Industrial Decentralization - Reflections on the New Initiatives¶
C. L. McCarthy (1982). Industrial Decentralization - Reflections on the New Initiatives. South African Journal of Economics 50(3).
| Type | journal article |
| Scope | Economics and economic history of apartheid |
| Method | policy analysis |
| Period studied | 1960–1982 |
| Regions | South Africa, Homelands / black states, Border areas, PWV metropolitan area |
| Themes | manufacturing industrial policy, bantustans removals influx, business and state, politics ideology elections, demography urbanisation |
| Citation key | mccarthy1982 |
Argument¶
McCarthy asks whether the regional development policy that took effect on 1 April 1982 can do better than the traditional industrial decentralization policy, which failed to create jobs. He attributes the earlier failure to strong metropolitan agglomeration economies, too many growth points, relocation of large capital-intensive firms, small multipliers, and above all the separate-development ideology that defined regions ethnically. The new policy marks a real advance in demarcating functional regions and prioritising them, in emphasising small enterprises, and in planning to replace the section 3 'stick' with fiscal measures. He argues, however, that it still carries a strong separate-development bias, with incentives favouring points in the black states and a large number of designated development points. He doubts the social-cost case for decentralization and concludes that the policy is hard to achieve under the best of circumstances and harder still when it is tied to a political model.
Evidence and method¶
Official policy documents (Good Hope Plan, the 1982 Information Newsletter and White Paper), secondary literature (Van der Berg, Maasdorp, Palmer, Datta-Chaudhuri, Richardson, Stillwell, Van Eeden, Du Toit), and a few statistics on metropolitan concentration and job creation. There is no original data.
Place in the debates¶
- 'Work to the workers' (decentralization) versus 'workers to the work' (migration) as regional policy (see Homelands, separate development, decentralisation and removals). Sceptical of the social-cost case for decentralization, which he says is unproven. He allows that taking work to the workers has merit if overall employment and income growth is not sacrificed and locational economics is respected.
- Political ideology versus economic rationale in South African regional policy (see Homelands, separate development, decentralisation and removals). Argues that the separate-development ideology underlies or transcends the other obstacles. He holds that the 1982 reforms only partly escape it.
- Concentrated growth poles versus proliferating growth points (see Homelands, separate development, decentralisation and removals). Sees the proliferation of growth points, and the 49 development points in the new policy, as undermining the agglomeration advantages needed for self-sustained growth. He is sceptical that infrastructure can be politically concentrated on a few.
- Large-scale relocation versus small-scale appropriate industry (see Homelands, separate development, decentralisation and removals). Favours small-scale industry for its stronger regional linkages and lower investment per job. He doubts that the incentive package can really foster the informal sector.
Findings¶
-
Decentralization created far fewer jobs in the border areas and homelands than the labour market required. (p. 1)
150 000 jobs in total were created in the border areas and in the homelands during the period 1960-80; this compares unfavourably with the approximately 115 000 new annual entrants to the labour market in the black states.
-
Manufacturing was highly concentrated in four metropolitan areas, with the PWV region dominant. (p. 1)
In 1976 the four metropolitan areas, consisting of the Pretoria - Witwatersrand - Vaal triangle (PWV) area, the Western Cape, Durban-Pinetown and Port Elizabeth-Uitenhage, were responsible for 80 per cent of value added in manufacturing industry and for 76 per cent of manufacturing employment.
-
Incentives could not offset metropolitan agglomeration economies. (p. 2)
available incentives did not compensate for the agglomeration economies of the metropolitan areas, especially the PWV area.
-
The apparent successes were deconcentration of industry near the metropoles, not true decentralization. (p. 2)
in effect represent deconcentration, i.e. development outside the metropolitan area as such but near enough to provide a low-cost link-up with the markets, services and industries of the metropolitan areas.
-
Growth points proliferated, diluting resources. (p. 2)
in 1977 twenty designated growth points qualified for scheduled concessions.
-
Development corporations favoured large, probably capital-intensive, relocated industries over small industries. (p. 3)
a concentration on the decentralization of large-scale (and one may suspect capital-intensive) industries to the detriment of the development of more cost-effective small industries
-
The political ideology of separate development is the deepest constraint on the policy. (p. 3)
In the final analysis the political ideology of the policy of industrial decentralization either underlies or transcends the other factors that restrict its effectiveness.
-
The new policy designated many development points, contrary to expectations of a small number of growth poles. (p. 4)
the approval of 49 industrial development points (21 in "white" South Africa and 28 in the black states) came rather as a surprise to many
-
The new policy ranks regions by development need, giving the PWV the lowest priority. (p. 4)
region D, comprising the Eastern Cape, Ciskei and the southern part of Transkei, was given the highest priority, and region H, the PWV area, the lowest.
-
Incentives still tilt toward points in the black states, which shows a continued separate-development bias. (p. 4)
incentives vary within demarcated regions at different development points and consistently favour the developing points located in the black states.
-
The author challenges the claim that the PWV is overconcentrated. (p. 5)
even as far as the latter area is concerned the region may not, from an economic point of view, be overconcentrated.
