Chapter 7. The potential for farm/non-farm linkages in the cassava subsector in Ghana - Ramatu Al-Hassan and Irene Egyir

INTRODUCTION

This chapter focuses on the cassava subsector in Ghana. The choice of cassava for a study on farm/non-farm linkages was made because of its growing importance as a crop in Ghana and because it is almost always processed in some form for sale. The need to process cassava enhances its potential as a commodity linked to the non-farm sector. In recent years, there has been an increase in the market for cassava chips for export. Although the chip market has become more uncertain in recent times, it continues to influence the market for traditionally processed cassava products and has fostered a number of linkages between the farm and non-farm sectors.

The analysis of the evolving market for cassava products in this chapter is organized as follows. In the second section, background information on farm/non-farm linkages and cassava production in Ghana is presented. In the third section, information is given about the two districts that form the basis of this study. Linkages between the cassava market and the rest of the economy are examined in the fourth section; this is done by examining each of the markets for processed cassava. The fifth section examines the recent evolution of the cassava market and how this evolution has affected rural producers and the linkages within the market. The sixth section examines the role that institutions have played in forming linkages and supporting cassava production, particularly the Government. The final section presents conclusions.

CONCEPTUAL FRAMEWORK AND BACKGROUND

The weak farm/non-farm linkages in the Ghanaian economy have constituted one of the major development issues since independence. In a report on growth and poverty in Ghana, for example, the World Bank noted the following: “Unless substantial improvements are made in the performance of agriculture, the potential for growth elsewhere, particularly in agro-industries and transport, could be restricted because of intersectoral linkages.” (World Bank, 1993). Yet there has been very little study of linkages in the Ghanaian economy; the research has to tended to focus on the macro level, using input-output analyses (Stryker and Dumenu, 1986; Jebuni, Asuming-Brempong and Fosu, 1990); these studies identify weak linkages between agriculture and the industrial sector. The Plan Consult report, for example notes that intersectoral linkages in Ghanaian agriculture amount to only 13 percent of cross-sectoral activities, and that even though agriculture employs 55 percent of the active labour force, its contribution to gross domestic product is only 45 percent. Jebuni, Asuming-Brempong and Fosu (1990) report that early input-output tables on the Ghanaian economy show that low use of intermediate inputs in the agricultural sector is a major reason for the weak production linkages between agriculture and industry. The intermediate input demand of the agricultural sector, excluding cocoa, forestry and fishing, was estimated at 5 percent of agricultural output; intermediate input use of agricultural output by the rest of the economy was estimated at 7 percent. Demand for agricultural output as an intermediate input came mainly from food and beverage processing, tobacco and textiles. Similar conclusions have been reached by other studies, which estimate that intermediate inputs, mainly seed and hand tools, account for less than 10 percent of the value of cocoa and less than 20 percent of the value of maize and rice (Stryker and Dumenu, 1986).

To assess demand linkages, Jebuni, Asuming-Brempong and Fosu (1990) estimated an econometric relationship between non-farm expenditures and an index of agricultural output, agricultural terms of trade, food terms of trade and non-farm output. Their results show that a unit increase in agricultural output induces an increase of 0.52 in demand for non-farm output; a unit change in agriculture’s terms of trade stimulates a unit change of 0.49 in non-farm consumption expenditure. Unlike the weak production linkages, these estimates suggest fairly strong demand linkages between agriculture and the non-farm sector. To examine this further, Table 1 presents data on expenditure shares by income class. In general, households spend over half of their budget on food items, although poorer households use more home-produced food than richer households. Richer households spend a smaller share on food than poorer households, but spend a greater share on non-farm items. Although a marginal budget share analysis is required to give more conclusive results of the linkage implications of the expenditure patterns, the budget profile suggests that increases in income levels will lead to greater non-farm demand.

Despite high shares of food in the budget shares of all income classes, the domestic market for agricultural output remains rather small because of the small population and the low purchasing power of domestic consumers. Nearly 50 percent of households had expenditures below the minimum wage between 1988 and 1989; as a result, the Ministry of Agriculture has sought an export-driven growth strategy for agriculture since the late 1980s (Ministry of Agriculture, 1990). Such a strategy needs to be accompanied by product development to diversify markets and increased opportunities for value-added to agricultural products, whether for domestic markets or export. The activities associated with export diversification, product development and processing are expected to create the additional demand for agricultural output and generate additional employment in the sector, thereby raising incomes and spurring further demand for agricultural and non-farm consumer goods and services.

TABLE 1
Expenditure shares by income class, 1987/88 and 1988/89

(a) 1987/88 Ghana Living Standard Surveys (GLSS) data; (B) 1988/89 GLSS data.
Source: Seini, Nyanteng and van den Boom (1997).

The cassava subsector has the potential to generate this type of growth. Cassava has for a long time been grown by smallholders as a food-security crop because of its resilience to adverse climatic conditions and its low input requirements. Production of the crop has grown steadily, mainly by area expansion through replacement of fallow and as a substitute for other crops such as yams. Average yield per hectare has improved from 7 mt in the early 1980s to about 12 mt in the 1990s. Ministry of Food and Agriculture statistics show that since 1994 Ghana has produced about 1 million tons of cassava as excess supply (Ministry of Food and Agriculture, 1996). Cassava has potential as an export commodity because of world market demand for cassava as animal feed. Cassava chip exports increased from 500 mt in 1993 to over 20 000 mt in 1997 (Trade and Commodity Group [TCG], personal communication, 1996). These figures exclude exports of gari, cassava dough and tapioca, which is granulated cassava starch, as ethnic foods to Europe and America, and through informal cross-border trade, particularly Togo. Actual exports of cassava products are likely to be much higher than reported.

Cassava has a high potential for product diversification, because it can be processed into various forms for human consumption and made into chips for animal feed; starch and starch derivatives can be extracted for industrial uses. Cassava utilization in Ghana is still mainly for human consumption, however, either as unprocessed roots or in the traditional processed forms of gari, dough and flour. The bulk and perishability of the crop mean that demand for transport is high. These processing and transport requirements suggest that cassava will have substantial forward linkages.

The cassava subsector has been increasing production over the last few decades. The present estimated yield of 12 mt/ha is still well below the achievable yields of 30 mt. Recent yield increases suggest that the cassava production system is being transformed from an extensive subsistence system to an intensive commercial system. If this is occurring, it is worthwhile determining the factors driving the change and the input demand being created from it.

