Setting the Agenda: A Critique of the World Bank's Rural Restructuring Programme for South Africa¶
Gavin Williams (1996). Setting the Agenda: A Critique of the World Bank's Rural Restructuring Programme for South Africa. Journal of Southern African Studies, Vol. 22, No. 1, Special Issue: State and Development. Taylor & Francis.
| Type | journal article |
| Scope | Wider apartheid literature |
| Method | policy analysis |
| Period studied | 1913–1995 |
| Regions | South Africa, Kenya, Zimbabwe, KwaZulu, KaNgwane, Transkei |
| Themes | agriculture land, transition post apartheid, business and state, bantustans removals influx, comparative theory, macro fiscal finance debt |
| Citation key | williams1996 |
Argument¶
Williams argues that the World Bank's 1993 'Options' land reform and rural restructuring proposals rest on misleading intellectual foundations. Its claims that large South African farms are inefficient and that small farms would be more efficient are not supported by much of the evidence it cites. The Bank's Kenyan model credits planners' intentions (Swynnerton) rather than the unintended outcomes, and its proposals revive the social engineering of the Tomlinson and Swynnerton reports. Its 'surprisingly small' cost estimates depend on indicative models with unrealistic assumptions on interest, administration costs and land prices, and proved unaffordable within the RDP. The models serve an ideological purpose, legitimating a plan that amounts to another vast resettlement project. He accepts the case for liberalising marketing and for land reform, but on grounds of restitution, poverty and insecurity rather than efficiency.
Evidence and method¶
Close reading of World Bank documents (Options, 'South African Agriculture', Executive Summary, Terms of Reference) and the LAPC and Royal Swazi Sun background papers. Also Moll's thesis, Thirtle et al.'s TFP estimates, Farmer Support Programme evaluation reports, Kenyan and South African historiography, Board of Tariffs and Trade and Kassier Commission reports, press reports and RDP/DLA documents.
Place in the debates¶
- Farm size, efficiency and scale economies (inverse farm-size/productivity relation) (see Agricultural policy, productivity, farm size and land reform). Rejects the Bank's claim that small farms would be more efficient than large ones. Cites Moll for scale economies in grain farming and attributes the Bank's cross-size comparisons to differing crop mixes. The case for land reform should rest on restitution and poverty instead.
- Causes of farm mechanization and labour shedding (Habakkuk-type/Junker-path account) (see Farm labour: mechanisation, labour shedding, wages and labour control). Criticises the Binswanger-Deininger-Feder schema as evolutionist. Argues that state interest-rate, tax and pricing policies in the 1970s encouraged combine harvesters and labour displacement, and that mechanization initially raised labour demand.
- Agricultural liberalisation and deregulation (see Agricultural policy, productivity, farm size and land reform). Accepts strong arguments for liberalising marketing, but doubts that removing subsidies will 'level the playing field' and warns that deregulation leaves concentrated agro-food firms intact. Questions the Bank's claims about cheap imports of staples.
- Kenya as a model for land reform (see Agricultural policy, productivity, farm size and land reform). Argues Kenya's agricultural success came from opening highland markets and White Highlands to smallholders, not from consolidation, registration and planned resettlement that the Bank credits.
- Productivity growth in South African agriculture (TFP) (see Data reliability and econometric method). Disputes the Thirtle et al. and World Bank reading of productivity growth after 1983, noting that the 'rapid growth' uses the 1983/84 drought nadir as base, so weather explains it more simply than policy.
Findings¶
-
The 30% land transfer figure in the RDP originated in the World Bank's Options, where it was only an illustrative example. (p. 2)
The figure originated in Options for Land Reform and Rural Restructuring in South Africa (hereafter Options) put forward by the World Bank.""""""""""""""""""""""""""""""""""
