Agricultural policy, productivity, farm size and land reform¶
Did state marketing control and support for white farmers help or harm agriculture, how should productivity be measured, and what land and farm-size policies suit South Africa?
Works engaging the debate: 30
Positions¶
- Marketing control costly and captured by producers (Vink, Kirsten, Van Zyl)
- Mixed developmental results subverted by populism
- Two agricultures as a policy construct (Lipton) and small-farm efficiency vs scale economies
- Market-led vs state-led land reform; causes of land degradation
Evidence base¶
The debate rests on a narrow quantitative core surrounded by commission reports and secondary synthesis. The core is the official agricultural statistical system: the Agricultural Census (from 1918, with the post-1949 series and the 1955, 1960, 1965 and 1969 rounds most used), the Abstract of Agricultural Statistics (from 1958), Union Statistics for Fifty Years and the Population Censuses of 1921-1970. The most systematic reworkings of these sources come from a handful of studies. Liebenberg (2012a) and Liebenberg (2012b) rebuild national output, input and public R&D series for 1910/11-2010/11 from commodity-level prices and quantities, ledger files of the Directorate of Agricultural Statistics, mechanisation sales surveys and Union/RSA Estimates of Expenditure, and compute Fisher and Törnqvist-Theil indexes. Kirsten et al. (2007) computes nominal and relative rates of assistance by comparing domestic prices with border reference prices, using SAGIS, FAOSTAT, IMF and customs data with an authors' spreadsheet running back to the 1950s. Simkins (1984a) and Simkins (1984b) interpolate six population census observations against Agricultural Census employment, production and wage data to reconstruct African population, labour supply and cash-and-kind income on white farms, 1923-1969. They anchor incomes on the 1937-39 Native Farm Labour Committee and deflate by the CPI to 1946 prices. De Klerk (1990) builds financial ratios for 1970-1988 from Department of Agricultural Economics and Marketing series and the 1983 SAAU farmer survey. Obi (2006) runs Engle-Granger cointegration on a 1955-2003 annual real land-price series. On degradation, (Hoffman and Todd (2000)) scores 348 magisterial districts from 34 extension-officer workshops in 1997-98 and regresses the scores on census and GIS variables. Wessels (2011) uses 1 km AVHRR NDVI for 1985-2003 and MODIS for 2000-2005 against rainfall surfaces.
A second layer consists of commodity-specific institutional records. Brits (1969) uses Mealie Industry Control Board annual reports and the Stabilisation Fund. The anonymous wheat thesis (Author not named in text (MSc candidate (2012)) uses Wheat Board annual reports for 1939-1996. Dubb (2012) uses South African Sugar Yearbooks and SACGA grower records. Two farm-level studies date from the interwar period: Neethling (1930) is a one-year (1927-28) survey of 109 Free State maize farms, and Lamont (1931) uses Union and world maize trade tables. Archival case studies supply the qualitative evidence: Schirmer (1999) draws on Central Archives Depot extension and Land Bank files and evidence to the Du Toit Commission, (Delius and Schirmer (2000)) on Native Affairs files and Soil Conservation Board reports, (Sato (2012)) on Weenen Bantu Affairs files and AFRA papers, and Beinart (1982) on Cape magistrates' and traders' records. Beinart (1994) adds interviews and Land Bank district figures. Much of the policy argument, however, comes from works that cite rather than analyse data. Examples are (Vink (2012)), which pairs 1930s SAJE predictions with the 1992 Kassier Committee, and Vink et al. (2012), (Lipton and Lipton (1994)), (Binswanger and Deininger (1993)), (Williams (1996)) and Crush and Jeeves (1993). The commission corpus most cited across the debate comprises the Native Economic Commission (UG 22/1932), the Social and Economic Planning Council Report No. 4, the Du Toit Commission (1959-60) and Kassier (1992).