THE STUDY AREA

The study was conducted in Nkwanta district in the Volta region and Atebubu district in the Brong Ahafo region, which were selected for the study because they are high-potential agricultural areas and currently involved in processing cassava chips for export as animal feed. They are therefore well placed for an assessment of the effects on the farm and non-farm sectors of an increase in demand for cassava. A brief description of the economies of the two districts is given below.

Atebubu district

Atebubu district has an area of 6 720 km2; the population increased from 50 225 in 1960 to 141 181 in 1996, at growth rates of 2 percent between 1960 and 1970, 5.4 percent between 1970 and 1984 and 4 percent between 1984 and 1995. The population is 60 percent rural; the remaining inhabitants are distributed among towns of 5 000 or more people situated along the main road linking the population centres of Yeji and Kumasi. Population density is only 21 persons per km2. The district experiences low out-migration of 14 per 1 000, but high in-migration from the north and south. Migrants from the north come mainly for farming; those from the south come to trade in agricultural produce, especially yams. The Atebubu district is in the forest/savannah transitional zone, where there is bimodal rainfall with peaks during May and June and during September and October. Mean annual rainfall ranges from 125 cm to 140 cm. Atebubu serves as a transit point between north and south; the towns of Atebubu and Yeji play important roles, although traffic through the district has diminished drastically since the construction of the Kumasi-Tamale road via Kintampo.

Soils in the district support rice, yams, cassava, maize, sorghum, groundnuts, soya beans, cowpeas, cotton, kenaf, vegetables and tobacco. The bulk of agricultural produce is sold in the production areas. The second most important sale point is the weekly district market. Market interaction outside the regions is limited, probably because the roads are bad and farmers are unable to move produce themselves; traders are able to use tractors to transport produce from buying centres.

Agriculture is the dominant sector of the economy, occupying 60 percent of the population. It is estimated that crop production generates 65 percent of the district’s income. Expenditure shares are 42 percent on food and 23 percent on farming, suggesting that the non-farm sector is small and its components classified as:

• agriculture-based: milling, brewing, fishing, gari processing and distilling;

• wood-based: carpentry and charcoal burning;

• textile-based: tailoring and dressmaking;

• clay-based: pottery;

• metal-based: ironwork.

Agriculture-based activities account for 37 percent of the non-farm sector, followed by wood-based activities with 33 percent.

The district is poorly served with infrastructure. The only major road runs through the district capital from Kumasi in the south to Yeji in the north. There is no piped-water network in the district. Five banks serve the district, including two Ghana Commercial Banks, one Agricultural Development Bank and two Rural Banks; the main beneficiaries of bank services are salaried workers and traders. Farmers tend not to save in banks, and loans granted to farmers have been difficult to recover.

Nkwanta district

The Nkwanta district is the largest district in the Volta region, with an area of 4 530 km2. The projected population in 1995 was 150 000 with a density of 33 people per km2, which is sparse compared with the regional density of 70 persons per km2. Processing of different types of cassava products is associated with ethnic groups, which include the Konkomba, Ntwumuru, Challas, Atwode, Ntrubu and Adele. The district stretches from the savannah woodlands of the north to the forest zone in the south. Mean annual rainfall is bimodal and ranges from 130 cm to 170 cm.

The various agro-ecologies in the district offer good growing conditions for a wide range of crops. The main crops grown in the savannah ochrosols include yams, cassava, maize, groundnuts, cowpeas and sorghum. The associated bottom soils in the river valleys are used for rice production. The more fertile forest ochrosols and oxysols are planted with cocoa, citrus, oilpalm, cola, plantain, bananas, cocoyams and pineapples.

The economy of the district is largely agricultural; the Department of Agricultural Extension estimates that more than 80 percent of the population is engaged in agriculture, although this estimate includes all agriculturally related activities, such as processing and marketing. These two activities, and small trading in non-food consumer items, are the major non-farm activities. Although the term non-farm is used to describe all post-harvest activities, some food processing is carried out on farms as a matter of convenience.

Nkwanta district has a very bad road system; even the main road from Jasikan through the district capital to Damako in the north is sometimes impassable during the rainy season.

Summary

The two study villages have a very small non-farm sector. The few non-farm activities are concentrated in the district capitals. In villages there are only one or two small traders with kiosks selling consumer items such as cigarettes, matches, soap and sugar. The situation urgently needs ways to create a vibrant non-farm sector, which reinforces the relevance of the present study. Both districts have relatively low population density and poor infrastructure, which have serious implications for the strength of linkages. Low population density limits the size of local markets and causes land-surplus agriculture in which per capita land availability is much higher than the threshold for transfer of labour from the farm to the non-farm sector. Poor infrastructure, especially roads, creates transport bottlenecks that will adversely affect forward linkage activities.

LINKAGES AND THE CASSAVA MARKET

This section examines the variety of linkages between cassava production and the rest of the economy. The cassava market is divided for this purpose into the various types of products made from the crop, which allows a clearer evaluation of the links between cassava and the rest of the economy and is a mechanism for comparing linkages across products. The information gathered for this analysis is based on interviews conducted in October and November 1997 in the cassava-producing regions of Atebubu and Nkwanta with informants including farmers, processors, traders and buying agents, transporters and officials of the Ministry of Agriculture and district assemblies, who were asked specific questions about cassava production and the market for cassava products. Discussion of cassava product markets is preceded here by some general information gathered from the interviews.

Cassava has only recently emerged as an economically important crop in the study districts. For a long time, yams were the staple crop in these regions as well as the major cash crop; cassava was intercropped with yams. In Atebubu, cassava was left for the women to harvest after the yams had been harvested. The economic status of cassava in the two districts began to change after the 1982-1983 drought and the ensuing famine. In the Nkwanta district before the drought and famine, ethnic groups in the southern part of the district processed cassava into gari, roasted granulated cassava. The ethnic groups in the north processed cassava into kokonte, dry cassava chips or flour with informants.

Cassava is primarily produced by small-scale farmers on fields of up to 0.8 ha. In Nkwanta district, smallholders constitute 70 percent to 80 percent of farmers. Farms with 6.0 ha and above are considered large-scale and account for only 5 percent of farmers. In Atebubu, average holdings are about 2 ha; two thirds are under cassava. Holdings of 8.0 ha and above are considered large-scale, accounting for about 10 percent of farmers. Cassava is normally intercropped with yams, maize or beans, but the dominant mixture is yams and cassava. There is little difference in cropping systems or technology levels between small-scale and large-scale farmers. Yam mounds are prepared from October to December; the yams are planted by February or March. Cassava is planted when the yams have taken. Yams are harvested after four to six months and the cassava is left to ripen. Depending on the variety, cassava can be harvested after 6 to 12 months. The very late varieties can take between 18 and 24 months to ripen. Ripe cassava can normally be stored in the ground for several months and harvested as needed, a feature that enables cassava to play an important role in household food security.