-
Mechanization first raised and only later reduced farm labour demand, contradicting the Bank's schema; combine harvesters were the key shedding technology. (p. 11)
mechanization initially increased the demand for labour in South Africa by increasing the ploughed area; only with the introduction of combine harvesters was the overall demand for labour, as well as the incidence of labour tenancy, reduced.""""""""""""""""""""""""""""""""""""""""""""""""
-
State interest-rate, tax and pricing policies in the 1970s encouraged combine harvester adoption and labour displacement. (p. 14)
In the 1970s state interest rate, taxation and pricing policies encouraged farmers to acquire combine harvesters and reduce employment of labour"""""""""""""""""""""""""""""""""""""""""""""
-
Farm indebtedness and 'inefficiency' derive from state policies shaped by farmers' lobbies, not intrinsic inefficiency; subsidies and relief continued into the 1990s. (p. 14)
State policies, shaped by farmers' organisations and agro-industrial interests, encouraged an unsustainable structure of farming. The source of the consequent 'inefficiencies' was in these policies.""""""""""""""""""""""""""""""""""""""
-
Farmers' share of consumer grain prices fell sharply in the 1980s, which the Bank's subsidy-based explanations cannot account for. (p. 15)
The producers' share of the consumer price for grains (and most other crops) fell consistently between 1980 and 1990 (from 42% to 25%)""""""""""""""""""""""""""""""""""""""""
-
The Bank's claim of 'surprisingly small' costs rests on models rather than evidence; Williams argues the programme's cost exceeded RDP resources. (p. 24)
The 'surprisingly small' cost proved to be surprisingly large when set against the resources available for the Reconstruction and Development Programme.""""""""""""""""""""""""""""""""""""""""
-
Under more realistic interest assumptions, beneficiary incomes in three of four models would fall markedly, so costs are likely higher than projected. (p. 25)
If more realistic assumptions are made about the actual annual interest payments farmers are expected to meet from their annual earnings (say 15%), farmers incomes would fall by between 9% and 29%""""""""""""""""""""""""""""""""""""""""""""""
-
Despite avoiding the term, the Bank's plan amounts to a large resettlement scheme of the Tomlinson type. (p. 28)
it is difficult to see how a programme to move 600,000 families over five years onto some 30% of South Africa's medium to high quality land could be anything but 'a vast resettlement project'."""""""""""""""""""""""""""""""""""""""""""""""
-
Farmer Support Programmes in the bantustans failed to produce emergent commercial farmers, unlike sugar outgrower schemes. (p. 8)
In two cases, farmers who did not participate in the FSP were more successful than those who did. Hardly any commercial farmers 'emerged' from the process."""""""""""""""""""""""""""""""""""""""""""""""
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| World Bank, 'Options for Land Reform and Rural Restructuring in South Africa' (LAPC, Johannesburg, October 1993) | commission or official report | 1993 | Primary object of critique: its proposals, indicative models, cost estimates and 30% redistribution example are examined. |
| World Bank, 'South African Agriculture: Structure, Performance and Options for the Future' (Washington, 1994; draft 5 August 1993) | commission or official report | 1993-1994 | Source of the Bank's claims on efficiency, scale, TFP, subsidies, marketing and prices that Williams assesses against the evidence cited. |
| World Bank, 'Executive Summary, South Africa Rural Restructuring Programme' and 'Terms of Reference' (mimeo, 1993) | commission or official report | 1993 | Used to show the programme's guiding aims, resettlement emphasis and administrative options. |
| Roth, Dolny and Wiebe, 'Employment, Efficiency and Land Markets in South Africa's Agricultural Sector' (Land Tenure Centre, Wisconsin, draft 1992) | commission or official report | 1992 | Identified as the main report of the Bank's agricultural sector initiative and origin of passages repeated by Christiansen. |
| Royal Swazi Sun (1992) papers: Binswanger and Deininger, Binswanger/Deininger/Feder, Kinsey and Binswanger, Duncan and Jones, Lipton, Christiansen (World Development 21, 9, 1993) | secondary literature compilation | 1992-1993 | The Bank's international-experience evidence on farm size, land policy and resettlement, critiqued for its model of agrarian change and use of comparative evidence. |
| Union of South Africa, Summary of the Report of the Commission for the Socio-economic Development of the Bantu Areas (Tomlinson Report, 1955) | commission or official report | 1955 | Comparator for the Bank's proposals: its 'economic farm units' and creation of a Bantu farming class are likened to the Bank's plan. |