-
Decentralization is intrinsically difficult, and the political model makes success less likely. (p. 7)
Because of the very strong influence of agglomeration economies, industrial decentralization is a difficult objective to achieve under the best of circumstances; encumbered with a specific political model as its rationale the chances of success diminish still further.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Good Hope Plan for Southern Africa (summary of proceedings of meeting between Prime Minister and business/community leaders, Cape Town, 12 November 1981) | commission or official report | 1981 | Source for the stated objectives of regional policy, the new regional approach, incentive differentiation favouring the black states, and the 'co-operation' concept. |
| The Promotion of Industrial Development: An Element of a Co-ordinated Regional Development Strategy for Southern Africa (Information Newsletter), 1982 | commission or official report | 1982 | Primary policy document used to describe the new strategy's growth points, incentives, small-enterprise emphasis and the admission that overconcentration applies only to the PWV area. |
| White Paper on the Information Document about the Promotion of Industrial Development as an Element of a Co-ordinated Regional Strategy for Southern Africa, 1982 | legislation gazette | 1982 | Quoted to show the multi-sectoral emphasis, selective concentration of infrastructure on a few points, and planned abolition of section 3. |
| Environmental Planning Act / Physical Planning Act, section 3 | legislation gazette | 1960s-1982 | Describes the 'stick' element of traditional decentralization policy (ministerial approval for new factories and extensions) and its proposed replacement. |
| Van der Berg, S., 'Job creation by the Development Corporations in the homelands' (forthcoming, Development Studies Southern Africa) | secondary literature compilation | 1960-80 | Supplies the estimate of 150,000 jobs created and the analysis of development corporation investment favouring large-scale industries. |
| Van Eeden, F. J., 'n Kritiese ontleding van die vordering met ontwikkelingsprogramme in die nasionale state' (Development Studies Southern Africa) | secondary literature compilation | 1975 | Source for the 44 percent purchasing power leakage out of the homelands in 1975. |
| Maasdorp, G., 'Industrial Decentralization: Ideal, Theory and Reality', IDS Bulletin 11(4) | secondary literature compilation | c.1980 | Used for the assessment that traditional policy produced deconcentration, and for the view that only two or three growth points are viable. |
| Tomlinson Commission estimate of annual industrial job requirement | commission or official report | 1950s | Benchmark of 20,000 industrial jobs per year against which achieved job creation is compared (cited via Van der Berg note). |
| Unattributed manufacturing statistics for 1976 (metropolitan share of value added and employment) | official statistical series | 1976 | Documents high industrial concentration (80% of value added, 76% of employment; PWV 50% and 45%). |
| Palmer, P. N., 'Industrializing the National States: The Bophuthatswana Example', Finance and Trade Review 14(3) | secondary literature compilation | c.1980 | Quoted evaluation that policy emphasized employment in ethnic regions over proximity to raw materials and markets. |
Inference¶
Quantification: Illustrative figures.
Conceptual and policy-document analysis: the author applies regional economics theory (agglomeration, private vs social costs, social overhead capital, absorptive capacity) to the official texts of the 1982 policy and argues by reasoning and by reference to others' evaluations of the earlier policy, with no new empirical testing.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Traditional decentralization policy failed to create adequate jobs: about 150,000 jobs in 1960-80 against about 115,000 new labour-market entrants annually in the black states. (p. 1) | Van der Berg's study (forthcoming), cited by the author; Tomlinson benchmark noted in footnote. | Partial data |
| Manufacturing is highly concentrated in the metropolitan areas, especially PWV, which justified decentralization policy. (p. 1) | Unattributed 1976 manufacturing figures for the four metropolitan areas. | Partial data |
| Traditional policy failed mainly because incentives could not offset agglomeration economies; the successful growth points (Brits, Babelegi, Hammersdale) represent deconcentration, not true decentralization. (p. 2) | Argument from agglomeration theory plus Maasdorp's assessment; examples named without data. | Illustrative or anecdotal |
| Small multipliers in the homelands result from purchasing-power leakage, linkages of decentralized industries to metropolitan areas, and low absorptive capacity. (p. 3) | Van Eeden's leakage figure, Datta-Chaudhuri's theory; the author's hypothesis. | Illustrative or anecdotal |
| The new strategy retains a strong separate-development bias, as incentives favour development points in the black states and the proliferation of 49 development points is politically motivated. (p. 4) | Reading of the Good Hope Plan, Information Newsletter and White Paper texts. | Partial data |
| The economic rationale of decentralization (high social costs of metropolitan concentration) has not been proven, and the high unit costs of backward-region infrastructure are ignored. (p. 5) | Author's own reasoning and Datta-Chaudhuri's theoretical argument; absence of research cited. | Asserted |
The author's caveats¶
The author notes that the incentives are not discussed or evaluated, and that the regional demarcations and priority ranking are not assessed. He states that he knows of no research proving the high social costs of metropolitan concentration. Several of his explanations are explicitly hypotheses.
Gaps in the evidence¶
The argument lacks quantitative evaluation of incentive effectiveness, of firm-level location decisions, and of metropolitan versus decentralized cost comparisons. Establishing the claims would require firm-level surveys, development corporation records and employment series by growth point, much of which existed through the Department of Statistics and the corporations' own reports, as well as the cost of infrastructure in the PWV. The 1976 concentration figures and the job estimates are unsourced or rest on an unpublished paper.
Replicability: Data poorly documented.