Farming in Nkwanta is based on minimal technology; hoes and cutlasses are the main farm implements. There is just one serviceable tractor in the district. Animal traction for ploughing and transport was introduced by the International Fund for Agricultural Development (IFAD) Smallholder Credit Input and Marketing Project in 1994; as of November 1997, there were three sets of animal-traction technology working. Although tractor ploughing is more common in Atebubu, cassava fields are not ploughed because ploughing requires complete clearing of the land, which exposes cassava to the elements. Fertilizers are little used, and certainly not on cassava. It is claimed that soils in both districts are rich; crops do well without fertilizer. No pesticides are used on cassava and farmers are encouraged by extension personnel to use pest-free and disease-free planting material. Weeding is done by hand, although in Nkwanta it was claimed that one or two farmers use herbicides. The organization of cassava production thus presents few opportunities for backward linkages. Lack of demand for inputs such as mechanized equipment, fertilizer and pesticides means that little business activity can currently be generated in these areas to serve the cassava economy. Although hoes and cutlasses are in high demand in this low technology farming system, the demand generated for them cannot be attributed to cassava alone.

The most important purchased input in cassava production is hired labour for land preparation - clearing and mounding - weeding, harvesting and processing. The labour requirement for cassava production using traditional technology is estimated at 35 person-days/ha. The 17 000 ha cassava area in Atebubu creates about 600 000 person days of employment, about 75 percent of which is hired. Hired labour in Atebubu is mainly migrant labour from northern Ghana; in Nkwanta the hired labour comes mainly from Benin. Even for hired labour, therefore, the employment created does not directly benefit local inhabitants. It is not clear whether there is an indirect benefit to the local economy from the income earned by migrants, because seasonal migrant labourers return home with earnings for their own farming activities. The local economies would benefit if migrant labourers purchased farm implements or consumption goods while they were in the region.

Cassava processing offers the best opportunity for linkages from the farm sector to the non-farm sector. The description of the organization of cassava processing reveals the extent to which this potential for farm/non-farm linkages is being realized in the study area. Cassava products may be classified as traditional or modern. Traditional products, such as gari, kokonte, cassava dough and starch, are produced either for subsistence or for sale in local markets. Modern products, such as cassava chips, are produced for sale to processors for commercial purposes. The dominant traditional cassava products in the study area are kokonte in Atebubu and gari in Nkwanta. The major modern product in both districts is cassava chips for animal feed. Cassava chips are a new product, produced solely for one company. The different types of product markets and the linkages within those markets are now discussed.

Kokonte processing in Atebubu

Kokonte is only processed on-farm as an individual activity. Labour for processing is hired on a share-production basis whereby farmers hire labourers to uproot and process the crop; the product is then shared equally between farmers and labourers. There is little other productive activity generated from processing kokonte, because the technology is labour-intensive, based on knives and fuelwood. There is, however, considerable potential for employment generation locally because the labour hired for processing kokonte is local youth, mainly men, who form themselves into gangs. Some work gangs may include two or three women, who help to headload and peel the cassava. The young men who sell their labour for kokonte processing have small farms and therefore spare time to earn income from other sources. Women are involved more in the home processing of another type of kokonte called chorchor. The cassava is peeled at home, crushed in a mortar and dried in the sun. This type of processing is done on a very small scale at a rate of about one bag per week, equivalent to half a tractor load of smoke-dried kokonte for on-farm processing.

The marketing of kokonte involves a network of buying agents, handlers, transporters and traders. The role of each of these actors is set out below, but it is not easy to assess the number involved in each of these activities or the wages paid.

The buying agent is the link between farmer and trader, who is usually from a large town or city. Each agent has a number of regular traders for whom they identify sources of kokonte. With the help of handlers, agents provide bagging services for the trader. Handlers are hired on a casual basis; the number hired depends on the quantity of cassava to be bagged. When a kokonte buyer is identified, the farmer arranges for a tractor to transport the product from the farm to the house for bagging. The tractor driver is helped by workers; they assist the farm labourers, who processed the kokonte, to load the loose product onto the trailer. The charge for the tractor includes the services of the tractor workers. One load is between 40 and 50 bags; there may be up to ten trips per farm. The trader pays only after the product has been bagged. Traders operate in groups of five or six, and buy the product weekly or fortnightly. The duration of one buying trip depends on the availability of kokonte. It is understood that some traders are agents themselves, working for others in big towns or cities.

The next stage in the marketing chain is transporting the product out of the district. Depending on the volume, the trader may arrange for a truck for long-distance haulage, or another tractor to take the product to the market in the district capital to be transferred to a bigger truck. The trucks are served by handlers called loading boys. Weekly haulage of kokonte from the district market ranges between two and five trucks with capacities of 7-10 mt or between 120 and 180 bags. One trader is capable of purchasing one truckload of the product at a time.

As explained earlier, the proceeds from the kokonte are shared equally between the farmer and the gang of labourers. The value shared is the net return after deducting the cost of tractor services. The tractor and truck services and handlers are supplied through the local branch of the Ghana Private Road Transport Union (GPRTU), which sets prices for these services. Transport charges are paid by the bag and depend on the distance covered. The rates for tractor haulage varied considerably as stated by GPRTU informants and farmer and agent informants. The GPRTU informants quoted rates between 400 Cedis and 1 000 Cedis per bag; farmers and agents quoted rates between 1 500 Cedis and 3 000 Cedis. The charge for bagging the produce is 500 Cedis per bag; the fee for loading sacks of kokonte onto trucks is 200 Cedis per bag; both are paid by the trader, who also pays the cost of trucking out of the district.

