Previous Page Table of Contents Next Page

South Africa

Agribusiness and its Place in the South African Economy

Until early in 1998 the marketing of most agricultural products in South Africa was extensively regulated by statute. One of the main characteristics of the control that was exercised was isolation from world market forces. Most products were regulated under the 22 marketing schemes introduced from 1931 and especially from the time of the 1937 Marketing Act, although some products, including sugar, wine and ostriches, were regulated by those industry’s own institutions under separate legislation.

Beginning two decades ago, the industry faced increasing pressures for deregulation, a process that was accomplished in two phases over this period. The major change in the first phase was the extensive deregulation of state agricultural marketing schemes within the framework of the Marketing Act of 1968. The origins of this change can be found in the shift in monetary policy in the late 1970s and fiscal strategies in the 1980s, which undermined the complex structure of protection, price support and cross-subsidies on which the system of agricultural support was founded. Yet isolation from the world market, accompanied by the increased isolation of the country in social, cultural, political and intellectual spheres during the 1980s, meant that the deregulation steps that did take place were aimed at the domestic market (Vink and Kassier 1991, Francis and Williams, 1991; Vink, 1993; Kirsten and van Zyl, 1996; Vink, 2000).

Further, the market in Africa was not considered to hold any real potential, partly because it was regarded as too small, but mostly because of the anti-apartheid measures that were in place. Foreign trade still largely consisted of managing imports and exports in order to manipulate domestic prices (e.g. maize, wheat), or of monopoly export schemes (e.g. for fruit). The first real steps in opening the agricultural sector to world market influences came with the Marrakech Agreement of the GATT in 1993, when all direct controls over agricultural imports were replaced by tariffs.

The government of national unity, elected in 1994, ushered in new policies across the entire range of state activities. In agriculture, some tended to follow the direction of changes already under way (Hall and Williams, 2000). Major direct policy changes had to wait until after the National Party, and its Minister of Agriculture, Kraai van Niekerk, withdrew from the government in 1996. New policy initiatives included the land reform programme; laws protecting agricultural workers and labour tenants against eviction and extending their rights; liberalization of international trade and agricultural marketing; the Marketing of Agricultural Products Act of 1996; a new rural development policy; and institutional restructuring in the public sector. South African agriculture has now taken its place as a fully integrated member of the global trade environment, and its trade patterns have shifted considerably over the past decade.

The field crop sector has been most affected by the process of deregulation. Field crop production has declined from almost half of total output in the three year period centred around 1978/79 to less than a third in the three-year period centred on 1999. In this time animal production increased its share somewhat, while horticultural production increased its share by 10 percentage points.

Table 2: The value of commercial agricultural production in South Africa (1999)[2]

Gross value of production

Number of farms

Number of hectares


(R per ha)

(t per ha)


5 484 468

3 868 000

10 613 000




437 720

16 000

29 700

27 358



144 905

51 000

167 692

2 841


Soya beans

196 179

94 000

152 600

2 087


Field crops

13 666 600

11 992

9 528 309

1 434



11 714 500

8 039

3 898 486

3 005


Cattle slaughtered

3 983 215

495 000


Animal production

18 938 700

31 442

63 384 734




44 319 800

60 938

82 209 571


Sales in the South African manufacturing sector grew by some 2.5 percent per annum in real terms in the period 1996-2001, a rate close to the overall real rate of growth of the economy. By contrast, sales of the food and beverages industries grew by about half that rate, making it one of the worst performers in this sector. Recent sales growth in this subsector has been third highest among the components of the manufacturing sector. Production in the food and beverages group accounted for about 18.5 percent of total manufacturing output for the country in 1996, while employment was 15.9 percent of total manufacturing sector employment and the wage bill 13.5 percent of total manufacturing sector wages. A more detailed breakdown of the subsector is provided in Table 3, while the degree of concentration in 1996 is reflected in Table 4, the most recent available data.

Table 3: The South African food and beverage sector

Sales (Rm)


Exports (R’000)

Imports (R’000)


78 079

225 527

6 205 634

5 524 284


80 131

219 155

6 752 412

6 291 720


83 886

221 426

8 286 938

6 625 716


83 607

209 686

8 247 898

7 471 358


81 896

201 594

9 061 613

6 989 492


81 759

203 211

9 122 024

6 468 007


79 757

187 882

10 270 184

6 556 806


84 689

184 187

12 225 957

6 742 894

The South African agro-food complex, which consists of primary production plus the input and agro-processing sectors, accounts for around 14 percent of the GDP. In 2000 the agro food complex exported about R16 billion worth of primary and processed food products, or nearly 10 percent of South Africa’s total exports. Almost all the productive and social activities of rural towns and service centres are dependent on primary agriculture and related activities. This includes increasingly popular and economically significant agro-tourism and game farming activities. Taking all of this into account it is true that more than half of the provinces and about 40 percent of the country’s total population are primarily dependent on agriculture and its related industries.

Linking Arrangements between Poor Farmers and Agribusiness - Case Studies

Due to its historical legacy South African agriculture is characterized by extreme dualism. There are around 50 000 large commercial farmers who are predominantly, but not exclusively, drawn from the white population. Commercial farms employ about 1 million workers, which is 11 percent of total formal sector employment in the country. Many of these workers live on commercial farms and their children receive education in farm schools. These commercial farms compare very well with the large farm enterprises in the US where packing and processing is sometimes done on the farm. Many farmers have recently moved to contractual arrangements with larger processing and retailing companies to secure markets and less volatile prices.

On the other hand retailers and agribusiness nowadays prefer such arrangements in order to ensure reliable supplies and less fluctuation in prices (as we discussed in the introduction). This trend has been very apparent following the process of abolishing the controlled marketing arrangements in the country in the post 1994 period. The linkages that have emerged range from formal delivery contracts to production under strict instructions to informal marketing arrangements.

Apart from the large-scale commercial farmers, there are also 240 000 small farmers (mainly black) that provide a livelihood to more than 1 million of their family members and occasional employment to another 500 000 people. These farmers supply local and regional markets where large numbers of informal traders make a living. Furthermore, an estimated 3 million households who are primary located in the communal areas of the former homelands, produce largely to meet part of their family’s total needs.

With the change in the political dispensation there has been a need to commercialise the agricultural activities of these households. Some earlier efforts by parastatal development corporations and some agribusinesses are commendable but the challenge of black empowerment in agriculture is so huge that much more needs to be done. Improving on-farm productivity for increased sales could be one way of stimulating commercial activity and thereby linking them to markets. However our experience with development efforts over the years has clearly shown that this approach is not sufficient because access to markets (and finance) seems to be more important for economic success.

Poor developed links with markets (and thus with agribusiness per definition) have reduced incentives in agriculture to such an extent that farmers in many cases have abandoned farming activities. This has been a major problem not only amongst farmers of perishable commodities such as dairy, fruits and vegetables but also amongst grains, oilseeds and beef. The lack of market access is often attributed to poor infrastructure and communication. But sometimes it is just poor quality or quite often lack of trust that creates the perception that these farmers’ products do not comply with the basic minimum requirements in order for it to be sold.

It therefore becomes quite important for agribusiness in South Africa to develop stronger links with disadvantaged farming communities to ensure that true economic empowerment materialises. It is argued that some special actions from the business community are needed to tackle this major challenge in South African agriculture. Maybe the lessons emerging from this paper can make a contribution in this regard.

Over the years, agribusiness have invested in some initiatives to ensure a much stronger engagement of farmers in disadvantaged communities in the production of especially industrial crops such as sugar, tea, cotton and timber. Lately there have also been a number of initiatives by agribusinesses and some facilitators to provide more and better market opportunities for disadvantaged farmers in the production of tomatoes, other vegetables, wine, fruit, grains, wool and livestock. These are all new initiatives and as a result it is not possible to document them. For that reason the following case studies are presented to illustrate the nature of the links of small-scale farmers with agribusiness:

In addition a very novel approach or concept to link small-scale grain and oil seed producers to the market is discussed. These producers are poor, do not have collateral and are therefore excluded from producing grains and oilseeds on a commercial scale. By linking the credit market, input market, output market and the futures market this is now made possible.

The analysis of the case studies is approached from a strict New Institutional Economics perspective. This approach allows to unpack the nature of the arrangement in a much more structured framework and also enables the debate the appropriateness of the current arrangement given the characteristics of the product, transactions and the partners.

Transvaal Sugar Company

The South African sugar industry is regulated by the Sugar Act (1978), which grants statutory powers of self-government to this sector of the agricultural economy. The affairs of the sugar industry are controlled by the South African Sugar Association (SASA) who administers the production and supply of sugar cane to the millers and also the production, marketing and distribution of sugar. The principal sugar growing areas of South Africa primarily include Kwazulu-Natal, Mpumulanga and the Eastern Cape where the climatic conditions are more favourable for the growing of sugarcane.

The South African sugar industry produces an average of 2.5 million tonnes of sugar in 2000/1 of which 50 percent was exported to markets in Africa, the Middle East, North America and Asia. These exports contributed R 1.9 billion to the country’s foreign exchange earnings. The sugarcane industry directly employs approximately 85 000 people and provides employment for an additional 265 000 in industries that are linked to this sector of the economy. The number of growers and area under sugar cane are illustrated in Table 7. South African cane growers produce over 23 million tonnes of sugar cane annually that are processed in 15 different milling areas. Large growers (3 percent) are responsible for over 70 percent of sugarcane produced whilst small growers (97 percent) account for less than 30 percent.

