series | 1910-1960 | Source for city population comparisons (Johannesburg vs Cape Town; Durban vs Pietermaritzburg) showing Witwatersrand-led shifts. |
| 1960 population census (as used by Board, Davies & Fair) | population census | 1960 | Cited as the primary data underlying the most systematic 1960 study of the space economy and the 1960 benchmark picture. |
| Board, Davies & Fair 1970, 'The structure of the South African space economy' (GDP per square mile, 1960) | secondary literature compilation | 1960 | Supplies the 1960 baseline map of GDP per square mile (Figure 1) establishing the core-coast 'spokes on a wheel' structure. |
| Abedian and Standish 1992, Economic Growth in South Africa (sectoral GDP shares 1911) | secondary literature compilation | 1911 | Used for the claim that agriculture and mining made up about half of GDP in 1911. |
| DBSA 1999 map of Spatial Development Initiatives (SDIs) | maps spatial data | 1999 | Figure 2 used to compare projected growth areas with SDI corridors and as a policy factor in scenarios. |
| Ramos 1999, address 'GEAR Targets' (Director General of Finance) | other | 1999 | Cited for the idea of 'incidence analysis' showing varying rates of return to public expenditure. |
| Urban Foundation 1990, Urban Debate 2010 | other | 1990 | Cited as an example of literature assuming metro/non-metro differences drive spatial restructuring. |
New data¶
The author assembled a file of G.G.P. by magisterial district for 1968, 1972, 1981, 1991 and 1993/4, classified districts as coastal/core/mid-zone and metro/non-metro, computed growth changes by period, and generated a GIS map of projected growth per magisterial district.
Inference¶
Quantification: Statistical analysis.
Descriptive comparison of G.G.P. growth across district groupings (coast/core/mid-zone; metro/non-metro) over four periods, with rank-ordered 'acceleration' measures, simple regressions of 1990s growth on past growth, population growth and sectoral growth, then extrapolation to 'most likely' and 'less likely' scenarios by applying the 80% regression-derived persistence share to past shares, supported by qualitative arguments on path dependence and global forces.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| The 1960 space economy was dominated by the Witwatersrand core with a corridor to Durban and secondary port nodes, a 'spokes on a wheel' structure. (p. 3) | Board, Davies & Fair 1970 GDP per square mile map from the 1960 census, plus Fair 1983 and Union Statistics for Fifty Years population figures. | Partial data |
| Coastal, core and mid-zone districts differed in growth acceleration over time: the mid-zone led in the 1960s/70s but ranked last thereafter, while the coast moved from last to first. (p. 5) | Author's calculations of G.G.P. changes by magisterial district for 1968-1994, rank-ordered (Table 2). | Systematic data |
| Non-metropolitan areas did relatively better in the 1960s-70s, when apartheid decentralisation was strongest, than in the 1980s-90s; growth momentum shifted to metros. (p. 7) | G.G.P. by magisterial district with a strict metro definition; non-metro growth as a percentage of metro growth (Table 4: 89%, 95%, 84%, 85%). | Partial data |
| Spatial variance of 1990s growth is best predicted by past spatial growth patterns (61-80% variance explained), not population growth (<10-20%). (p. 7) | Regression analyses on G.G.P. growth by magisterial district across decades; results reported without coefficients, sample details or specification. | Partial data |
| Manufacturing, construction, trade, transport and finance are the strongest 'predictor' sectors of 1990s spatial growth, while mining and agriculture are weakest. (p. 8) | Correlation of sub-sector growth rates across magisterial districts in 1993/4 (Statistics SA 1995). | Partial data |
| About 80% of future growth will occur where it did in the recent past, with a slight shift towards coast and metros; projected shares are given in the Table 5 scenario. (p. 9) | Extrapolation from regression R² values and the 1990s baseline, with author-chosen adjustments and assumptions about SDIs and global forces. | Asserted |
| Growth is extremely concentrated: 58% of 1990s growth occurred in core metro districts covering under 4% of area, and 24% in the Johannesburg-Pretoria channel (about 0.2% of area). (p. 15) | Author's calculations from the G.G.P. by magisterial district file. | Partial data |
The author's caveats¶
G.G.P. figures are acknowledged to be imperfect measures of economic activity, especially the informal sector. The 1996 G.G.P. data were unavailable. Raw totals are unadjusted for inflation, and the coast/core/metro boundaries are admitted to be strict or arbitrary.
Gaps in the evidence¶
The paper gives no regression specifications, sample sizes, definitions of the district groupings or the full G.G.P. series, so claims on acceleration and predictability cannot be checked. Real (deflated) district-level series, constant district boundaries across the 1968-1994 period and sector breakdowns would better test the persistence claim, and these exist in Statistics SA and HSRC data. The forward scenarios rest on assumed persistence shares and policy effects without an independent model, and the 1960-1968 gap and the informal sector are not covered.
Replicability: Data poorly documented.