| Swynnerton Report, A Plan to Intensify the Development of African Agriculture in Kenya (Nairobi, 1954) | commission or official report | 1954 | Used as the Kenyan parallel to Tomlinson and as model for the Bank's 'yeomen farmer' approach. |
| Development Bank of South Africa Farmer Support Programme evaluation reports (Kirsten, van Zyl, Sartorius von Bach; Fischer, Dederen, Ficq; Cross and Evans 1991) | commission or official report | c.1987-1991 | Evidence that Farmer Support Programmes in the bantustans largely failed to produce commercial farmers. |
| Thirtle, Sartorius von Bach and van Zyl, 'Explaining Total Factor Productivity Growth in South African Agriculture, 1947-91' (Reading Working Paper 1993; Development South Africa 10, 3) | secondary literature compilation | 1947-1991 | TFP estimates cited by the Bank; Williams recomputes growth rates and questions base-year choice (1983 drought). |
| P. Moll, DPhil thesis, Oxford 1988, 'The Structure of Commercial Agriculture in South Africa' | secondary literature compilation | 1980s | Used to show economies of scale exist in large-farm sector, contra the Bank's reading. |
| Van Zyl/Schalkwyk land price studies (LAPC 1993; 'The South African Land Market: an Analysis of Land Prices') | secondary literature compilation | 1970-1990 | Model and time series of real land prices used to discuss land acquisition costs. |
| Board of Tariffs and Trade, 'An Investigation into the Price Mechanism in the Food Chain', Report 3273 (Pretoria, 1992) | commission or official report | 1992 | Source of Bank argument that food marketing margins reflect wage inflation; Williams finds it unconvincing. |
| Republic of South Africa, Report of the Committee of Enquiry into the Marketing Act (Kassier Commission, 1992) | commission or official report | 1992 | Outline of statutory marketing schemes and criticism of monopolies such as Unifruco. |
| Agricultural Marketing Policy Evaluation Committee (Basson Committee), Reports 1 and 2 (1994) | commission or official report | 1994 | Shows the committee's partial acceptance of tariffs over import controls while retaining marketing boards. |
| Price parity studies: Jones/Duncan/Beynon LAPC Policy Paper 2; van Zyl 'Agricultural Pricing'; van Zyl and Niebuhr, de Kock and Laubscher (Agrekon) | secondary literature compilation | 1985-1991 | Import/export parity price estimates used to contest the claim that domestic grain prices exceed world prices. |
| Van Heerden and van Zyl, 'Aggregate Measure of Support for Maize in South Africa, 1985/86 to 1990/91' (mimeo, 1992) | secondary literature compilation | 1985-1991 | Producer subsidy equivalents for maize showing subsidies were eliminated by 1990/91. |
| M. Rimmer, 'Debt Relief and the South African Drought Relief Programme', LAPC Policy Paper (1993) | secondary literature compilation | 1981-1991 | Supports figures on state subsidies and debt/drought relief to farmers. |
| Legislation: 1913 Natives Land Act, 1936 Native Trust and Land Act, 1991 Abolition of Racially Based Land Measures Act, Act 126 of 1993, Constitution Act 200 of 1993, Restitution of Land Rights Act 1994 | legislation gazette | 1913-1994 | Used to describe the legal framework of land segregation, abolition, property protection and restitution that conditions the Bank's proposals. |
| LAPC background papers and Proceedings, Land Redistribution Options Conference (1993/94) | secondary literature compilation | 1993-1994 | Papers (Lund and Wakelin, de Wet, Adams et al., Njobe, Klug, Scott, Nieuwoudt) used to show the Bank selectively drew on and diverged from its own commissioned research. |
| World Bank Options Annexes 1-4 and de Klerk models (assumptions, imputed rentals, interest rates, administration costs) | commission or official report | 1993 | Williams re-examines model assumptions (interest rates, administrative costs, R3.5bn per annum) to show costs are understated. |
| ANC, Reconstruction and Development Programme (1994); RDP White Paper; DLA 'Draft Land Policy Principles' (1995); Land Reform Pilot Programme overview (1994) | commission or official report | 1994-1995 | Shows adoption of the 30% target and later scaling back to R15,000 grants; RDP budget used to gauge affordability. |
| Press sources: Business Day, Farmer's Weekly, Weekly Mail and Guardian, Perspective: The Media on Agriculture (1993-1995) | newspapers periodicals | 1993-1995 | Used for statements by Bank officials and for later developments in maize prices, land claims and disputes. |
Inference¶
Quantification: Illustrative figures.