Transvaal Sugar Limited (TSB), which employs around 4 000 people, was founded in 1965 and operates in the province of Mpumalanga with offices in Johannesburg and Durban. TSB is a 100 percent owned subsidiary of Hunt Leuchars and Hepburn (HL & H) which, in turn, forms part of the Rembrandt Group of companies. The company has the capacity to produce 350,000tonnesof sugar annually from its two factories and sugar production has increased from 109 500 tonnes in 1975/76 to the current level of approximately 300 000 tonnes in 2000/1. Assets employed have increased from R 80.5 million in 1986 to R 586 million in 1995 with an estimated current replacement value of R 2.3 billion.

The sugarcane supply-processing operation of the Transvaal Sugar Company consists of the factory processing operation and a range of growers. The factory processing operation consists of two sugar mills. The first mill is located at Malelane and the second mill is located south of Komatipoort both in Mpumalanga. The growers include the company estates and a range of contracted large-medium and small-scale suppliers.

The differences between the types of growers can largely be categorised on the basis of the differential farm size and the level of capital investment. In the case of the company estates, the milling company farm large tracts of land. This operation is categorised by a modern capital-intensive mono-cropped sugarcane production system with high levels of management input and control.

The second category of grower, namely, contracted medium-large growers, are also characterised by a modern capital-intensive mono-cropped sugar cane production system with high levels of management inputs. These farmers are contracted to the Transvaal Sugar Company by way of a long-term specification contract and supply in excess of 64 percent of the total volume of sugarcane delivered to the two mills. Most of these farmers operate farms that are in excess of fifty hectares and in many cases, sugar cane is one of the farm enterprises together with sub-tropical fruit and vegetables. These farmers are largely autonomous and the growing and delivery of sugarcane is self managed with ad hoc inputs from the agricultural division and the factory cane supply division who co-ordinates the timing of the harvesting and delivery operations.

The third category of grower, namely the contracted small-scale growers, includes in excess of 1 000 small-scale farmers belonging to thirty-two different supplier groups. The average farm size of these growers is 6.8 hectares with the smallest farm recorded as around 3.7 hectares and the largest 30 hectares.

Linking arrangements

The supplier contract. The contractual arrangement between out-growers and the Transvaal Sugar Company (TSB) is controlled by a cane delivery agreement. All growers must adhere to the conditions and obligations that are specified in a comprehensive specification contract that binds the respective parties over long periods of time. The price paid to out-growers is determined by the specifications of the South African Sugar Association who determine the grower-miller split from the proceeds of sugar sales.

Transaction characteristics. The processing of large volumes of a perishable commodity like sugarcane requires the co-ordination of the activities of the growers with the optimum use of high fixed cost processing facilities that are unique to the sugar industry. The volume of supply, the nature of the suppliers and the industry specific nature of the processing plant and equipment, in turn, influences the dynamics of the firm's transactions. In order to identify and evaluate the transaction characteristics of the supply chain a number of assumptions were made. The annual number of sugarcane deliveries to the mill was employed to determine the transaction characteristic of frequency whilst the value and co-ordination requirements of company fixed assets were used as a basis to determine asset specificity. The transaction characteristic of uncertainty has been evaluated on the basis of analysing the conditions of supply.

Transaction frequency. The TSB milling operation processes around 19 000 tonnes of sugarcane per day or an annual volume of around 3.6mtonnesof sugarcane that are transported by way of 136 000 deliveries. The continuous nature of the processing operation, combined with the large volumes of sugarcane processed and the perishable nature of the raw commodity, do not allow the processor to stockpile sugarcane. On the basis of this evidence, this study has concluded that the transaction characteristic of frequency should be classified as high.

Asset specificity. The mill group currently employs over R500 million of fixed assets at gross book value. These assets have a current value in excess of R2 billion and are highly specific with low opportunity cost outside the sugar industry. The assets, moreover, are relatively immovable and are also site specific as they have been centralised in relation to the company estates and out-grower suppliers. Conversely, the assets of the contracted growers consist largely of irrigation and general farming equipment that can be applied relatively more easily outside the sugar industry. TSB management is required to manage and co-ordinate the harvesting and delivery of sugarcane from the company estates, over 140 medium-large scale commercial farmers and, in excess of 1 000 small-scale growers. On the basis of the value of sugar specific plant and equipment, combined with the high level of co-ordination required to synchronise the sugarcane supply-processing operation, it is quite evident that high levels of asset specificity exist in the supply chain. The milling company has therefore some vested interest in ensuring good relationships with the growers due to the fact that many (especially the larger growers) can easily switch to higher value enterprises such as mangoes and citrus if the incentives in sugar are not appropriate.

Uncertainty of supply. The uncertainty of sugarcane supply is a function of the reliability of the growers combined with the uncertainties of agronomic and climatic variables. Because the supply of sugarcane from the company estates is internalised in the company structure, low levels of uncertainty attach to this category of supplier. The reliability of supply from contracted growers, however, is often a function of alternate opportunities and sugar prices that are struck in volatile markets. The level of uncertainty attached to this category of grower is, therefore, higher than the company estates. The high fixed cost nature of the processing operation, moreover, also promotes uncertainty because of impact this has on the volatility of cash flows (Drury, 1996, Horngren, 1999). On the basis of this evidence a low to moderate level of uncertainty is assumed.

The governance structure for the sugar supply chain. The process of matching a set of transaction characteristics with the optimal governance structure i.e. how the supply chain and the links with growers are managed is a central tenet of transaction cost theory. Organizations that have transaction characteristics that reflect high levels of asset specificity, combined with the need to carefully co-ordinate a complex input-output function, require higher levels of managed co-ordination than the open market mechanisms (Williamson, 1981; 1996; 2000; Petersen & Wysocki, 1997; 1998).

In the case of sugar production the transaction characteristics suggest an optimum structure somewhere between specification contracting, using outgrowers with contracts, and full vertical co-ordination. The actual organization structure of sugarcane supply in the case of TSB, however, displays an aggregated structure that is closer to specification contracting than full integration because 82 percent is acquired by means of grower contracts.

The transaction cost of supply. The objective of this section is to demonstrate that the transaction cost of small-scale growers exceeds that of medium and large-scale growers in the TSB cane supply operations. The differential cost of the smallholder supply operation, in general, is illustrated by the annual budget of R3.2 million of the small farm department of the company. The TSB grower transactions include set-up transactions, cultivation transactions, harvesting and delivery transactions and administration transactions.

Firstly, smallholders incur higher levels of start-up cost. Start-up transaction cost is incurred by all contracted growers with respect to the initiation of the contract. However, only small-scale farmer projects receive high levels of company assistance with respect to their interaction with local authorities, assistance with financing, training, land preparation and the installation and maintenance of irrigation equipment. Medium and large growers, largely, self establish their legality and ability to be engaged as contracted suppliers.

Secondly, smallholders incur higher levels of transaction cost in all phases of the growing operation. Small-scale growers, again, require higher levels of company inputs from the agricultural and small farm commercial departments with respect to the planting, weeding, fertilising, irrigation and ripening of cane. Small-scale growers, moreover, require higher levels of company inputs to guide and co-ordinate harvesting and delivery transactions. Conversely, larger farmers interact on an occasional basis with TSB with respect to growing technologies. Although larger farmers also receive regular visits from the TSB cane supply department, transaction cost for the company when dealing with the smaller growers is much higher.

Finally, smallholders also incur higher levels of administration cost per ton delivered. The administration of suppliers’ accounts is managed by way of a creditors system that generates a weekly payment, in cheque form, for the weekly tonnes delivered to the mill. The creditors system also adjusts this payment if the supplier has drawn items from stores. Large suppliers, for instance, only generated 76 accounting transactions against a delivery volume of 47 840 tonnes for the 2000/01 season. Conversely, medium size suppliers delivered 10,623 tonnes against 138 accounting transactions and small-scale suppliers 978 tonnes against 26 accounting transactions.

Grower performance. The grower performance, analysed between 1998 and 2001, indicates that small-holder production efficiency matched that of the company estates in this period. The cost structure of TSB and the four sets of selected farmers, although comparable from a total cost perspective, is significantly different in terms of operating and overhead cost. TSB displays a total overhead cost of R27 per tonne whilst small-scale farm overhead costs have been estimated at R13.3 per tonne. The small-scale growers appeared to contract more efficiently for a range of field and transport services than the company estates. Conversely TSB displays total operating costs of R86 per tonne compared to small-scale grower costs of between R96 and R100. The company estates appear to have lower cultivating costs than the small-scale farmers indicating, somewhat paradoxically, that the estate productivity exceeds that of the family farms. This could be partially explained by the fact that the production of sugarcane is not particularly labour intensive and that small-farmers may need to incur incremental cost for the use of mechanised inputs.

However the small-scale grower programme has had tremendous impacts in the Nkomazi valley. Since the mill at Komatipoort was erected and the irrigation infrastructure established the physical and socio-economic landscape transformed within two seasons. The link of smallholders with the Sugar Company has improved livelihoods and alleviated poverty in many households in the community. The multiplier effects are also visible as more small enterprises got established. Thus a large benefit for the community has materialised out of small-grower agribusiness linkage.

