South Africa's Border Industries: The Tswana Example¶
Alan C. G. Best (1971). South Africa's Border Industries: The Tswana Example. Annals of the Association of American Geographers, Vol. 61, No. 2. Taylor & Francis for the Association of American Geographers.
| Type | journal article |
| Scope | Economics and economic history of apartheid |
| Method | mixed |
| Period studied | 1955–1969 |
| Regions | Tswana homeland (Bophuthatswana), western Transvaal, northern Cape, Rosslyn, Brits, Rustenburg, Zeerust, Mafeking, Pretoria region |
| Themes | manufacturing industrial policy, bantustans removals influx, labour markets wages, business and state, politics ideology elections, demography urbanisation |
| Citation key | best1971 |
Argument¶
Best assesses the border industries scheme, which gave incentives to industry in white towns adjoining the homelands, using the Tswana homeland and its five border towns as a case. He argues the scheme has produced only minor industrial decentralization and little homeland development, and that white investors, white towns and white services have benefited more than the Bantu areas. Industries rely on white capital, technology, raw materials and urban markets, so linkages and multiplier effects are weak. The scheme is politically rather than economically motivated; it is rational only within a commitment to territorial segregation. He reviews alternatives (Bantu entrepreneurs, government corporations, white investment on an 'agency basis'), but concludes that real regional growth cannot be expected while apartheid persists.
Evidence and method¶
Fieldwork in June-August 1969, including interviews with officials and employers and field questionnaires, and unpublished files of government departments (Bantu Administration and Development, Bantu Investment Corporation, Industrial Development Corporation, Tswana Territorial Authority) and municipal offices. Also published government reports (Tomlinson Commission, Viljoen Commission, Permanent Committee for the Location of Industry annual report) and secondary economic literature. Tables cover state assistance, employment, and business establishments in the new townships. The author notes discrepancies between the data sets of different offices.
Place in the debates¶
- Economic viability and purpose of industrial decentralization / border industries (see Homelands, separate development, decentralisation and removals). Sceptical: the scheme yields marginal decentralization and negligible homeland development, is politically motivated, and the new centres are mere appendages of the metropolitan areas, in line with critics such as Nieuwenhuysen and Van der Horst.
- Tomlinson Commission vs government policy on industrializing inside the homelands (see Homelands, separate development, decentralisation and removals). Sees the government's rejection of Tomlinson's homeland-based industry as a missed opportunity and the later 'agency basis' as a belated partial return to its recommendations.
- Separate development vs economic integration and rationality of apartheid (see Homelands, separate development, decentralisation and removals). The scheme is rational only if total territorial segregation is the goal; the apartheid structure itself blocks real regional growth.
Findings¶
-
State assistance to border industries 1960-68 was large, mostly in loans, share capital and factory buildings. (p. 5)
State assistance to industries and local authorities (1960-1968) totaled Rands 79,735,000 (Table 1).
-
A firm moving to Rustenburg under the scheme received cheaper black labour as an explicit concession, at about half the Roodepoort wage. (p. 9)
reduced Bantu wage rates, the latter being approximately half those in Roodepoort.
-
Rosslyn, the model border town, has no local linkages, limiting the multiplier. (p. 8)
No industries are geared exclusively to the local Tswana market, thus minimizing the multiplier effect the border industries scheme theoretically possesses.
-
Border industries mainly benefit whites and are tied to white capital, inputs and markets. (p. 12)
Whites have prospered through Bantu wage control and transport subsidy; and Whites subscribing to separate development feel less threatened by the mere presence of the Bantu
-
The scheme produced no homeland manufacturing and minimal spillover; the main flows between white and Bantu areas are labour, consumer goods and wages. (p. 13)
No manufacturing has materialized within the homelands as a direct response to border industries.
-
Border industries and satellite townships perpetuate migrant labour while halting migrants nearer their source areas and weakening chiefly authority. (p. 13)
The new townships and border industries are perpetuating the migrant labor system by drawing enterprising workers away from the interior of the homelands.
-
The first decade's results were negligible for homeland development and marginal for decentralization. (p. 13)
The first ten years of the border industry scheme have met with negligible success if rapid and diversified development of the Bantu homelands was the primary objective.