There are often special financial arrangements between farmers, traders and agents. A trader may provide a loan of between 100 000 Cedis and 200 000 Cedis to a farmer against the supply of kokonte. The amount of a loan depends on the quantity of kokonte a farmer expects to process. The agent assures the buyer that the farmer has cassava for processing. Another type of financial arrangement is a farmer providing kokonte on credit to a trader, who can take delivery of the kokonte and pay on the next purchasing trip, which may be in a week or two. The maximum of such credit is usually 55 bags or a full tractor load. It is usually large-scale farmers who engage in these financial transactions. A large-scale farmer receiving a loan has a better chance of guaranteeing it with available cassava; a large-scale farmer is more likely to be able to supply cassava on credit as a loan. Farmers who can provide this kind of credit establish regular demand for their product and can protect themselves against falling prices. These financial arrangements are sufficient provided the scale of processing remains at the current level. Any increase in scale must come through adoption of improved processing technologies, which will require higher levels of financing that traders alone may not be able satisfy.

A number of farmers indicated that income from kokonte is used principally to pay for land preparation in yam and cassava fields. Farmers are thus able to increase production of yams, which is the main staple and cash crop. The first of two peaks for kokonte processing in November to March reflects farmers’ need for cash to pay for land preparation. The second peak in June to August is a period when there is less farm activity.

Gari processing in Nkwanta

Gari processing is an individual income-generating activity that relies solely on family labour. Unlike kokonte processors, gari processors tend to buy cassava to add to their own production. Farmers sell cassava by fields of 0.4 ha. The price depends on the ripeness of the crop, but is linked to the price of a 50 kg bag. There are no buying agents in this trade. Although farmers with bigger fields sell cassava, farmers tend not to have a preference as to the size of processor they sell to. Farmers supply cassava to processors, who pay after the processed product has been sold. The longest repayment period is normally two weeks. The financial arrangements between farmers and processors are based on trust, developed through long-standing relationships. The cassava is purchased from local farms at distances of up to 10 km.

Gari processors do not hire labour, even for the harvesting. There is no labour supply, because people prefer to engage in production for themselves, no matter how small the scale. Processors call on relatives to assist them if they need more hands. One person working alone is able to process 50 kg of gari per week; with two workers, 50 kg can be processed in two days.

Grating is the only part of the process that is mechanized; a few processors use mechanical presses to remove the liquid from the grated cassava. Processors prefer mechanized grating because it is faster and produces a finer product. Manual processing of one unit of cassava field would take ten days; it can be grated in a day using a mechanical grater. Mechanical graters are normally owned by an individual who provides the grating service for a fee. Some farmers now use mobile powered graters, which permit them to process on farms. In such cases, the processors may hire workers to help with the roasting. This latest system of processing and the related linkages need to be investigated in more detail.

Although processors can sell to retailers and wholesalers, their main customers are traders from big towns and cities such as Accra, Ho and Nkwanta. There are also buyers from Burkina Faso and Niger. Processors also sell in markets at the district capital by carrying on their heads enough to meet current cash needs. The buying season for traders from outside Ghana is from January to July. Traders from towns in Ghana buy less during the rainy season, when driving on the main road to the district becomes difficult. The price of gari varies according to the availability of other foods: prices are relatively high in July and August; they are very high in October to early December.

TABLE 2
Rates of payment to actors in the gari trade

Source: Author interviews.

The system of gari marketing in Nkwanta also uses agents, who identify sources of the product and buy and package it for export out of the district. A trader advances money to the agent based on the current market price. Traders often advance money to processors to guarantee supplies. Price variations do not affect the financing contracts. There are sometimes barter arrangements whereby traders provide processors with non-farm consumer items such as clothes in return for gari. Conflicts can arise from these contracts when processors default. Agents act as intermediaries between the contracting parties, disbursing funds and ensuring that traders get the product within the specified period.

After buying gari, agents engage handlers to bag the product and sew the bags. Transporters move the product into temporary storage. Agents sometimes provide storage space; otherwise, there are property owners with spare rooms who provide the service. Payments to different actors in the processed cassava business are shown in Table 2. All payments are piece-rate and depend on the number of bags, irrespective of time taken. Agents prefer to deal with foreign traders from the West Africa subregion, because they make prompt payment for services; the larger volumes purchased by foreign traders create more business.

Chip processing in Atebubu

The TCG, the sole exporter of cassava chips, is responsible for organizing farmers to produce chips. About 4 000 farmers supply chips to the company; of these, 2 500 are considered regular and reliable suppliers who have taken up cassava chip production as a regular source of income. These farmers do not buy cassava to process; the company does not encourage farmers to buy cassava for chipping, because it increases costs for the farmer and introduces a risk of irregular supplies to the company. The farmers include small-scale and large-scale farmers with farms of between 1.2 ha and 16 ha. In terms of chip production in a season, farmers can be grouped as shown in Table 3; some farmers in the large-scale category can chip up to 900 bags per season.

TABLE 3
Chip output by farmer category

Source: TCG area office, Atebubu.

Chip production is an individual income-generating activity. Farmers may use family labour or hire labour for harvesting, but they usually do not hire labour for chipping; some large-scale farmers may hire labour for chipping on share terms similar to those for processing kokonte. There is no use of reciprocal labour in chipping; the chippers’ association in Atebubu district does not engage in communal chip production.

Chips are processed manually and by machine. TCG has 31 chipping machines, although it is estimated that 90 percent of farmers do manual chipping. Most farmers prefer to chip by hand because the chips are easier to dry and hand chipping gives higher yields. The chipping machine tends to produce small chips, which require more careful drying and become powdery more easily. Only large-scale farmers use the chipping machines; more farmers opt for machine chipping at the beginning of the farming season, when demand for farm labour is at a peak. Processing kokonte is hard work, which is why farmers hire labour to do it. Processing chips is much easier and can be done casually by farmers in their free time. An estimated 40 percent of chip processors are women. Before the advent of the cassava chip market, men donated their cassava fields to women to process the crop into kokonte. Now that cassava has become valuable and cassava is fully integrated into the cropping system, men’s attitudes have changed with regard to disposal of their crop; the result is that women themselves are growing it. There is no problem of access to land; the only problem that farmers face is the high cost of labour.

Farmers tend to grow only varieties with high dry-matter content for chip-ping. Most of the local cassava varieties ripen in a year, but there is one that ripens later and can remain unharvested for about three years. This variety gives farmers the flexibility of chipping at any time during the year. It was only during the 1997 cropping season that one of the varieties released by the Crops Research Institute was introduced for testing by a few farmers in the district.

TABLE 4
Rates charged for hauling chips

Source: TCG Area Office, Atebubu.

The export company is the sole buyer of chips. The company organizes processing areas into buying districts, made up of zones and each zone has a buying centre. There are four buying centres, comprising 16 zones. In the Atebubu area, the cassava chip-buying district includes the administrative districts of Atebubu and Sena west.