Summary and conclusion

The case study is an example of a successful smallholder contracting arrangement in the sugar industry. The results illustrated that the transaction characteristics of the cane supply-processing operation influenced the level of managed co-ordination required, that small-holders incur higher levels of transaction cost and that they are able to compete, in terms of production efficiency, with larger contracted growers. The usefulness of these results can be debated from a number of perspectives.

Firstly, the use of transaction cost theory can contribute to a more informed process of selecting the right level of managed co-ordination for a given set of contractual conditions. Secondly, the case study clearly demonstrated that differential small-scale farmer transaction cost could be broken down into cost elements. This can be useful in a number of ways. Incremental cost, once identified, can be charged back to the grower, form the basis of an agribusiness approach to the state for assistance or, lastly, provide the motivation to organise farmer inputs by way of a farmers association. These findings can be used as a basis to convince agribusiness that small-scale operators can operate as viable business partners and at the same time create an economic externality to the benefit of the whole of society.

Small-Scale Growers in the Timber Industry: Sappi’s “Project Grow”

The South African forestry industry is an important player in the South African economy. This industry consists of two primary segments, namely, the growing of timber, which falls into the agricultural sector and the processing of timber, which falls into the manufacturing sector. The timber industry contributed a total of 2 percent of the total national gross domestic product in 2000/1. This contribution, over a twenty two-year period, has grown from 1.4 percent in 1979/80 to its current level. The growing of timber, currently, contributes towards around 8 percent percent of the national agricultural gross domestic product and the manufacture of pulp and paper products accounts for 9 percent percent of the national manufacturing gross national product. This industry, moreover, makes a significant contribution to foreign trade and forestry products generated an annual net trade surplus of R 3.3 billion, or 8 percent of South African exports, in 2000/1. The industry has demonstrated a consistent annual growth of 8 percent over the last ten years and is one of the most internationalised industries in the South African economy.

The South African forestry sector is a major employer of labour in South Africa. In 2000/1 the industry employed 135 000 people who were either engaged in the primary production of timber or the processing of wood. This sector employs 5.7 percent of the total working population and includes the support of over two million dependants, many of whom live in rural areas. The estimated linkages effect generated by the industry would suggest an average multiplier of four resulting in a total employment potential of 500 000 people. The industry, moreover, is a major contributor towards the development of rural infrastructure and provides R 15 million per annum to the provision of housing and R 40 million per annum to the provision of health care. Other contributions to rural infrastructure include R15 million per annum towards the provision of schooling and bursaries and R10 million per annum towards the maintenance of provincial rural roads.

The South African forestry industry, with a capital base of R25 billion and an annual turnover of R12 billion, is a major supporter of the small-scale grower sector. The industry has promoted the development of 15 000 emerging timber growers, in addition to, the promotion and support of forestry contractors and entrepreneur development programs. There are two principal types of commercial woodlot afforestation that embrace the small-scale farm sector. The first type is co-ordinated and sponsored by agribusiness and includes examples like Sappi’s Project Grow, Mondi’s Khulanathi Project and initiatives set up by the South African Wattle Growers Association. The second type of small-scale grower operations is typified by ad hoc, uncoordinated individual plantings where no records exist and no authorisation has been received from the relevant authorities.

A summary of small-scale timber production reveals that 50 percent of these growers are located in Zululand, 29 percent in the Natal Midlands and 11 percent in Southern Natal. Small-scale growers have, therefore, developed predominantly in Kwazulu-Natal with a small percentage in the Eastern Cape. The incidence of forestry as a farming practice appears to be influenced by land tenure systems where in Kwazulu-Natal woodlots are individually owned whereas in the Eastern Cape, land is owned on a communal tenure basis. At present some 24 205 hectares have been planted in managed small-scale grower schemes. The estimated total area planted, however is 43 455 hectares suggesting that 44 percent of the total area under small-scale farm afforestation is unauthorised and that 35 percent of farmers are operating without permission.

Some estimates have been made of the potential to expand the small farm sector and project that this sector can grow at an annual area of 17 630 hectares between 2001 and 2005, resulting in a total increase of 93 100 hectares involving 10 197 new applicants. The principal areas of new afforestation are projected to be the Zululand area where the industry would expand by 27 225 hectares and the Eastern Cape at an estimated additional 62 000 hectares. These schemes represent an investment of more than R30 million, which is expected to generate some R 60 million when clear felling takes place.

Linking arrangements

The Sappi-Saiccor timber supply-processing operation includes a processor, namely, Sappi-Saiccor, a company in the Sappi Forest Products Group, and a range of company and contracted timber growers in Kwazulu-Natal, Mpumalanga, the Eastern Cape and the Highveld. The company produces dissolving pulp, a product that is made entirely from hardwoods, namely, eucalyptus and wattle. In 2001 a portion of the production process was subcontracted in order to reduce the level of fixed cost and induce capacity flexibility.

The growers: The entire timber supply operation is co-ordinated by the Sappi Forest Division and Sappi-Saiccor. Sappi Forest Division controls the activities of all the growers including the company plantations and a range of small-scale to large contracted growers. The first category of grower, namely, Sappi Forest, owns and manages 500 000 hectares of plantations in Southern Africa that primarily grow eucalyptus and softwoods.

The second category of grower includes medium to large scale contracted farmers. These growers are typically medium to large size family farms with an area under timber in excess of 50 hectares. These farmers are generally involved in a number of agricultural enterprises, which typically include timber, sugarcane, tea and fruit. Their timber plantations, thus, constitute only a part of the farming operation. This category of grower has been registered as a contracted supplier with the Sappi Forest division and the conditions of supply are dictated by a long-term specification contract. These growers are largely autonomous with respect to the growing of timber but the felling and delivery operations are controlled and co-ordinated by Sappi Forest Division who controls the supply of all timber to the Sappi-Saiccor mill. Sappi Forest personnel, however, can be consulted by the grower on an ad hoc basis. The timber operations of these farmers are less capital intensive than the company plantations with respect to plant, vehicles and equipment and in addition the farmers also subcontract a majority of the felling-delivery operations.

The third category of grower includes managed small-scale farmers incorporated in Sappi's Project Grow program. This category of grower, occupy on average 0.6 hectares each, is mostly located within a 100 kilometre radius of the company mill. Project Grow is a tree farming scheme that has the objective of converting rural subsistence farmers into emerging commercial operations. This project was launched in 1983 by Sappi Forests, the Gencor Development Fund and the Kwazulu Department of Agriculture and Forestry with a view to developing viable small scale timber operations in rural Kwazulu-Natal. Since 1989, Sappi Forest division has contracted out the management of this project to a rural development organization called Lima, which is a non government organization registered under Section 21. This organization receives an annual fee from Sappi Forest division, to administer the Project Grow initiative with two management staff, two administration staff, six extension officers and eight field assistants.

The Sappi Project Grow arrangement provides small farmers with financial assistance, seedlings, technical advice and a guaranteed market. Sappi Forest, via the management company Lima, provides an interest free loan of up to a maximum of R2 700, calculated on a per hectare basis, for farmers to establish trees where all contracts have the approval of the local tribal authority. Thereafter, advances are paid out to the farmer for completed certified work over the growing period of the trees to ensure that operations are funded over the growing cycle. The contracted farmers are thus paid to do the land preparation, planting, weeding, coppicing maintenance and the management of firebreaks. Lima extension officers visit the growers frequently after the trees have been established to provide further assistance with weed control and the preparation of fire breaks. Sappi Forest, if requested by the growers and Lima, may also assist during negotiations with harvesting and transport contractors. At the time of harvesting Sappi Forest buys the timber from the farmers at a market related price less the advances paid out during the growing period. Additional benefits from growing timber include fencing, building material and firewood which is obtained as a result of coppicing when the trees are between one and a half years and three years old.

Sappi Forest Division has invested in excess of R10 million in Project Grow in terms of loans and an additional R5.2 million in seedlings. A summary of the growth of this project indicates that the number of growers has increased from 101 to 7 100 in the period 1989 to 2001. A total area of 4 224 hectares is currently being managed under this programme.

Timber deliveries in excess of 1 500 tonnes of timber per month were achieved for 2001 generating some R2.6 million for the season. The estimate turnover for 2002 has been set at R3.5 million. It estimated that an additional 1 120 people are employed by contractors who assist the growers with the planning and harvesting of their plots. The project contributes towards the uplift of women in this area as some 80 percent of the growers registered with the project are women. The project generates considerable revenue for local communities with an estimate of 50 percent of turnover retained within the community as a result of payments to local contractors, 42 percent retained by the grower and 8 percent refunded to Sappi as loan repayment.