Contested evidence¶
On marketing control, (Kirsten et al. (2007)) and Vink (2012) read a support peak of about 31% NRA in 1980-84, the perpetually indebted maize stabilisation fund (a bail-out of roughly R400 million) and the Kassier findings as proof of capture by large producers and cooperatives. Brits (1969) documents nine consecutive seasons of maize export losses from 1955/56, from Maize Board accounts. Commodity evidence cuts the other way, however. The Wheat Board series in Author not named in text (MSc candidate (2012) show domestic wheat prices exceeding import parity only after 1966, and only marginally until the mid-1970s, and depict a bread subsidy redistributing toward black consumers. Kirsten et al. (2007) itself concedes that its scheme-level distributional claims for maize, red meat and fruit rest on secondary literature rather than board records. Schirmer (1999) reads the same SEPC Report No. 4 productivity tables and Land Bank material as evidence of developmental intent subverted by populist price-raising, not of capture from the outset.
On productivity measurement the dispute is directly about indexes. Lipton and Lipton (1994) relies on the Thirtle, Sartorius von Bach and van Zyl TFP series for 1947-91 to argue that deregulation raised white-farm productivity. Williams (1996) recomputes the same series and attributes the apparent acceleration to a drought-year base. Liebenberg (2012b) shows that the spliced-Laspeyres Thirtle method doubles measured MFP growth (1.84% against 0.86% a year, 1945-2010). It also finds that official machinery formation figures were up to 52.5% too high in 1981 and fertiliser 7-44% too low.
On farm size the sides rest on incompatible evidence. Binswanger and Deininger (1993) and Lipton and Lipton (1994) argue small-farm efficiency from international cases, chiefly Kenyan settlement surveys, with no South African farm-level comparison. Against this, (Williams (1996)) notes that Moll's farm-level thesis does not support the World Bank's inefficiency claim. Schirmer (1999) uses Wickens and the 1951 and 1956 Agricultural Statistics to find large farms more productive in the 1950s. Neethling (1930) finds operator earnings rising with size in a single 1927-28 cross-section. The simulation models summarised in Vink and Kirsten (2001) predict negligible employment gains from redistribution.
Lipton's two-agricultures construct is challenged by Simkins (1984b), whose district map of per capita incomes shows parts of white rural areas giving Africans higher incomes than the reserves. Simkins's argument in turn depends on inferring de facto occupation from the gap between Agricultural and Population Census counts after 1955. Sato (2012) likewise finds the census undercounting Weenen's African population.
On degradation, (Delius and Schirmer (2000)) judges the Native Economic Commission's desertification predictions impressionistic, while Houghton and Dagut (1973) reprints Henderson's Victoria East comparison as evidence of decline. Hoffman and Todd (2000) finds poverty in susceptible terrain, rather than tenure, the predictor, but its index is perception-based. Wessels (2011) attributes degradation to apartheid population density without any pre-1985 or demographic data.
Underused and missing sources¶
The most consequential gap is farm- and district-level Agricultural Census data. Liebenberg (2012b) says the microdata are no longer accessible, so black and homeland output is trend-interpolated. Schirmer (1999), (Beinart (1994)), (Stavrou (1988)) and Binswanger and Deininger (1993) all lack the farm-size breakdowns needed to test scale economies. The Abstract land price index (Table 102, 2000 edition) offers a check on the land-price series in Obi (2006).
Control board accounts and minutes for the Maize, Meat and other boards are used only by Brits (1969) and the wheat thesis, although (Vink (2012)) and Kirsten et al. (2007) make distributional claims about them. Also missing are the Marketing Act texts and National Agricultural Marketing Council reports, Union Statistics for Fifty Years, the Statistical Year Book, and Hansard. Beyond these lie transaction-level Deeds Registry records by buyer and land use, needed for the land-market claims of Obi (2006) and Beinart (1994). So do Land Bank account-level arrears data and the 1968 labour-tenant registers flagged by Sato (2012). Pre-1985 aerial photography and erosion surveys would test the degradation claims of Delius and Schirmer (2000) and Wessels (2011), and constituency election returns would test the bread-and-1948 argument.
Strength of key claims¶
| Basis | Share of key claims |
|---|---|
| Systematic data | 14% |
| Partial data | 60% |
| Illustrative or anecdotal | 17% |
| Asserted | 9% |
Comparative figures for all debates are on the Literature overview.