-
Bantu Investment Corporation lending was overwhelmingly to trade, not manufacturing, showing the weak industrial base in the homelands. (p. 14)
Trading, services, and manufacturing accounted for eighty-five, thirteen, and two percent of the loans, respectively.
-
The conclusion ties the limits of the scheme to apartheid itself. (p. 16)
real regional growth cannot be expected as long as apartheid persists. Separate development means a duplication of infrastructure, the non-maximization of Bantu skills and resources, and White supremacy at the expense of the Bantu.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Tomlinson Commission, Summary of the Report of the Commission for the Socio-Economic Development of the Bantu Areas Within the Union of South Africa (U.G. 61, 1955) | commission or official report | 1955 | Sets out the recommended homeland industrialization alternative that the government rejected in favour of border industries. |
| Government Decisions on the Recommendations of the Tomlinson Commission, White Paper F-1956 | commission or official report | 1956 | Documents the government's rejection of internal industrialization and adoption of industry at White border towns. |
| P. S. Rautenbach, 'Industrial Decentralization in the Republic of South Africa', Tegnikon Special Issue (1967) | secondary literature compilation | 1960s | Supplies the figures on concentration of industrial output and establishments in four metropolitan regions. |
| H. F. Verwoerd, Statement on the Establishment of Industries in Border Areas (mimeographed, Industrial Development Corporation files, 1960) | commission or official report | 1960 | Source of the official definition of 'border areas' and the first major policy statement. |
| Viljoen Commission, Report of the Commission of Enquiry into Policy Relating to the Protection of Industries (U.G. 36, 1958) | commission or official report | 1958 | Cited for its conclusion that decentralization fits government policy but economic scales were weighted against it. |
| Permanent Committee for the Location of Industry, Report of Activities 1 January 1968 to 31 December 1968 (S. 40166, 1969) | official statistical series | 1960-1968 | Source for state assistance totals (Table 1), border-area establishments, employment, investment and Bantu job-creation figures. |
| The South African Financial Gazette, 24 January 1969, p. 8 (and 11 October 1968) | newspapers periodicals | 1968-1969 | Source for the estimate of IDC annual investment in border areas and the decentralized share of metropolitan investment. |
| Unpublished and published materials from government departments and corporations (Dept. of Bantu Administration and Development, Dept. of Planning, Tswana Territorial Authority, Bantu Investment Corporation, IDC, Bantu Affairs Commissioners) | state archives | 1960s-1969 | Supplies the economic and demographic data on the Tswana homeland and border towns. |
| Unpublished data in files of the Bantu Investment Corporation, Pretoria, and field investigation (Table 2) | company records or reports | 1969 | Establishes the scarcity of non-agricultural employment and establishments in the Tswana homeland. |
| Unpublished BIC and Tswana Territorial Authority reports and field observations (Table 3: business establishments in five new Tswana towns) | company records or reports | August 1969 | Documents the small tertiary businesses, population and housing in the satellite Bantu townships. |
| Field questionnaires and unpublished municipal office data, Brits, Mafeking, Rustenburg (Table 4) | author survey or fieldwork | August 1969 | Gives employment percentages by industry and race, used to show industry is minor except in Brits. |
| Personal interviews with officials and managers (Du Plessis, Wolvaardt, Pieterse, Kloppers), July-August 1969 | interviews oral history | July-August 1969 | Supports statements on government plans, Brits farm labor, Rustenburg industrial siting and Mafeking's Tswana capital. |
| Bantu Investment Corporation of South Africa, Annual Report and Accounts for the Year Ended 31st March 1968 | company records or reports | 1960-1968 | Source for BIC loan numbers and amounts and their sectoral distribution. |
| Unpublished reports in the files of the B.I.C., Pretoria (agency-basis investment terms) | company records or reports | c. 1968-1969 | Source for conditions on White investors under agency-basis industry. |
| Senate Hansard, No. 4, 1964, Column 839 | parliamentary debates | 1964 | Evidence of the Minister of Labour's statement that no skills ceiling applies to Bantu in border areas. |
| Legislation: Bantu (Urban Areas) Consolidation Act 1945, Group Areas Act, Promotion of Self-Government Act 1959, BIC Act 1959, Homelands Development Corporation Act 1965, Promotion of Economic Development of Bantu Homelands Act 1968, Physical Planning and Utilization of Resources Act 1967, Industrial Conciliation Act 1956 | legislation gazette | 1945-1968 | Describes the legal framework of influx control, homeland development and industrial location. |
| Press items: The Pretoria News 9 July 1969; News From South Africa 4 March 1970; Bantu vol. 17 (April 1970) | newspapers periodicals | 1969-1970 | Supports claims on the Rosslyn brewery, Mabopane resettlement plans and Temba as agency-industry pilot. |
| J. P. Nieuwenhuysen, 'The Industrialization of the Tugela Basin', South African Journal of Economics 31 (1963) | secondary literature compilation | 1963 | Provides the backward/forward linkage classification applied to industries in the study area. |
| D. Page, The Urbanization of the Bantu Homelands of the Transvaal (dissertation 1964; book 1966) | secondary literature compilation | 1964-1966 | Cited in support of the feasibility of homeland industrialization and urbanization. |
New data¶
Author's 1969 fieldwork (June-August) in the Tswana homeland and border towns: questionnaires on employment by industry and race in Brits, Mafeking and Rustenburg, interviews, and compiled unpublished department and BIC data on establishments, townships and population; sample size not stated.