The supply contract: All contracted suppliers are required to enter into a timber purchasing agreement with the Sappi Forests division. The purchasing agreement specifies the exact location of the grower as well as the commencement and duration of the relationship. The agreement indicates the total tonnage to be delivered to the mill during the period of the contract and also stipulates the annual tonnage. The contract specifies the price that the company will pay for the tree species to be delivered or alternatively that the parties shall agree to an annual price. The supplier must adhere to quality specifications as determined by the company mill. The supplier is required to obtain the necessary permits, license or statutory authority from the Department of Water Affairs and Forestry, the National-Provincial Environmental Authority and the Department of Agriculture. The conditions of delivery, risk, ownership and payment are outlined in the contract with the risk only passing to the mill once the specified timber has passed over the company weigh-bridge. The date and mode of payment for a timber delivery is also specified. The company undertakes to supply seedlings of a specific species on condition that the supplier gives proper notification and that transport costs are to be borne by the supplier. The company also undertakes to provide free technical advice during the growing cycle of the timber however, the supplier must provide reasonable notification to the company and access to the growing site. A clause is inserted to cover both parties from “force majeure” and outlines the terms and conditions of the suspension or waiving of contractual liabilities. The enforcement of the contract is stipulated by way of written notice to the defaulting party and the supplier may not sub-contract or cede any of the terms and conditions of the agreement to a third party. Finally, the contract specifies the domicilia of the parties and outlines further miscellaneous legal clauses to the purchase agreement.

In certain cases, suppliers enter into a financial assistance agreement with Sappi Forests. This agreement stipulates the background of the applicant, the duration of the arrangement and an exact schedule of the growing and harvesting of specific species of trees. This agreement, moreover, stipulates the rate of interest to be paid to the company together with notification of a liability for finance charges. The terms of repayment are specified by way of a deduction of the financial assistance received from payments made with respect to the supply of timber under the timber supply agreement. Furthermore, the conditions that apply in the event of the non-supply of timber to the company, for whatever reason, are outlined. The supplier, applying for financial assistance, should be the registered owner of the stated property, and, if a loan in excess of R50 000 is made, then the supplier is obligated to register a covering mortgage bond in favour of Sappi Forests. In certain cases where the supplier plants in excess of 500 hectares the company may enforce a timber servitude on the supplier as an additional measure to enforce the contract. The contract, moreover, stipulates the general obligations of the supplier and includes conditions that enforce the supplier to comply with all environmental and silvi-cultural requirements. The contracted supplier, moreover, must sell the specified timber to Sappi Forests when the trees are at a specific age at a market related price relative to the area in which the mill is established. The financial assistance also includes the provision of free technical advice to the supplier up to a stated number of visits per year and outlines the risk-insurance requirements to be met by the grower who shall forward a copy of the insurance agreement to the company.

A different supply arrangement is used to contract small-scale farmers under the Project Grow arrangement. The Grow Agreement involves an arrangement whereby the Sappi Forest division supports the supplier both financially and technically. The duration of the arrangement is specified and the terms of assistance outlined. Assistance is received in the form of an initial interest free loan for planting, maintaining and weeding the timber. Sappi Forests also undertakes to provide seedlings. The grower must demonstrate they have all the necessary permits, licences and authority to grow timber on the said property including the compliance of the Department of Water Affairs, the National-Provincial Environmental Authority and the Department of Agriculture. The grower undertakes to meet a range of obligations that include compliance with Sappi Forest's environmental and silvicultural practices and access to inspection by all stipulated parties. The grower is obligated to sell the timber to Sappi Forest and this timber must comply with the stated mill specifications. The supplier must also comply with Sappi Forest's instructions to harvest the timber at a specific age. The price paid for timber is negotiated between the parties and corresponds to the prevailing market price. All risk of damage remains with the grower until it has crossed the weighbridge although timber that does not meet mill specifications may be rejected. The agreement, furthermore, cedes the grower’s rights to the purchase price as a measure to provide additional security to the company. Finally, the contract outlines the conditions relating to the breach of the contract by the grower and the manner in which the contract will be enforced. The grower is not allowed to cede any rights or obligations to third parties and all notices to the grower are to be delivered personally by the company or at monthly Project Grow meetings.

Transaction Characteristics. The interface between growers and the company mill, with respect to the continuous supply of large volumes of timber, generates a unique set of transaction characteristics. The number of deliveries of timber to the mill can be classified as the transaction characteristic of frequency and the value-degree to which the assets of the processor-grower are tied to the timber supply operation can be classified as transaction characteristic of asset specificity. Finally, the level of supply uncertainty has been classified as the transaction characteristic of uncertainty. The actual Sappi-Saiccor timber supply transaction characteristics, illustrated in Table 5, were developed for the period 2000/1 and are based on the delivery of over 1.6 million tonnes of timber to the Sappi-Saiccor timber yard.

Table 4: Transaction Characteristics of the timber supply chain

Transaction Characteristic


Types of Growers

Estate, Large, Medium, Small, Micro


> 500 000 hectares

Tonnage Crushed

1.64 million tonnes

Number of Deliveries



5.3 tonnes/transaction

Asset Specificity


Co-ordination Level

12 months/year/24 hrs/day 7 days/week road-rail, wood chips perishable, mill requirement 6 000 tonnes/day

Value of Estates

> R 3.8 billion (net operating assets)

Value of Plant

> R 5 billion (replacement cost)



Company Estates

Legislation, environmental issues, cost of inputs, physical variables, land constraints

Medium-large Growers

Timber Prices, physical variables, limited additional land, water cost, environmental

Small-scale Growers

Different time horizons, land tenure, cost of inputs, legislation, lack of access, moral hazard, theft


High Degree of leverage


Volatile markets, changing nature of industry > high uncertainty

Transaction frequency: The transaction characteristic of frequency, illustrated in Table 5, has been developed on the basis of the number of timber deliveries that are made to the timber supply yard. The frequency of timber deliveries is high because of the need to maximise the use of mill capacity of approximately 6 000 tonnes of timber per day, combined with the limited ability of the timber supply yard to stockpile the commodity. The timber supply yard, therefore, needs to be replenished on a daily basis to ensure that stock-out does not occur. The frequency of deliveries indicates that 46 669 truckloads of timber resulted in the delivery of 1 663 520 tonnes of timber involving 1 489 203 tonnes of gum and 174 317 tonnes of wattle. The number of deliveries by road exceeds the deliveries by rail with 33 176 delivered by road transport rigs and 13493 deliveries (railway trucks). Of this total, approximately 86 percent of road deliveries and 91 percent of rail delivery, consist of gum, which accounted for 90 percent of the timber processed by Sappi-Saiccor in 2001. On the basis of these records, this study found that a level of delivery frequency of in excess of 125 truckloads per day occurred for the full calendar year. According to the comparative levels of transaction frequency, developed by Williamson (1975) and Petersen and Wysocki (1997), this study has classified the level of transaction frequency as intermediate-high.

Asset specificity: The mill group, which has a capacity of processing an average of 6 000 tonnes of timber per day, employs four mills of different capacity. The net operating assets employed are currently valued at R 1.3 billion, on a historical cost basis, that translates to a current replacement cost in excess of R5 billion for the year ending September, 2001. These assets are highly specific and have a low opportunity cost outside the timber industry. The assets, moreover, are relatively immovable and are also site specific as the they have been located in close proximity to certain suppliers, harbour, rail and road facilities. The finished product is largely exported by ship and therefore the mill is site specific to the coast and harbour facilities. The factory assets also demonstrate high levels of asset specificity as a result of the need for high levels of co-ordination in order to maximise capacity usage. The timing of the delivery stream is, thus, co-ordinated in order to stagger the arrival of over 46 000 vehicles-railway trucks per annum in order to maintain capacity, as well as, reduce the turnaround cost of the timber supply yard. The logistics of timber supply requires the processor to co-ordinate a continuous delivery stream by road and rail from a diverse group of growers that are widely dispersed. The company must, moreover, ensure that each supplier is adhering to their contractual conditions. The high levels of co-ordination are also further influenced by the perishable nature of the semi processed raw commodity, namely, wood chips that are stored in an open yard and loose quality over time, especially in hot humid conditions. Finally, the level of asset specificity is increased by the timber specific knowledge and skills that have evolved, and are locked into, the timber processing industry. On the basis of the historical records, in conjunction with the grading of transaction characteristics it would appear that the intermediate-high levels of frequency, as a transaction characteristic, are matched with high levels of asset specificity in the processing operation group.

Uncertainty of supply: The uncertainty of supply has, historically, been relatively low due to a number of factors. Firstly, the uncertainty of supply has been reduced by the monopsonistic nature of the timber industry where Sappi Limited is a major player. Secondly, the company estates have, historically, produced more than 50 percent of the timber processed by Sappi-Saiccor and uncertainty of supply was further reduced by the site specificity of many growers who are located within a 50 kilometre radius of the Saiccor Mill. Thirdly, uncertainty is reduced by the long-term nature of timber production. Sappi Forest division is, in this regard, able to manipulate the supply of timber according to annual mill requirements and standing timber can, therefore, be felled if required or maintained until a future time when it is required. The economic viability of the standing timber is not affected due to the annual growth rate of this commodity. The uncertainty of supply has, therefore, been a function of the reliability of the company plantations and medium and large growers, combined with the normal agronomic and climatic variables that prevail in the timber growing industry. The uncertainty of supply from medium and large suppliers, in this respect, is reduced by a stringent specification contract whilst the activities of Sappi Forest are fully internalised in the structure of Sappi Limited. Finally, deteriorating international markets for dissolving pulp, combined with increased levels of competition, have resulted in Sappi-Saiccor reducing capacity by 25 percent. Given the same volume of timber supply, the level of future uncertainty looks set to decrease as a result of downstream market pressures. Finally, uncertainty is partially a function of asset specificity. The reason for this is that the high level of fixed cost of the processing operation contributes towards increasing the degree of leverage, which in turn, increases the volatility of company cash flows. Since increased levels of volatility are associated with higher levels of uncertainty, the high level of asset specificity will tend to increase supply uncertainty. On the basis of the analysis of the historical records, combined with the current concerns of the timber, this study concludes that the current level of supply uncertainty can be classified as low to intermediate. Should Sappi Limited divest out of the timber growing industry, as has been suggested, the projected level of supply uncertainty is likely to increase.