Best-evidenced works¶
Liebenberg (2012b), Kirsten et al. (2007), Simkins (1984b), De Klerk (1990), Hoffman and Todd (2000), Author not named in text (MSc candidate (2012)
Most-used primary sources¶
Holdings are matched by machine against the archive's file manifest; a dash means no match was found, not that the project certainly lacks the source.
Works and their positions¶
| Work | Position |
|---|---|
| Author not named in text (MSc candidate (2012) | Rejects the view that white farmers were simply favoured at consumers' expense. Domestic wheat prices were not far above import prices until the mid-1970s, and the main long-run beneficiaries were the grain co-operatives, millers and bakers. |
| Author not named in text (MSc candidate (2012) | Sides with the sceptics (ANC 1992, Kassier, Lipton) who warned of monopoly. Deregulation transferred control from the state to a few agribusinesses, and the losers were small farmers and working-class consumers. |
| Beinart (1982) | Rejects the view that overstocking reflects a 'cattle cult'. The reluctance to sell in the 1920s-30s is traced to East Coast fever regulations, low prices, loans and bridewealth under mass migrancy. |
| Beinart (1994) | Argues that the land market, suitably nudged by Land Bank controls, taxation changes and state support, could provide scope for redistribution. He warns that the costs of reform need careful consideration and that capital for new farmers is the critical constraint. |
| Beinart (1994) | Land accumulation is tied to extensive stock farming, and the biggest owners are not the most intensive producers. Medium-sized farms show more diversity and intensity, so taxing large holdings or limiting farm size would be unlikely to harm overall production. |
| Binswanger and Deininger (1993) | Holds that economies of scale are rare beyond what a family with a medium tractor can manage, and that large wage-labour farms persist only because of state support. Rejects the efficiency case for the commercial sector. |
| Binswanger and Deininger (1993) | Favours market-assisted reform with grants or vouchers to the poor, willing sellers, removal of subsidies, and a judicial route for black-spot restitution. The pure land market alone cannot redistribute land. |
| Brits (1969) | Critical of the one-channel scheme: prices are set too high, consumers subsidise producers, and the scheme is rigid. He favours more flexibility, such as a floor-price/surplus-removal scheme like the one used for kaffircorn. |
| Brits (1969) | Argues that small farmers inflate the average cost of production and that the price should not sustain them. He wants small producers to leave farming and large farms to produce at lower unit cost. |
| Brits (1969) | Holds that producers dominate the Board and Cabinet and that no political party can ignore organised agriculture, so the Minister rarely overrides it. |
| Crush and Jeeves (1993) | Reports de Klerk's view that the crisis has structural and cyclical parts and is hastening a long overdue restructuring as the state withdraws patronage. |
| Crush and Jeeves (1993) | There is general assent to redistribution but disagreement on means. The editors are cautious about uncritical enthusiasm for the peasant or smallholder option (Lipton) and warn about World Bank generic solutions. |
| De Klerk (1990) | Argues drought is only the short-run cause. Interest rates and above all the terms of trade (47% of the debt increase 1980-85, citing the Economic Advisory Council) are more structural. |
| De Klerk (1990) | Holds that state policy to keep white farmers on the land, not market logic, sustained over-investment. Net protection benefited agriculture. Industrial capital has not moved into farm operation, so 'the age of monopoly capitalism in South African agriculture is not on the doorstep'. |
| De Klerk (1990) | Predicts that market forces are creating a core of black commercial farmers (sugar outgrowers, farm managers, covert tenancy). He speculates the state will manage this without repealing the Land Acts. |
| Delius and Schirmer (2000) | Rejects the argument (Jones and Muller) that this stemmed mainly from the declining influence of farmers within the United Party. Stresses instead the psychological and social impact of the Second World War and the reconstruction ideal (following Marwick), while granting that farmers' influence remained important. |
| Delius and Schirmer (2000) | Argues that authors such as Jones and Muller mistakenly assume opposition. The South African Agricultural Union and most farmers' representatives supported conservation, seeing it as a route to more state assistance. |