Cassava chips are normally processed on-farm, as with kokonte. If rains set in when farmers are chipping, however, the cassava is transported home or sent to company warehouses for chipping. In such cases, the company provides transport at a cost to the farmer shown in Table 4. Although fresh cassava is transported, the rate charged depends on the volume of chips produced; farmers face the same transport cost whether they processes on-farm or off-farm. There is therefore an element of subsidy to the farmer who processes at home.

Linkages to the local economy for chipping are limited. Farmers hire labour for land preparation, mounding, weeding and harvesting of cassava. But much of this labour is seasonal migrant labour. TCG creates employment for local people as agents, loaders and tractor operators. Little business activity has emerged to serve farmers or processors. The main types of business activities are tractor hire and transport. The chip export company provides subsidized tractor and transport services. The company hires tractors and 7-ton trucks for hauling chips from farms to weighing centres.

Many farmers suggested they had increased expenditure on consumer durables and farm implements from cassava chip earnings. TCG realized this, and provides some of these items to farmers on credit against delivery of chips, including cutlasses, hoes, rubber boots, cement, roofing sheets, bicycles and iodized salt. Some farmers have requested television sets, radios and sewing machines.

Yam production is increasing, because the higher earnings from cassava are reinvested in bigger yam fields, which means interplanting with more cassava. About three chippers have purchased pickups to provide transport between Atebubu and Kumasi. Farmers are also investing more in children’s education; many cited their ability to pay school fees and other education-related costs as a benefit of processing chips.

Before the introduction of chips, kokonte was the major cassava product in the region. Estimates from various informants suggest that between half and two thirds of annual cassava production is now processed into chips. Although kokonte is strictly for human consumption, cassava is sometimes sold to the chip market when the kokonte market is down or when farmers need cash in lump sums. The major attraction for chips is these lump sum payments. It appears that the chip boom is experiencing a lull, however, because cash-flow problems have limited the ability of the company to make timely payments. Maintaining farmers’ interest in processing chips will depend on the ability of the company to avoid delays in payments.

It is generally believed that cassava production is increasing because of the alternative market for chips. Large-scale farmers switch between processing chips and kokonte, with the choice depending on:

• the processing preferences of hired labour, because hired labour is paid in kind;

• the weather, because it is easier to process kokonte in the rainy season;

• farmer’s need for cash at the harvest time;

• distance between farms and villages, which determines transportation costs paid by the farmer; it is more expensive to transport kokonte from distant farms.

Although few farmers plant cassava only, the density of cassava in yam fields is increasing. Income from cassava is increasing because of increased output and better prices. The chip market has helped to boost kokonte prices, because chips and kokonte have become competing alternatives, putting pressure on kokonte traders to improve prices. During the 1996/97 season, for example, chips were bought at 4 500 Cedis per bag and kokonte at 12 000 Cedis per bag. By the 1997/98 season, the price of chips was 5 000 Cedis while kokonte was already selling for 20 000 Cedis per bag. This trend is another threat to the viability of the chip industry: farmers are beginning to ask for higher chip prices because of the relative increases in the price of kokonte. Traders in kokonte are also experiencing higher transaction costs for each purchasing trip, because it takes more days to organize a trip now than before farmers went into chip processing.

Cassava chip processing in Nkwanta district

Although the organization of the cassava-chip market in the two districts is similar in many respects, some significant differences arise from the alternative uses that farmers have for their cassava.

Participation in chip production in Nkwanta is much lower, estimated at about 400 core suppliers, most from the northern and eastern parts of the district. The south of the district is dedicated to gari processing, and farmers are still nervous about the chip market. As in Atebubu, cassava chipping is an individual activity, using cassava from people’s own farms. About 15 percent of the major chip producers are women, and numbers are increasing. Processors may be classified as small-scale, producing from 1 kg to 500 kg per season, medium-scale, producing from 500 kg to 2 500 kg, and large-scale, producing over 2 500 kg.

Farmers are advised to harvest cassava at 9 to 12 months, because this creates the high dry-matter content. Technologies for chipping include manual and mechanized chipping. The company provides mechanized chipping services at a small fee, which is reflected in a lower price than for manually produced chips. In the 1996 season, for example, the cost of mechanized chipping was 12 percent of the price of 4 500 Cedis per 50 kg. As in Atebubu, farmers prefer hand chipping, but for different reasons. In Nkwanta, farmers assist each other with labour through the cassava chippers’ associations. Labour used for chipping is family labour or communal/exchange labour. There is therefore no employment created for chip processing apart from the company’s own labour force. The dominance of hand chipping also limits any direct production linkages. Chippers require no services that the non-farm sector can provide.

Chip processors tend to spend their earnings on roofing sheets, child education and bicycles. Many more people are building houses, although they are still made of mud. There is also conspicuous social spending on such events as lavish funerals at which abundant food and beverages are served.

The major problem facing the chip industry now is the company’s lack of funds to purchase chips, especially during the peak season. A growing demand for gari, resulting from an influx of traders from the West African subregion, and a cultural preference for gari in the southern parts of the district present a challenge for the buying company in terms of improving its cash flow. Gari processing is increasing, and in 1997 the price was fairly stable. This trend is probably a result of the export marketing of gari that started sometime between 1996 and 1997 and caused the decline in chip production in Nkwanta district, as reported by TCG. Although farmers say that low chip prices are the disincentive, the difficulty of the work and the costs of processing chips and gari are not the same. Requirements for labour and other inputs for gari are high, but do not appear to worry farmers. Farmers nevertheless claim that processing chips is a better alternative when cassava begins to rot, because while gari processing can only be done in small amounts, a farmer can uproot a complete field of cassava and chip it all at once.

RURAL PRODUCERS, LINKAGES AND THE EVOLUTION OF THE CASSAVA MARKET

As noted in the previous section, the cassava market has evolved considerably in recent years. In this section, the prospects for expansion of cassava chips exports and the effect of changes in the market on farm/non-farm linkages are considered.