The governance structure for the timber supply chain. According to Petersen and Wysocki (1996; 1997) and Williamson (1981; 1996; 1998; 2000), transaction characteristics that include high levels of frequency and asset specificity, combined with a measure of uncertainty, require higher levels of managed co-ordination than the spot market mechanisms. On the basis of transaction cost theory, the control of timber supply could be co-ordinated by a range of structures from specification contracting to full vertical integration on the vertical co-ordination continuum of Table 6. The contract characteristics of the supply arrangement suggest, however, that this structure would involve medium to high levels of managed co-ordination. The actual structure of control at Sappi-Saiccor does not necessarily contradict this and indicates that 50 percent of supply is controlled by specification contracting and 50 percent from the company plantations indicating an aggregated control structure in the centre of the continuum.

Table 5: Governance structure in the timber supply chain

Theoretical Optimum Actual Structure

Spot Market

Specification Contracting (50 percent)

Strategic Alliance XX

Formal Co-operation

Full Vertical Integration (50 percent)

Level of managed co-ordination

0 percent





Governance Form

Contract Growers

Company Estates

Summary and conclusion

The case study demonstrated that large numbers of micro growers, namely, in excess of 7000 farmers, can be incorporated in the timber growing operation of an agribusiness partner. The timber case study also demonstrates that transaction cost theory can be practically employed to test the level of managed co-ordination in an agricultural supply chain. A number of worrying concerns emerged from this case study. Firstly, the high level of company support to the growers has resulted in a cost to Sappi Limited of R10 million since 1989. The project appears to have been conceived as a result of dual political economy-economic company objectives that translated into a paternalistic agribusiness approach to managing the smallholders. Secondly, the withdrawal of the administrative support of the agribusiness could result in the abandonment of the project. The smallholders, at this stage, appear unlikely to organise their own affairs by way of a farmers association. Thirdly, the case study also supports the widespread contention that smallholders generate incremental transaction cost. Finally, the company’s control systems are also, in many instances, unable-reluctant to disclose the differential cost of dealing with smallholders.

Small Growers in the Tea Industry: The Sapekoe Story

This case study evaluates a tea supply chain in Sapekoe Estates (Pty) Ltd - a company located in the Northern Province of South Africa. The case study focuses specifically on one of Sapekoe Estates’s plantations, namely Tshivusa Plantation situated in the former Venda Homeland. The Tshivase Estate, along with the rest of the company actively promoted small farming after 1987 and by 1992 some 157 small farmers were engaged to supply green tea leaves to the factory. In January 1999, the estate had engaged 330 small farmers occupying an area of 192 hectares. Each farmer occupied an area of approximately 0.6 hectares where 0.5 hectares was under tea. An annual contract was drawn up between the small farmer, as a grower, and the company as a processor. This contract was translated into the language of the farmer. The profile of the farmers indicated that a high percentage (as high as 96 percent) of mini farmers were females between the age of 22 and 53 years. Their profile indicated as many as 50 percent were illiterate whilst 38 percent had been educated at a junior primary school level only.

A series of quarterly reviews maintained between January 1999 and April 2001 reveals a trend of a reduction in company involvement with the mini farming project. The principal reasons included small farmers not adhering to plucking schedules, poor field practices, low yields and a lacking of an entrepreneurial ethos. In March 1999, it was decided to reduce small farm projects from 330 people occupying 192 hectares to 90 farmers occupying 67 hectares. In the period 1/4/1999 to 30/4/2001 the mini farmers complied with standards and demonstrated high levels of productivity.

In 1999 the small farmers improved productivity from 63.7 kg of green tea per labour unit per day to 99.9 kg. In 2000, weeding-fertilising operations were reported as satisfactory and plucking and pruning productivity often exceeded that of the estate. During this period all standards were maintained. The number of farmers reduced to 87, then 80 on the same area and by 31/10/2000 the company had the intention of expanding this number to 100. The mini farming project on Tshivase was terminated in June 2001 due to a change in legislation. The reasons for failure include the poor selection of participants, the poor management skills of the small farmers, the limited duration of the contract, volatile world prices and major strikes in 1996 where company workers intimidated the contracted farmers. The company also displayed a history of paternalism in the corporate culture, a high level of scepticism of company officials and the small farmers, who were originally employees, never became independent farmers.

Linking arrangements

Tshivase Small Farm Contract: The agreement is embodied in the form of a tea harvester agreement between the estate and the designated participant. The agreement stipulates a specified area as designated in the company records and actions both parties are regulated in terms of Section 45 Act 32 of 1994. This act controls the institution of any legal action providing for the enforcement of any rights under, or arising from the agreement, in the Magistrate Court which has jurisdiction in respect the estate.

Transaction characteristics: The Tshivase Factory has processed an average of about 7,000tonnesof green leaf tea in the period 1999 to 2001. The harvesting and delivery are co-ordinated and synchronised in order to maintain constant use of factory capacity which is estimated at 74 tonnes per day in two shifts. Capacity utilisation is hence 51.4 percent. The harvesting and delivery of green tea leaves co-ordinates some 36 different fields occupying an area of 577 hectares. Each field is plucked on an 8 to 12 day cycle indicating a high level of labour management as the teams move from field to field according to the designated plucking programs. The plucking cycle results in each worker’s basket being weighed in the field and placed in a covered light delivery vehicle or tractor-trailer for delivery to the factory. The deliveries are weighed on entry to the processing operation and average out at around 1 500 kg per delivery. This would indicate that the factory receives in excess of 4 700 deliveries of green tea per annum. Each delivery should be offloaded in 20 to 30 minutes and the perishability of green tea leaves increases the need for high levels of co-ordination between the factory and the growers. In addition, the number of company weighing-inspection transactions in the field averages at around one transaction per 45 - 60 kg per labour unit (June 2001 Plucking Summary). There is a minimum of 25 weighing-inspection transactions in the field for an average delivery of 1 500 kg or some 117 500 field transactions per annum.

The high levels of frequency, as a transaction characteristic, are matched with high levels of asset specificity in the factory group with lower levels of asset specificity for the agricultural division. The assets of the Tshivase Estate factory were valued at R 7.56 million at gross book value in 2001. These assets are highly specific to the tea industry and have a low opportunity cost outside the industry. The assets, moreover, are relatively immovable and are also site specific as they have been centralised in relation to the company estates and out-grower suppliers. The factory assets also demonstrate high levels of asset specificity as a result of the need for high levels of co-ordination in order to maximise the use of capacity. The co-ordination of over 4 700 deliveries from 36 different fields has been that every field is plucked on an 8 to 12 day basis. And every delivery is scheduled to have a turnaround time of 30 minutes and less. The high levels of co-ordination are further influenced by the perishable nature of the green tea leaves that rapidly loose quality if they are exposed after cutting.

The uncertainty of supply has, historically, been relatively low due to the fact that the company estates have produced more than 96 percent of the processed tea leaves. The uncertainty of supply has therefore been a function of the normal biological and climatic variables experienced in the agricultural sector. The uncertainty of supply has, at times, been increased by the unavailability of labour, labour disputes and other forms of absenteeism. Small farm supply has been less reliable than the company estates and is cited as one of the factors contributing to the abandonment of this form of supply. The uncertainty of future supply could be affected by the land tenure situation of the company who currently leases all of its land from local tribal authorities. The downstream uncertainty in the tea industry is, however, elevated because of the volatility of world tea prices. The transaction characteristic of uncertainty is also increased because of the high level of fixed cost of the processing facilities where the need to maintain mill capacity is vitally important. The company therefore has a high degree of leverage with respect to its fixed cost structure and any type of breakdown in the processing facilities, or alternatively, major shifts in climatic patterns, have a significant impact on the company cash flows.

Governance structure: The tea supply chain of Tshivase Estate displays the transaction characteristics of high levels of frequency, asset specificity and a moderate level of cane supply uncertainty. These transaction characteristics, in turn influenced by crop specific characteristics, clearly demonstrate that the Tshivase Estate requires high levels of managed co-ordination in order to co-ordinate large volumes of a highly perishable raw commodity that are supplied by the growers. It is therefore logical to argue that the Tshivase Supply Chain would not be able to function without high levels of managed co-ordination. The adopted governance form has resulted from the need to co-ordinate the high volumes of supply to maintain mill capacity at around 38 000 kg of green tea per day. The financial performance of the Tshivase Estate is severely affected if plant capacity is not utilised during the year, as there are moderately high levels of fixed factory cost. This study therefore has demonstrated that governance form is a function of transaction characteristics and that governance forms evolve in order to minimise total cost, rather than just transaction cost as stated by the proponents of transaction cost theory. Organization structure will thus impact on financial performance. In the case of the Tshivase Estate it would be clearly impossible to obtain and co-ordinate the supply of green tea leaves on the spot market.