| Delius and Schirmer (2000) | Both extremes failed. Coercion in black areas bred resistance and ignored local adaptation, while accommodation of white farmers' interests rewarded inefficiency. The authors favour a balance between the two. |
| Delius and Schirmer (2000) | Sides with critics of the Native Economic Commission's diagnosis. The authors stress inadequate land, the centrality of migrant labour, and thin scientific evidence of imminent catastrophe. |
| Dubb (2012) | Rejects it as 'profoundly misleading'. FAF functioned as a miller-controlled forward-purchase system embedded in the regulatory structure and KwaZulu state support. |
| Dubb (2012) | Drought is proximate. The decline stems from the deregulation of registration and subsidies, which made the earlier growth a bubble. Industry blame of grower imprudence or fraud is called 'specious'. |
| Dubb (2012) | Treats plot size as mainly a question of land redistribution. It stresses declining terms of exchange and dependence on off-farm income over tenure. |
| Dubb (2012) | Differentiation is unstable and reversible and depends on non-farm income. Only contractor-growers show expanded reproduction. The paper argues for differential support and regulatory intervention rather than institutional fixes alone. |
| Hoffman and Todd (2000) | Rejects tenure as a cause. Argues that communal tenure per se is not responsible, and that poverty, unemployment, dependency and steep, hot, erodible environments better explain degradation. |
| Hoffman and Todd (2000) | Agrees with Roux that the eastern Karoo margin is no longer the main focus. Rejects the claim that veld degradation is rising everywhere, noting that it has been reversed in some commercial areas through state intervention and education. |
| Hoffman and Todd (2000) | Argues that past work was local or sectoral and neglected homeland areas, producing incomplete maps and discriminatory solutions. |
| Houghton and Dagut (1973) | Includes Viljoen and McLoughlin defending the Marketing Act, against Richards and the UCT economists' protest. |
| Kirsten et al. (2007) | Finds apartheid-era controls costly, fiscally burdensome and redistributive towards large producers. Argues that some 1950s-60s Afrikaner-farmer support programmes deserve re-examination for black empowerment, in WTO-compatible 'green box' form. |
| Lamont (1931) | Opposes compulsory control, price fixing, export subsidy and stabilization as unsound and likely to stimulate overproduction. Accepts voluntary co-operatives as legitimate and beneficial. |
| Lamont (1931) | Argues that making Union consumers cover export losses would act as a tax on all maize users and would restrict domestic use, including for livestock feeding. |
| Lamont (1931) | Argues for adjusting maize output to the farm's other enterprises, namely cattle, sheep, pigs and poultry, rather than expanding the one-crop system. |
| Liebenberg (2012a) | Holds that exclusion of black and homeland farmers distorted official commodity statistics, and that discriminatory policy suppressed black farmers' contribution, helping explain low measured 1947-68 productivity growth. |
| Liebenberg (2012b) | Holds that support was biased towards white commercial farmers (Land Act 1913, Co-operatives Act 1920, credit and settlement exclusion) and that suppressed black output helps explain low measured productivity growth 1947-68. |
| Liebenberg (2012b) | Suggests post-war productivity gains partly reflect pre-war investments and settlement schemes; R&D investment stagnated from 1971/78 and productivity effects lag by decades. |
| Lipton and Lipton (1994) | Argues pessimists confuse cause with effect. Discrimination removed the incentives, part-time and diversified farming is normal worldwide, and recent easing of restrictions has already raised Black output. |
| Lipton and Lipton (1994) | Favours a reformed labour-intensive agriculture as a lower-cost source of livelihoods, complementary to urban growth and not an inferior 'hewers of wood' option. |
| Lipton and Lipton (1994) | Favours market-oriented deregulation of marketing boards and prices, combined with active state support services and land measures such as a progressive land tax and land ceilings. |
| Neethling (1930) | Finds a direct relationship between size and operator's earnings, and a strong positive link between the maize share of receipts and earnings. Concludes that extensive farming must be on a large scale and that cattle on a commercial scale did not pay in 1927-28. |
| Obi (2006) | Sceptical: the proxy results do not support government concern about foreign buyers, and the land market is argued to be segmented, though the data are weak and the conclusion tentative. |