As in most of Africa, cassava utilization in Ghana is still primarily for human consumption, either as unprocessed roots or in the form of traditional cassava products - gari, dough and kokonte. The expected negative effect of urbanization on demand for some traditional cassava products can be overcome with adaptations of the products to suit the needs of urban consumers. The Food Research Institute has, for example, developed a cassava flour whose main advantage is reduced meal-preparation time. Cassava flour can be used as a wheat flour substitute for the baking industry (Day et al., 1996). Even with these developments, external markets will be necessary if a substantial increase in cassava demand is to be achieved. In addition to exports of small quantities of traditional products, Ghana began to export cassava chips for livestock to Europe in 1994; earnings from chip exports recorded a dramatic increase from US$284 000 in 1994 to over US$2 million in 1997, followed by an equally dramatic fall to US$208 000 in 1998. The new export avenue for cassava presents opportunities for agro-industrial development, but access and responsiveness of rural producers to the new market opportunities are significant factors that set the pace of agro-industrial development. In the case of cassava, where the export product faces competition from processed traditional products, response to export demand will depend on the level of incentives derived from this market compared with those derived from processing for traditional markets. To understand these incentives, we need to investigate how rural producers fit into the cassava market.

Cassava farmers in the two districts prefer to process the crop themselves rather than sell the fresh root. In both districts, over 98 percent of fresh cassava processed by farmers is from their own farms; purchased fresh roots represent less than 20 percent of fresh cassava processed. These results are consistent with data from the Collaborative Study of Cassava in Africa (COSCA), which suggests that producers in remote areas such as the study districts process a high proportion of their produce because of the bulkiness of fresh root (Nweke, 1996). Farmers do not specialize in cassava processing because:

• they can transfer farm labour during the off-season to the non-farm sector, a strategy essential for overcoming problems of labour redundancy that characterizes seasonality in farm production;

• processing improves cassava income and stabilizes overall income because of improved shelf-life;

• farmers lack confidence in the reliability of cassava supplies from other farmers; this also occurs because of variations in supplies as farmers respond to annual price variations.

Because they do not specialize in processing, many small-scale processors widely dispersed across the producing areas use household labour and traditional technologies. The small size and dispersal of processors in turn require an efficient network to assemble adequate volumes.

The marketing chain for chips in Ghana is short, with the exporter dealing directly with the thousands of small-scale processors. In order to ensure assembly of exportable volumes of quality chips, exporters use an extensive network of agents and complicated logistics. To ensure stable supplies, exporters give farmers incentives such as credit for consumer items and subsidized transport between farm and local assembly depots. As the company reduces explicit cash payments to farmers, however, its own costs of assembly, transport to port and associated administrative costs account for 25 percent of cost of export brokerage (Table 5).

The scattered distribution of producers and the poor rural road networks account for the high transport costs. It is estimated that while production costs in cassava producing areas of Brong Ahafo are about 48 percent lower than costs in Greater Accra Region, the cost of transporting chips from Brong Ahafo to the port of Tema is 80 percent more than the costs to transport chips from Greater Accra to Tema (TechnoServe Ghana, 1994). Dadson, Kwadzo and Baah (1994) have also shown that in general, the competitiveness enjoyed at the farm gate by root and tuber crops is lost at the level of urban market because of high transportation costs.

TABLE 5
Component costs per ton, tcg (1998)

¹ FOB: free on board.
Source: TCG area office, Accra.

Freight charges from Ghana are not globally competitive, ranging between US$35/mt and US$40/mt, compared with only US$9/mt from Thailand (David Pessey, TCG, personal communication, 1996). The high cost of freight has been attributed to inefficient handling at the port in Ghana. The loading rate at the port of Tema is 600 mt per day, compared with 20 000 mt achieved by the competition. The FOB price received by Ghanaian exporters is therefore far below the average world price; the difference was US$47/mt in 1996 and US$11/mt in 1997. In addition to a volatile average world price, the price received for Ghana’s chips has declined steadily from an average of US$120/mt between 1993 and 1994 to US$86/mt in 1997. The effect of low FOB prices and high costs of haulage are a price squeeze on producers, which affects the relative profitability of processing chips for exports.

Returns from processing various cassava products are used as an indicator of the relative incentive to process each product. The budgets for cassava production and processing of gari, kokonte and chips are estimated under two scenarios: i) farmers process their own cassava with hired labour, and ii) farmers process their own cassava with family labour. The latter scenario is based on the fact that farmers in the two districts generally process their own cassava without hired labour. Since farmers process their own cassava, fresh cassava is valued at the variable cost of production. Tables 6 and 7 show the results of these scenarios.

TABLE 6
Scenario 1: Farmers process their own cassava using hired labour

Source: Author’s calculations.

Net earnings from cassava processing are estimated at 31 000 Cedis for chips, giving approximately 60 percent return on investment, 40 000 Cedis for kokonte, giving 23 percent, and 122 000 Cedis for gari, giving 32 percent (Table 6). There are no previous estimates of returns on investment for kokonte or chip processing. The estimate of 32 percent for gari is higher than the 18 percent estimated by Agyako-Mensah (1985) but comparable to the 27 percent margin for Kumasi processors in Kreamer (1986). The difference in results is because of the valuation of fresh cassava in this study at variable cost rather than at the market price. The higher performance of chips over gari and kokonte is because of the lower cost of hired labour, which for all three products is costed at a third of the value of the final product.

The second scenario of farmers using their own labour is to demonstrate the returns to family labour engaged in processing (Table 7). These returns are uniformly higher than the returns using hired labour for all three products. The exceptionally high percentage margins, however, only illustrate the high family-labour component of total costs. The use of family labour is common in gari processing in Nkwanta, where earnings per mt of product are 1.2 times higher when family labour is used than when labour is hired. The difference between payments to hired labour and net returns to family labour is 41 percent in Atebubu, which is probably not high enough to entice family labour into the tedious task of kokonte processing; hence the widespread use of hired labour for processing kokonte.

The budget estimates demonstrate the low level of incentive for processing chips compared with gari and kokonte. Net earnings per day of work by family labour are lowest for chips. Farmers complained that the price of chips is too low and noted that the only reason they process chips was because of the possibility of receiving payment for their product in bulk, a payment structure which allows farmers to process chips to meet specific cash needs. Farmers admit, however, that the good prices they receive for the traditional products are a result of the introduction of chips.

TABLE 7
Scenario 2: Farmers process their own cassava using family labour

Source: Author’s calculations.