The company, moreover, is unable to expand its own capacity from estate type supply because of an inability to acquire additional permanent land inputs because of a limited land market in conjunction with land tenure problems. Mill capacity, moreover, exceeds own supply capacity so Tshivase must acquire additional supply from out-growers or the mini farmer source. The Tshivase Estate has, therefore, clearly, adopted the optimum governance form over time in support of the original proponent of transaction cost theory, namely, Ronald Coase (1937). The Company has, therefore, clearly, adopted the optimum governance form in order to co-ordinate the large range of growers.

Grower Performance. The comparison of the growers has been evaluated on the basis of the plucking performance summaries that have been summarised for the period 1996 to 2000. The plucking performance evaluates the productivity of the estate and the mini farmers on the basis of the kg of green tea leaves delivered to the factory per hectare farmed and on the basis of the kg of made tea per hectare farmed. A further measure of efficiency also examines the green leaves picked per day per worker and expresses this measure as the yield per labour unit. The results suggest that mini farmers have outperformed the estate between 1996 and 2001 except for the period 1998/9. The mini farmers, on average, have been 7.26 percent more productive per hectare in terms of green leaves delivered to the factory, as well as, made tea than the balance of the estate. The labour yield per unit for the estate has been calculated on the basis of the conventional hand plucking technique and again the results indicate that the mini farmers have outperformed the estate in the period 1996 to 2001 except for 1998/9. The results for 1999/2000 were not available but, on average, the mini farmer yield per unit was 13.9 percent higher than the balance of the estate.


This case study provided useful insights into what should NOT be done with respect to the design of a smallholder linkage or out-grower project. Firstly, the company officials never viewed the contracted growers as independent farmers but rather as former employees. The autonomy of contracted farmers was, therefore, extremely limited. The design of the project, moreover, was rather directed at avoiding labour union problems rather than stimulating small-scale supply. The second important feature was the micro nature of the growers who operated on less than a hectare suggesting the need to evaluate a critical minimum size farm in any crop category. In this respect, there is some correlation with the timber case study. Finally, the reasons for the failure of this project are contradicted by the apparent competitive grower performance of the contracted farmers suggesting that the official company reason for the closure of the project, namely, the incompetence of the growers, was prompted by other reasons.

Factors Contributing to Success

Screening of Partners

The careful screening-identification of future partners is a key success factor. Farmers who have a record of previous interaction with agribusiness appear to be more successful contracting partners (Levin, 1988; Porter & Phillips-Howard, 1997a; 1997b). The three case studies all appear to have developed a data bank of farmer details for each prospective new farmer. In the case of the sugar industry, however, the screening process took more cognisance of the entrepreneurial ability of the prospective applicant, whereas, in the timber industry, it would appear as if a majority of applicants were accepted.

Screening costs, involving large numbers of applicants, can be significantly reduced if these activities are assisted by a representative farmers association or, alternatively, if the agribusiness contracts with the farmers’ association rather than the individual farmer. The benefits of screening, moreover, can be increased if the process includes a business aptitude test, a credit check and a list of assets-collateral. Finally, the screening process should capture the location, logistics and communication channels of the applicant in order to ascertain the spatial dynamics of the project.

Understanding of Historical Legacies

An understanding of the historical legacies and institutional environment will contribute towards a better understanding of the future transaction cost of contracting with large numbers of small-scale farmers. Transaction cost theory suggests that the transaction characteristics of agricultural supply chains are a function of a range of historical-social variables.

All of the case studies confirm the pervasive long term influence on economic performance of historical legacies that have influenced transaction cost as a result of the effect of culture, the historic concentration of industry, the influence on property rights economics, the level of regulation and the concentration of political power. Appropriate design measures can then be taken to reduce transaction cost in two ways. Firstly, many industries have the power to lobby for changes in the institutional framework and Williamson (2000) suggests that this form of economising can significantly reduce transaction cost. South African agribusiness, in particular, has the bargaining power to lobby for property rights amendments and some form of subsidy-tax relief for undertaking smallholder start-up costs. Secondly, the design of organization structures can be undertaken more efficiently if an understanding of the dynamics of transaction characteristics is incorporated.

Creation of Mutual Asset Specificity

The creation of mutual asset specificity reduces uncertainty and raises the exit costs of both sets of contracting partners. The case studies in the sugar and timber industries indicate that the business partner is confronted with significantly higher levels of asset specificity than the contracted farmers. The industry and site specific processing assets, in the sugar and timber case studies, were valued at R2 billion and R5 billion respectively. Conversely, the contracted farmers owned fewer assets that were of a more general nature. The South African sugar firm, TSB, in particular has a very high level of asset specificity and relies on contracted out-growers for 80 percent of sugarcane supply.

Transaction cost theory would suggest that a higher level of managed co-ordination is needed in the absence of inducing higher levels of mutual asset specificity or other interlocking factors. Mutual asset specificity can be pursued by way of farmers associations undertaking the purchase of industry specific capital inputs. The Swaziland sugar farmers associations appear to have increased mutual asset specificity by investing in sugar specific plant and equipment that is too lumpy for the individual farmer. The agribusiness can attempt to act as a facilitator of finance, in this regard, to increase the interlocking nature of the arrangement. Finally, the agribusiness can examine other ways of influencing mutual asset specificity by way of configuring the technology of the grower-processor operations in such a way that only the agribusiness possesses the technology to perform a specific element of the growing operation. Contracted growers, for instance, in the processed tomato sector, require specific harvesting technology that can be owned and operated by the agribusiness (Rehber, 1998).

Design of Logistics

The design of the logistics of small-scale farm supply is a crucial success factor. An understanding of the relationship between commodity characteristics and logistics could be incorporated to reduce transaction cost (Delgado, 1999). The timber case study, in particular, illustrates that the increased level of transaction cost are generated by large numbers of small-scale farmers that are spatially dispersed. The agribusiness, at the outset, can evaluate the transaction frequency of visits, inputs and farmer deliveries with the distance from the processor, the nature of the roads and the available communication system. The transaction cost of logistics can be fundamentally reduced by allowing a farmers’ association to provide the necessary inputs, as well as, organise the logistics of small-scale supply.

The timber case study, in particular, suggests Sappi Forest should contract with a farmers association for an aggregated monthly volume of timber instead of contracting individually with the 7 100 micro farmers. In the absence of a farmers’ association the agribusiness can improve efficiency by establishing the nature of the roads, access and communication systems of the proposed project. The timber industry, in this regard, indicates that certain areas are impassable in the wet season and that declining levels of rural security and high levels of ethnic conflict have resulted in a lack of access except for local community members. Finally, the agribusiness can assemble and program the transaction cost of logistics by capturing the spatial dispersion of the farmers, the number of transactions and the average distance to the processor.

Assessment of Start-up Costs

The careful assessment and treatment of start-up cost is a key project evaluation procedure. In many instances, the agribusiness must commit long-term resources to establish small-scale farmer projects. Sappi Limited, for instance, has invested ten years and R10 million to establish small-scale tree farming in Kwazulu-Natal whilst the Transvaal Sugar Company in Mpumalanga incurs an annual cost of in excess of R3 million to ensure the viability of small-scale supply Start-up cost will, generally, be higher if non traditional crops are being introduced because of the need to train farmers and introduce complex technology.

The start-up cost, moreover, can include the linking of small-scale farmers to institutions like banks, insurance companies and suppliers. The lack of access to these facilities, in conjunction with infrastructure deficiencies, has been cited as a prime cause of project failure (Gittinger, 1982). The assessment and treatment of start-up cost will influence the investment decision. This study proposes that the incidence of smallholder contracting in South Africa will be negatively influenced if agribusiness is expected to bear this cost. The assessment of start-up cost, therefore, should form the basis of lobbying for government subsidy-relief or alternatively to charge back this amount to the farmers. If the contracted farmers are unable to amortize start-up cost, the viability of the operation should be questioned from the outset.

Commodity Characteristics

Certain commodity characteristics are better suited to contracting. Crops, in particular those which are labour intensive in the growing operation and display economies of scale in processing, are more suited to smallholder contracting (Delgado, 1999). The case studies in the sugar and timber industries did not demonstrate particular growing economies for smallholder family labour yet these growers, mostly matched larger growers with respect to the cost efficiency of production. The reason for this ability to compete with larger growers appears to stem from the ability to avoid overhead cost rather than the productivity of family labour. The results of the case studies suggest that smallholders maybe able to compete as growers with commodities that are not particularly labour intensive thus further relaxing the suggested product range of Delgado (1999).

Commodity characteristics can also be linked to transaction cost for design purposes. Commodities, for instance, that are perishable will require higher levels of co-ordination cost than those that can be stockpiled. Alternatively, commodities that have long growing periods may require a different contract structure from annual or shorter term crops. Growers, for instance, in the timber industry sometimes receive advances for work performed against the sale of the future crop. The design of contracting structures can thus consider developing a commodity characteristics profile and use this, together with processing capacity, as the basis for determining the transaction characteristics of frequency, asset specificity and uncertainty to determine an optimum structure.