| Obi (2006) | The results give no basis to conclude that prices hurt reform, but price spikes create uncertainty. The thesis favours state land release, debt-relief purchases and leasing, and stresses post-settlement support. |
| Obi (2006) | Follows Van Schalkwyk, Van Zyl and Bayley: support to white farmers was unsustainable, and deregulation began in the mid-1980s, before 1994. Chow tests (1985, 1970/80/92 breaks) are read as confirming this. |
| Pankhurst (1996) | Namibia is an exception: conservation discourse is barely present before the 1980s and did not produce state intervention in African production. |
| Sato (2012) | Sato extends the critique of 'community' (Walker, James) by showing that farm tenants differed from black-spot communities: their livelihoods were more firmly rural, evictions were driven largely by farmers, and they justified land demands by local residence and rural livelihoods. |
| Sato (2012) | Sympathetic to the AFRA/WECOP position that development should reinforce tenants' existing survival strategies on the land, against the TBDA camp-upgrading proposal and McIntosh's urbanisation proposal. |
| Schirmer (1999) | Rejects the either/or framing. Regulation had mixed results and some positive effects, but it under-delivered because populism and short-term politics subverted developmental aims. Critiques the uncritical anti-regulation stance of Wilson, Nattrass and Jeeves & Crush while remaining critical of what the state actually did. |
| Schirmer (1999) | Applies this framework to South African agriculture and finds that developmental intent was repeatedly diluted by populism and by farmer representation on the control boards. |
| Schirmer (1999) | Finds that large white farmers were invariably more productive than small ones within regions. He reconciles this with the small-farm efficiency literature by arguing that small white farmers lacked the capacity to intensify or cooperate. |
| Simkins (1984a) | Qualifies it: transitional belts are extensive and income ranges narrow, so he proposes three agricultures shading into one another (progressive capitalist, backward capitalist, reserve). |
| Simkins (1984b) | Questions the sharp dual thesis: transitional belts are extensive and the interquartile range of incomes is narrow. Proposes three agricultures (progressive capitalist, backward capitalist, reserve) shading into one another. |
| Simkins (n.d.) | Argues that the Land Acts still do work that urban segregation no longer does. Abolition would trigger re-occupation of land and might alter agricultural structure if small farming proved viable. If they are retained, homeland tenure would move along capitalist lines, speeding out-migration. |
| Stavrou (1988) | Accepts Marcus's fusion of agrarian and urban capital in broad terms, but admits that company ownership of farm units is only about 10 percent and the process is 'in the infant stage'. |
| Vink (2012) | Sides with the critics (Frankel, Richards, Tinley, Groenewald, Kassier). Their predictions were largely borne out, and the defenders' claims (De Swardt, Viljoen) were not. |
| Vink and Kirsten (2001) | Qualified support for Lipton: land reform alone yields little or negative net employment (multiplier losses, risk in semi-arid areas), and works only with the 'four reforms'. |
| Vink and Kirsten (2001) | Favours deregulation and export orientation; judges that agricultural productivity growth has not been matched by demand expansion, so agriculture is unlikely to be a major job creator unless exports grow. |
| Vink et al. (2012) | Follows Lipton (1977): the 'two agricultures' were created by state policy supporting large white farmers and suppressing small farmers, not by market forces. |
| Vink et al. (2012) | Sides with DBSA economists who held that small farmers were 'poor but efficient' and that support services, not large projects, should be the vehicle. |
| Vink et al. (2012) | Attributes failure to the vacuum left by the demise of coordinated support, weak provincial and local institutions, and ignored lessons of the FSP, rather than to lack of plans. |
| Wessels (2011) | Attributes degradation mainly to high population densities aggravated by the oppressive apartheid system, not to traditional communal pastoralism. |
| Wessels (2011) | Argues degraded areas look functionally stable and resilient, so there is no apparent danger of the historically predicted collapse; economic loss is undetermined. |
| Wessels (2011) | Notes that an extensive 1990s review found no evidence for it. |
| Williams (1996) | 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. |
| Williams (1996) | 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. |
| Williams (1996) | 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. |