Inference¶
Quantification: Descriptive tables.
Single regional case study (Tswana homeland and five border towns) using descriptive tables, official aggregates on state assistance, and local fieldwork, combined with qualitative linkage reasoning and comparison with Tomlinson recommendations; no statistical testing.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Border industries have produced only minor decentralization; Bantu job creation (about 5,000 per year) falls short of the 9,000 needed annually. (p. 332-333) | Permanent Committee for the Location of Industry 1968 report figures on border-area employment and the required job creation. | Partial data |
| Industry is unimportant in the Tswana border towns except Brits; Rosslyn is mere industrial extension of the metropolis, not decentralization. (p. 339) | Table 4 field questionnaires and municipal data, plus town-by-town descriptions from interviews and official files. | Partial data |
| Benefits of border industries accrue more to Whites than to the Tswana, through investment, services, wage control and transport subsidy, with minimal spillover into the homeland. (p. 339) | Qualitative inference from case descriptions and Nieuwenhuysen's linkage categories; no quantified distribution of benefits. | Illustrative or anecdotal |
| The Tswana homeland has very little non-agricultural employment; mining jobs mostly go to migrants from outside the region. (p. 333-334) | Table 2 from BIC files and field investigation; description of Rustenburg mines. | Partial data |
| Border industries have not relieved Witwatersrand/Pretoria congestion; resettlement into satellite townships is local movement, though it perpetuates migrant labour and weakens chiefs. (p. 340) | Author's observation of township origins, with no migration data (author admits the extent is hard to assess). | Illustrative or anecdotal |
| State assistance to border industries totalled R79.7 million in 1960-68, mostly loans, share capital and factory buildings, and about 135 assisted new industries located in border areas. (p. 332) | Table 1 drawn from Permanent Committee for the Location of Industry report. | Systematic data |
| BIC support to Bantu enterprise is concentrated in trading and in satellite townships rather than manufacturing inside the homeland. (p. 341) | BIC Annual Report 1968 loan data and townships' business counts (Table 3). | Partial data |
| Border industries are politically rather than economically motivated, and agency-basis industry differs from border industry only semantically. (p. 342-343) | Policy documents (Tomlinson, White Paper 1956, Viljoen) and author interpretation of legislation and concessions. | Asserted |
The author's caveats¶
Notes considerable discrepancies in economic and demographic data between Bantu Affairs Commissioners, the Department of Bantu Administration and Development, and BIC field and head offices, attributed to tardy reporting, boundary changes and redefinition of authority. Economic organization of the Tswana homeland is treated in general terms because of data restrictions. Table 4 industry and quarrying are combined, which is misleading; redistribution of people is hard to assess.
Gaps in the evidence¶
The argument lacks a counterfactual or before-after firm-level dataset of border-area establishments and employment, so effects of the scheme cannot be separated from general growth. Systematic data on migration origins, wages, and the distribution of subsidies between firms and workers would test the claim that Whites benefit more; such records existed in IDC and departmental files, and many are now in archives. Linkage claims rest on one external classification rather than input-output or trade-flow data, and the Tswana homeland income data are thin.
Replicability: Data poorly documented.