The case studies clearly illustrate the ability to plot actual transaction and contracting characteristics with the optimum governance form. The actual governance form can then be compared and moved along the vertical co-ordination continuum of managed co-ordination. Agribusinesses that incorporate both contracted growers and fully integrated estates can, possibly, employ a looser form of specification than those who rely solely on contracted supplies. Finally, the inappropriate choice of technology, a function of the commodity characteristics, has been cited a cause of project failure (Gittinger, 1982). The agribusiness often has a choice of technology alternatives and it has been suggested that if a labour intensive option does not detract from performance, then this option should be chosen to better suit the competencies of the farm family

Considering Transaction Costs

Contracting with large numbers of small-scale suppliers has been associated with higher levels of transaction cost (Runsten & Key, 1996; Rehber, 1998; Key & Runsten, 1999). The transaction costs of small-scale suppliers in the case studies clearly exceeded those of larger suppliers. The case studies demonstrated that transaction cost can be broken down into a series of cost elements including start-up cost, growing costs, harvesting-delivery costs and administration costs. The economics of contracting would suggest there is no reason for the agribusiness to choose small-scale suppliers over larger growers if all the contracted parties are paid the same price and deliver the same level of quality. This being the case, it is necessary to take specific steps to either avoid the cost or, alternatively, charge back the differential cost to the contracted party. Recurrent small farmer transaction cost can be avoided by contracting with a larger entity like a farmers association who undertakes the administration of its members’ interests.

The farmer association, moreover, can be responsible for configuring its members with agribusiness requirements including training, extension, technology acquisition, provision of commodity inputs and co-ordinating harvesting-delivery schedules. The agribusiness firm can increase the successful operation of the farmers association by acquiring representation in the management structure, as well as, allowing the association to be represented in it's own management structure. The agribusiness, moreover, can further influence the efficiency of the farmers association by ensuring this body maintains records, has no political agenda, is limited in size and that it contains sufficient professional management. A different approach to reducing transaction cost for the agribusiness can be engineered by way of charging back differential transaction cost to the small-scale grower by using activity based costing systems to identify the smallholders’ incremental use of company resources.

The case studies illustrate that differential administration cost can be charged back to the grower on the basis of identifying accounting transactions as the primary cost driver. Activity based costing can also be applied to charge back incremental growing, harvesting and delivery transaction cost. The timber case study clearly indicates that the agribusiness cost of timber from small-scale suppliers is higher than medium-large growers and the company plantations. Finally, activity based costing can be used as a basis to highlight the incremental cost of recurring smallholder transactions with a view to lobbying the South African government for assistance-relief.


The history of contracting demonstrates that the presence of competitive fresh markets for grower outlets increases the level of uncertainty of supply. Transaction cost theory explains the increased level of cost to the agribusiness in terms of higher levels of opportunism by the grower. The occasional opportunistic sale of timber in rural Kwazulu-Natal highlights this problem that was well documented in the Mexican tomato-growing sector (Runsten & Key, 1996). The unauthorised sale of the contracted commodity can especially problematic in the case of projects involving large numbers of smallholders in developing countries with poorly defined-upheld property rights economics.

The agribusiness, in this type of scenario, may not be able to legally enforce the contract because of the incremental cost of first dealing through an inefficient system and secondly because of the micro nature of the contract. The agribusiness can reduce unauthorised sales by securing an agreement with competitors with regard to the purchase of the commodity. Alternatively, the chances of contract enforcement are improved if market based prices are paid for the raw commodity. The agribusiness could also locate outside the area of the competitive fresh market.

Fafchamps and Minten (1999) suggest that trust based relationships can be a dominant interlocking factor that can contribute to contract enforcement. Transaction cost theory suggests that trust influences uncertainty as a result of its effect on the opportunistic behaviour of the contracting parties. Farmer distrust, combined with a perceived loss of autonomy and feelings of exploitation, has been widely cited as a major cause of contracting failures in developing countries (Glover, 1987; Clapp, 1994; Watts, 1994). It has been suggested, in this regard, that the level of trust that can be engendered between the parties will influence the success of future South African contracting arrangements. The development of trust is especially important given South Africa's history of colonialism and apartheid (Porter & Phillips-Howard, 1997a; 1997b).

Gow et al (2000) have demonstrated that contract innovations, or interlocking factors, can contribute towards reduced transaction cost. These factors include the administration of growers’ affairs, the company acting as financier and supplier of inputs and high levels of involvement in local communities. The timber industry case study demonstrates the interlocking nature of a contract that provides financial assistance and/or part payment for certain phases in the growing process. Colchao (1999) suggests the agribusiness can successfully induce contract enforcement by acting as a banker to the contracted farmer. The agribusiness, moreover, is able to compete in the banking sector as a result of better information combined with the ability to enforce contracts in alternate ways. In this regard, the business could attempt to own-control the assets and technology of the grower, as well as, play a role in the financing of these assets (Colchao, 1999). The timber case study, moreover, suggested that the small-holder management company, Lima, has effectively become a high cost interlocking mechanism because it is so integrated in the everyday affairs of contracted growers.

Contract enforcement is an important success factor with respect to small-scale farmer contracting. In many instances contract enforcement is difficult to ensure through the legal process and the logic of contracting with large numbers of smallholders is a questionable issue (Runsten & Key, 1996; Rehber, 1998; Sofranko et al, 2000). The representation of farmers’ interests by way of a farmers’ association will reduce the agribusiness cost of enforcement. Alternatively, the agribusiness can employ tribal authorities to enforce the contractual conditions. The logic, for instance, of attempting to legally enforce a set of contract conditions with respect to a farmer on less than a hectare of communal tenure land, as was the case in the timber study, is questionable.

The case studies in the sugar industry suggest that contract enforcement is achieved through the mutual interests of the parties rather than through the judicial system. The use of a renewable contract is suggested as a cost effective way to achieve enforcement (Key & Runsten, 1999). If suppliers have not performed in the previous year their contracts are simply not renewed the following year. Conversely, the firm in the timber case study appeared to pursue contract enforcement by way of registering timber servitudes or bonds, Williamson (2000) suggests that transaction cost can also be reduced by way of first order economising. This approach attempts to favourably influence the prevailing institutional environment in order to influence the economics of property rights. South African agribusiness, in this regard, has the potential to influence legislation that will reduce transaction cost. Legislation that could be amended includes land tenure, the water act, the role of tribal authorities, the labour act, rights of the female farmer, the national heritage acts and the conservation laws. The timber case study is an example of legislation that substantially increases transaction cost by way of a plethora of acts and legislation that must be complied with in order to register a new grower. The grower response to prices and other opportunities is suggested as a key long-term issue that can influence the stability of the contracting arrangements. A long-term perspective on prices could contribute to locking contracted growers into a commodity and ensuring continuity of supply in depressed conditions (Levin, 1988; Watts, 1994; Abbott, 1994).

Other Issues

A number of other issues influence the success of smallholder contracting schemes. These issues include the role of female farmer, the control of land and water, the role of the state, the household food security issue and land degradation. The role of the female farmer is especially important in many developing country contracting arrangements (Carney, 1988) and Gittinger (1982) suggests that a failure to understand the social environment is a prime cause of project failure.

The case studies in the sugar and timber industries suggest female household members supply high levels of labour inputs. Agribusiness in South Africa can contribute towards the future role of the female farmer by securing legislation that ensures full representation-rights for the female farmer in communal tenure areas. It is also suggested that agribusiness payment for the commodity should be directed to the household member responsible for supply. The role of tribal authorities in communal areas will also need to be configured to promoting gender access in rural areas.

The issue of who controls land and water in the contract relationship can influence the success of the arrangement. If the land and water is owned by the contracted growers then the decision making autonomy of the farmer is not affected, however, if they are owned by the agribusiness the conditions of use should be mutually developed and fully understood by both parties.

Another key issue is the role of the state. In many developing countries the state has been an active partner of small-scale contracting projects. It has been suggested that if the state is a partner then, at the very least, it should provide some form of financial assistance. The equity objectives of the state can also be investigated with respect to their impact on economic performance.

Finally, the issue of food security and land degradation can influence the long-term viability of contracting arrangements. Monoculture contracting has been associated with a reduction in food crops and an increase in pollution (Rehber, 1998; Pasour, 1998; Wolz et al, 1999). Agribusiness can respond to these threats by encouraging farmers to grow food crops on a percentage of their land and pro-actively investigating the long-term threat of growing technologies on sustainable land use.

Summary and Conclusion

The results of the case studies, combined with the potential of the South African agribusiness sector, suggest that large numbers of small-scale farmers could be linked to agribusiness partners by way of contracting arrangements. The proposals suggested, however, that a “fresh approach” to the design of these arrangements is required. This approach combined the lessons of history, conceptual developments in economic theory and the results of the case studies. The results, generally, suggested that the transaction characteristics of the grower-processing operations influenced the level of managed co-ordination required, that smallholder transaction cost exceeded that of larger suppliers and that smallholder production efficiency matched that of larger growers. The design of proposals, in response to these results, indicated a number of potential solutions that included the formation of a farmers association, the possibility of state assistance or the need to develop costing systems to identify, and charge back, incremental transaction cost to small-scale growers.

The case studies also suggest the pervasive, and long-term, influence of social-historical legacies on the economic performance of respective industry sectors. More specifically, the case studies demonstrate how the institutional structure in Southern Africa has been influenced by the earlier experience of colonialism-apartheid combined with the original concentrations of industry and infrastructure. The South African case studies, for instance, suggest that two hundred years of apartheid-colonialism have fundamentally influenced principal-agent costs, the concentrations of infrastructure and the property rights economics of the country. The timber case study, for instance, indicated the plethora of regulations-procedures required by a prospective grower to comply with the requirements of both local and national authorities before a water permit was granted. Clearly, the transaction cost incurred by the agribusiness of assisting these growers was influenced by both the degree of regulation involved, as well as, the inefficiency of local government authorities.