The structure and relative profitability of chip production and the market trends lead to identification of the following challenges for the expansion of chip exports, categorized as farm level, national and international constraints:

• Farm level. The price of chips and the profitability of chip production are lower than those of traditional cassava products, which have the added attraction of being food staples for farmers. Farmers are not familiar with the use of chips as animal feed. Small producers scattered over a wide area with poor roads and limited transport increase the costs of primary haulage. Although the use of a chipping machine can reduce labour input and contribute to production of chips of a standard size, it has not been taken up by farmers, who claim that the machine tends to slice the cassava too thinly, making the drying process more tedious. The machine tends to mash the cassava, thereby reducing potential yield of chips. These are technical problems, which ought to be addressed to make the technology more appealing. More important underlying reasons for the unattractiveness of machine chipping, however, are the small volumes produced and the fact that chip processing occurs during the slack labour period and poses no competition with farming activities.

• National level. High freight charges reduce the competitiveness of Ghana’s chip exports. High interest on borrowing, about 30 percent in 1998, limits capital mobilization. During fieldwork in 1997, for example, it was clear that cash-flow limitations on export brokers had caused frustrating delays in paying farmers. Insufficient domestic demand for chips is another factor limiting expansion of the market. Despite a wide range of local research on the use of cassava as an energy source in the feeding of poultry, pigs and small ruminants, maize remains the main source of energy in feed compounds. None of the 27 members of the Ghana Feed-Millers Association uses cassava as an ingredient. The reasons include lack of knowledge on the use of cassava as feed, additional costs to be incurred in adapting mills to cope with the larger bulk volume and dust emissions from milling cassava chips. With an annual output of only 400 mt per annum, the feed-mill industry cannot have a substantial impact on cassava demand. Finally, cassava utilization is highly dependent on the availability of other major staples: the higher the level of food availability, the lower the demand for cassava. Levels and stability of export supplies will therefore depend on the levels and stability of preferred staples, especially maize.

• International level. The cassava chip market has been in decline over the last decade. World exports declined from 11.9 million mt in 1989 to 3.3 million mt in 1997, driven largely by a decline in European imports. The Netherlands reduced imports from 2.5 million mt in 1980 to 1.6 million mt in 1989, and to about 1 million mt in 1997. Germany’s imports declined from 1.3 million mt in 1989 to about 148 000 mt in 1997. The decline in demand for cassava chips in Europe has been attributed to the level of intervention prices for feed grains in the European Union, which favours the use of domestic feed grains (Henry and Gottret, 1995). Ghana’s entry into the international chip market can be described as untimely, coming at a moment when the market was declining. As of 1996, the country was hoping to access a General Agreement on Tariffs and Trade (GATT) quota of 140 000 mt, but the will to participate in this market has declined.

The current state of the cassava chip market and the problems with Ghana’s international competitiveness in this market have the potential to alter the cassava market radically. Even now, there is a strong incentive for many farmers to produce traditional processed products instead of chips. If the export market continues to falter, this trend is likely to continue. Collapse of the export market could lead to a dramatic reduction in demand for cassava and a reduction in the price for all processed cassava products. This will limit the beneficial linkages between cassava production and processing and the employment market, erode incomes and reduce expenditure linkages.

INSTITUTIONAL SUPPORT AND THE CASSAVA SUBSECTOR

Institutional support for the cassava subsector comes from a number of sources, including the public sector, research centres and the private sector. Support to the sector includes identification of markets, development of new products and development and promotion of processing equipment. In this section, the support received by the sector and ways in which it could be improved are considered.

The Government’s declaration of 1995 as the Year of Cassava reflects growing interest in the crop at official levels. Following this, a 15-member multi-disciplinary Cassava Task Force was set up in March 1996 by the Ministry of Food and Agriculture to oversee development of the crop. The group’s terms of reference included identification of appropriate research for improving production and processing, and promotion of the crop to local and foreign investors. At a workshop on cassava processing held in September 1997, it was reported that US$11 million was being made available by IFAD to support development of roots and tubers, including cassava. It is expected that much of the funding will be devoted to multiplication of improved planting material. This support follows similar funding provided by IFAD and the Government since the late 1980s through the Root and Tuber Development Programme.

The Crops Research Institute released three cassava varieties in 1990, which have had varying degrees of success across the country. Adoption of the varieties has been more successful in areas where cassava is processed, for example the Volta region, than in areas where cassava roots are cooked without processing, such as the Ashanti region. The link between productivity increases and adoption of improved production technologies on the one hand and strengthening of demand for processing on the other have been reported in southern Brazil and Thailand, where growing demand for roots to process into starch induced farmers to adopt new production technologies (Henry and Gottret, 1995). Although these links are more downstream than upstream, it is expected that in areas where improved varieties are adopted, crop management practices may have to change if the yield potential of the new varieties is to be realized. These changes in crop management practices may bring about the necessary linkage activities in the non-farm sector.

It appears that the rate of spread of new varieties is rather slow. Trials for the introduction of new cassava varieties in the two districts of this study began only in the 1997 cropping season. These are both major cassava-producing districts, and it could be expected that such areas would be among the first to receive new production technologies.

The Food Resources Institute has developed fortified gari as a convenience food, and dry cassava flours as substitutes for bulky fresh roots and wet pastes. These products are primarily targeted to urban consumers. The products have not, however, gone beyond exhibition at fairs such as the Industry and Technology Fair. The reasons why businesses have not taken up these products should provide leads on the appropriate strategies for product development and promotion. The experience of the Food Resources Institute with uptake of these products by the private sector suggests that individuals are only interested in the distribution aspect, especially export as ethnic food, but not in the processing aspect. The major constraint has been lack of financing to acquire the expensive processing equipment needed for commercial processing. The Food Resources Institute has also identified the absence of an industrial extension as a limitation on adoption of new processed products by the private sector.

The Natural Resources Institute has developed a process for storing fresh cassava, which has a good market in urban areas. They have also studied the market for kokonte and are following this up with technologies to improve the quality of the product and assessment of consumer preferences and the implications of the technology on costs.

Another important development initiative from the Government, which will benefit cassava production and marketing, is the Village Infrastructure Project. An important component of the project is development of village-to-farm tracks, to be complemented by the introduction of intermediate means of transport to replace carrying loads on the head. Such a development can facilitate movement of cassava and help to strengthen the linkages between cassava farming and the non-farm sector.