Documentary critique of World Bank texts: Williams reads the Bank's reports against the sources they cite and comparative historical literature (Kenya, Zimbabwe, earlier South African commissions), re-checks selected statistics (TFP growth rates, interest-rate assumptions in the cost models), and reasons about the plausibility of the models.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| The Bank's claims about the inefficiency of large South African farms and the efficiency of small farms are not supported by much of the evidence it cites (Moll, Sartorius von Bach and van Zyl). (p. 149) | Comparison of the Bank's reports with Moll's thesis and other cited studies, with the Bank's own admission that comparison is 'difficult if not impossible'. | Partial data |
| The TFP evidence for poor farm performance is flawed: growth rates were miscalculated and the 'rapid growth' depends on a drought-year base, so weather explains it more simply than policy. (p. 150) | Thirtle et al. index, re-computed by Williams from mean annual increases over periods. | Partial data |
| Farm indebtedness and 'inefficiency' resulted from state policies and their withdrawal in the 1980s, not inherent farmer inefficiency. (p. 151) | World Bank 1994 report data, Vink, de Klerk, Rimmer on subsidies, debt and drought relief. | Partial data |
| The Bank's 'surprisingly small' cost claim rests on indicative models, not evidence, and its assumptions (low interest rates, modest administrative costs) understate costs; it proved well beyond the RDP's resources. (p. 161) | Options text and Annexes, de Klerk model, RDP budget figures. | Partial data |
| Using a more realistic 15% interest rate, farmers' incomes would fall by between 9% and 29% in three of the four Bank models. (p. 162) | Williams's own recalculation from model assumptions in Options Annex 1. | Partial data |
| Liberalisation and import-parity arguments are overstated: domestic grain prices are not uniformly above world prices, world prices are distorted, and white maize cannot be reliably imported. (p. 155) | Agrekon price-parity estimates, Jones and van Zyl papers, Bank's own acknowledgement on thin white maize markets. | Partial data |
| The Bank's proposals revive the Tomlinson/Swynnerton social-engineering approach, and its Kenya model misrepresents what produced Kenyan success. (p. 157) | Comparison of Tomlinson, Swynnerton, Binswanger and Deininger, and Kenyan historiography (Heyer, Cowen, Leo). | Illustrative or anecdotal |
| Farmer Support Programmes in the bantustans largely failed to create commercial farmers, whereas sugar outgrower schemes succeeded. (p. 145) | DBSA evaluation reports and Vaughan and McIntosh studies. | Illustrative or anecdotal |
The author's caveats¶
Williams notes that the published articles on Farmer Support Programmes tell little about outcomes and that the Bank's own comparisons of small and large farms in South Africa are inconclusive. He notes the cost models use statistical means that 'do not apply to a specific place', that no one knows the real cost, and that he is not offering a counter-model of arbitrary assumptions.
Gaps in the evidence¶
The critique draws almost entirely on secondary and grey literature and the Bank's own documents, with no direct engagement with farm-level or census data (e.g. agricultural censuses, farm survey microdata) to test scale-efficiency or productivity claims. Farm-level cost and price data, and land price and transaction records, would better test the cost models; some existed in the early 1990s but are not used here. Evidence on actual redistribution outcomes and costs became available after 1995.
Replicability: Data poorly documented.