The current, somewhat inflexible, labour act in South Africa is a further example of how the costs of labour contracting have been influenced by a government attempting to redress historical imbalances. The case studies also demonstrated how the original concentration of the sugar and timber industries in Southern Africa was located in limited areas that provided the necessary natural resources. The government of the time, international donor bodies and prevailing multinationals then "kick-started" these industries by providing major inputs, infrastructure and policy to protect these fledgling industries until they were able to compete. The original establishment of these industries, combined with a lack of incremental natural resources, presents an almost insurmountable barrier of entry for new entrants. These industries, moreover, display high levels of site, asset and human skills specificity as a result of the historic concentrations of economic development in specific regions of Southern Africa, as well as, the evolution of human skills and knowledge over the long term.

In conclusion, this report summarises the cost to agribusiness of assisting smallholders overcome the barriers of entry to high value cash crop sectors. The results of the case studies can be used by agribusiness with respect to acquiring a better understanding of the process and sacrifices involved. Smallholder contracting projects often involve many years of agribusiness inputs before supply commences. In many instances, moreover, the company is drawn into protracted equity issues involving a local community. The study, in particular, identifies some of the pitfalls and hidden costs that agribusiness can incur when embarking on small-scale contracting projects. The timber case study, in particular, is indicative of the difficulties of managing large numbers of micro farmers that appear to be unable to be consolidated as an economic entity. The withdrawal of the financed management structure of the agribusiness, in this instance, would result in the collapse of the project and the question needs to be asked, whether or not, the micro farmers have really overcome the barriers of entry, on a permanent basis, to this industry. Contracting projects, ideally, should result in the establishment of permanent growers that operate as viable business entities. Whilst support in the start-up phase is a necessary pre-requisite to overcoming the barriers of entry, the contracted farmers need to be weaned out of the company structure on a long-term basis.


Abbott, J.C. 1994. Agricultural processing enterprises: Development potentials and links to the smallholder. In: Joachim von Braun and Eileen Kennedy (eds.) Agricultural Commercialisation, Economic Development and Nutrition. John Hopkins University Press, London

Carney, J.A. 1988. Struggles over crop rights and labour within contract farming households in a Gambian irrigated rice project". Journal of Peasant Studies. 15 (3): 334-349, April 1998.

Clapp, R.A. 1994. The Moral Economy of the Contract. In: Peter D. Little and Michael J. Watts (eds.) Living under Contract. The University of Wisconsin Press.

Colchao, S. 1999. Will AG Banks prosper in age of vertical integration. ABS Banking Journal, Vol. 91 (11) 26-31

Delgado,C. 1999. Sources of growth in smallholder agriculture in sub-Saharan Africa: The role of vertical integration of smallholders with processors and marketers of high value-added items. Agrekon, Vol. 38, Special Issue, pp 165-189, May 1999.

Esterhuizen, D. 2001. Global competitiveness in the Agro-Food industry - A South African case study. University of Gent, Unpublished MSc in Agricultural Development Economics

Fafchamps, M. and Minten, B. 1999. Property rights in a flea market economy. MSSD Discussion paper No 27, International Food Policy Research Institute, Washington, DC. March 1999.

Gittinger, J.P. 1982. Economic Analysis of Agricultural Projects. John Hopkins University Press, Baltimore, Maryland.

Glover, D. 1987. Increasing the benefits to smallholders from contract farming: Problems for farmers’ organizations and policy makers. World Development, 15 (4): 441-448

Gow, H.R., Streeter, D.H. and Swinnen, J.F.M. 2000. How private contract enforcement mechanisms can succeed where public institutions fail: the case of Juhocukor a.s. Agricultural Economics, Vol. 23 Special issue: 253-265. September 2000.

Horngren, C.T., Foster, G., Datar, S.M., & Uliana, E. 1999. Cost Accounting in South Africa. Prentice Hall South Africa

Kandiwa, V. 1999. Economic performance of smallholder farmers using alternative vertical coordination mechanisms for horticultural crops. MS thesis, Department of Agricultural Economics, Cornell University.

Key, N. and Runsten, D. 1999. Contract farming, small-holders, and Rural Development in Latin America: The organization of agroprocessing firms and the scale of outgrower production. World Development Vol. 27 No 2: 381-401.

Kirsten, J. & van Zyl, J. 1996. The contemporary agricultural policy environment: undoing the legacy of the past. Agricultural Land Reform in South Africa, Oxford University Press, Cape Town, 1996.

Levin, R. 1988. Contract Farming in Swaziland: Peasant differentiation and the constraints of land tenure. African Studies 47 (2): 101-120.

Pasour, E.C. 1998. The potential impact of increased vertical integration on North Carolina grain farmers North Carolina State University, December 1998.

Peterson, H.C. & Wysocki, A. 1997. The Vertical Coordination Continuum and the Determinants of Firm-Level Coordination Strategy. Staff Paper 97-64, Michigan State University.

Peterson, H.C. & Wysocki, A. 1998. Strategic Choice along the vertical co-ordination continuum. Staff Paper 98-16, Michigan State University, East Lansing, Michigan 48824

Porter, G, & Phillips-Howard, K. 1997a. Contract farming in South Africa: A case study from Kwazulu-Natal. Geography: Journal of the Geographical Association, Vol. 82(354), Jan.1997, pp.1-38.

Porter, G & Phillips-Howard, K. 1997b. Comparing Contracts: An evaluation of contract farming schemes in Africa. World Development 25:227-238.

Reardon, T. and Barrett, C.B. 2000. Agro-industrialization, globalization, and international development: An overview of issues, patterns, and determinants. Agricultural Economics, Vol. 23 Special issue, September 2000: 195-205.

Rehber, E. 1998. Vertical integration in agriculture and contract farming. Regional Research Project NE-165: Private Strategies, Public Policies, and Food System Performance, University of Massachusetts, Dept. of Resource Economics, Amherst, MA 01003.

Rhodes, V.J. 1993. Industrialization of agriculture: Discussion. American Journal of Agricultural Economics, Vol. 75 (5): 1137-1140.

Royer, J.S. 1995. Potential for cooperative involvement in vertical coordination and value added activities. Agribusiness, Vol. 11(3): 473-481.

Runsten D. & Key, N. 1996. Contract Farming in Developing Countries: Theoretical aspects and analysis of some Mexican Cases. Research Report No. 3, August 1996. Report prepared for the United Nations Economic Commission for Latin America and the Caribbean, Santiago, Chile.

Sofranko, A.; Frerichs, R.; Samy, M. and Swanson, B. 2000. Will farmers organize?: Structural change and loss of control over production.

Tregurtha, N.L. & Vink, N. 1999. Trust and supply chain relationships: A South African case study. Agrekon. Vol. 38 (4): 755-765

Unnevehr, L.J. 2000. Food safety issues and fresh food product exports from LDCs. Agricultural Economics. Vol.23 (September 2000): 231-240.

Watts, M.J. 1994. Life under contract; Contract farming, Agrarian Restructuring, and Flexible Accumulation. In. Peter D. Little and Michael. J. Watts (eds.) Living Under Contract. The University of Wisconsin Press, Madison.

Williams, S. 1985. The Mumias Sugar Company: A nucleus estate in Kenya. Agribusiness and the Small-Scale Farmer, Westview Press, Boulder, Colorado, Inc, ISBN: 0-8133-0146-7 (1985).

Williamson, O.E. 1975. Markets and Hierarchies. New York: Free Press.

Williamson.O.E. 1979. Transaction cost economics: the governance of contractual relations. The Journal of Law and Economics. 22(2): 233-261.

Williamson, O.E. 1981. The economics of organization: The transaction cost approach. American Journal of Sociology 87: 548-577.

Williamson, O.E. 1985. The economic institutions of capitalism: Firms, Markets and Relational Contract. New York: The Free Press

Williamson, O.E. 1986. Economic Organization: Firms, markets and policy control. Wheatsheaf Books Ltd ISBN 0-7450-0015-0.

Williamson, O.E. 1988. The logic of economic organization. Journal of Law, Economics, and Organization., vol. No. 4, no. 1, Spring 1988.

Williamson, O.E. 1995. Transaction Cost Economics and Organization Theory. In O.E. Williamson (ed.), Organization Theory: From Chester Barnard to the Present and Beyond., expanded edition (Oxford University Press).

Williamson, O.E. 1996. Efficiency, Power, Authority and Economic Organization. Ed. J. Groenewegen, Kluwer Academic Publishers, Boston.

Williamson, O.E. 2000. The new institutional economics: taking stock, looking ahead. Journal of Economic Literature. Vol. XXXVII (September 2000) pp 595-613

Wolz, A., Kirsch, O.C 1999. Equitization of agribusiness in Vietnam. Options for small-scale farmers with special emphasis on coffee production in Daklak Province. Discussion Paper No. Research Centre for International Agrarian and Economic Development, Heidelberg, Germany, February 1999.

[2] Note: During mid 2002 the South African Rand traded at roughly R10 against the US dollar

Previous Page Top of Page Next Page