Recent private-sector entrants into the cassava industry are Glucoset Ghana Ltd, which has established a cassava plantation for processing starch, and TCG, the sole exporter of cassava chips as animal feed to the European Union. TCG has expanded operations from Nkwanta district in the Volta region to other parts of the country, including the Brong Ahafo, Ashanti and Northern regions. Exports have expanded from 30 mt in 1993 to more than 20 000 mt in 1996. The company is developing a partnership with farmers through incentives such as credit for consumer items and subsidized transport as a way of sustaining farmers’ interest in chip processing. As the company tries to cut down on cash payments for transportation by farmers, however, the high cost of transporting chips from producing centres to ports is one of its major constraints. Regions with a comparative advantage in the production of cassava chips lose the advantage by the time the product reaches the port, because transport costs are so high.

The Sasakawa Africa Foundation is disseminating cassava-processing equipment, especially to women’s groups. The equipment includes powered and manual equipment for processing chips and gari. The chipping machine, however, produces chips that are too thin for the export market. The gari processing equipment is more promising; it is now being manufactured by the Intermediate Technology Transfer Unit (ITTU) in Ho, the Volta regional capital. The portable set of equipment, which includes a grater, press, bagging stand, fermentation rack and sifter, is driven by a petrol engine. Capital requirements are high, however: unit cost of the machine as of November 1997 was 2.5 million Cedis, with a 700 000 Cedi import component on the engine. Apart from the engine, which is imported, all parts of the machine can be serviced by any technician. The machine can grate 1.2 mt of cassava per hour and has a fuel consumption of 4.5 litres per mt. It requires two people to operate; it can be assembled in 12 days by three people working approximately six hours a day. So far, three sets of equipment have been manufactured; two have been sold. ITTU aims to interest manufacturers of food processing equipment to take up commercial production of the equipment through training workshops.

Lakai Motor Company Ltd. developed a cassava chipper with the capacity to chip 46 mt of cassava per day. Initial sales were made to TCG, who provided services to farmers for a fee. Uptake of the chipping machine has been very slow; this was later revealed in the field to be because of lack of patronage by farmers.

CONCLUSIONS

Based on the information presented in this chapter, the following conclusions can be drawn:

1. Cassava production in Africa is not backward-linkage friendly. The technology applied in production does not generate demand for inputs other than labour. The employment created for fieldwork has limited linkage benefits for the local economy because of repatriation of earnings by migrant labour. The type of technology applied in cassava production is consistent with the requirements of the crop. Mechanized land preparation exposes the crop to the elements; fertilizers are not used because the soils are fertile. Cassava is noted for its adaptability to a wide range of growing conditions. Farmers are encouraged by extension agents to use clean planting material to prevent disease and pest infestation rather than to control them with chemicals. The conclusion is that promotion of low external input use as a strategy for sustainable agricultural production limits the potential of the farm sector to link up with the non-farm sector.

2. Forward production linkages from the farm begin with transport of produce to markets or processing centres. In Atebubu, marketing of fresh cassava is virtually non-existent. This is a deliberate strategy to reduce the bulk of the produce and make it more easily transportable. The poor condition of roads to production areas limits transport to a small number of tractors. In Nkwanta, where fresh cassava is sold to processors, the strategy is for farmers to shift the burden of transport to the processor by selling fields of ripe cassava instead of harvested produce. The processors harvest small amounts at a time according to batch size. The small amounts harvested are easily transported by trucks or headloaded to the processing centre, which is usually the processor’s home. The reported use of mobile powered graters on farms is an even more advanced response to the poor road conditions. Poor road infrastructure in the districts therefore limits forward linkage activities in transporting fresh cassava, and supports the need for infrastructure development as a prerequisite for rural growth.

3. One traditional and one non-traditional processed product are identified in each of the districts. Processing of both products generates employment for local residents, particularly during slack periods in the cropping season. Gari processing has direct production linkages, in the form of grating services. The size of this service sector is limited, however, by the scale of processing undertaken. Reliance on family labour for such a labour-intensive activity inevitably reduces the scale of processing.

4. Marketing of chips and traditional products generates additional employment primarily through handling and transport. Productive activities related to packaging are non-existent, because packaging materials are imported. The relative earnings of local agents and handlers in the marketing chain are probably much lower than earnings of traders and companies. The major inputs for transport - vehicles, fuel and spare parts - are imports; at best they are supplied from the major towns in the district. All these situations represent leakages from the local economy.

5. The propensity to spend on consumer and investment items from cassava earnings appears to be strong, especially when farmers receive bulk payments for sales of processed products. Investment expenditure on employing more labour to expand farms promotes growth of the farm sector; expenditure on children’s education promotes the development of human resources. Indications of consumption expenditure items, however, suggest that consumption preferences favour imports and may not be conducive to local growth.

6. Introduction of the chip market into the two districts and the export market for gari in Nkwanta have boosted prices of cassava products, and probably real farm incomes. The response to these new opportunities will determine whether the linkages between the farm and the rest of the rural economy will be strengthened. The response will depend largely on the level of support for local economies in terms of improved roads. With respect to labour-saving techniques, detailed assessment of farmers’ needs and preferences should be undertaken to guide any further research into technology development.

7. The chip market has increased demand for cassava and boosted the price of all cassava products, but weaknesses in the market for cassava chips, partially because of local conditions in Ghana, have the potential to reduce the importance of this market and thus reduce demand for cassava products. Such an occurrence could hurt farmers and those linked to the cassava market. To avoid this outcome, the government must consider actions such as improving the transport infrastructure and reducing freight costs, which would assist the competitiveness of Ghana’s cassava subsector.

8. The public and private sectors are giving increasing attention to the cassava subsector. Government programmes aim to improve productivity and production, but private-sector initiative is expanding demand sources. These strategies can complement each other if the bottlenecks resulting from poor roads are improved and processors and exporters gain better access to funding sources.

REFERENCES

Agyako-Mensah, N.A. 1985. The impact of small-scale gari processing industry on agricultural production in the Ashanti region. Kumasi, University of Science and Technology. (Unpublished thesis)

Dadson, J.A., Kwadzo, G.T-M. & Baah, K.O. 1994. Structural adjustment and marketing of roots and tubers in Ghana. Accra, FAO. (Report for FAO Regional Office for Africa.)

Day, G., Graffham, A.J., Ababio, J. & Amoako, M. 1996. Feasibility study: market potential for cassava-based flours and starch in Ghana. Accra, University of Ghana, Department of Food Science and Nutrition; Greenwich, London, Natural Resources Institute.

Henry, G. & Gottret, M.V. 1995. Cassava technology adoption: constraints and opportunities. In Cassava breeding, agronomy research and technology transfer in Asia: proceedings of the fourth regional workshop held in Trivandrum, Kerala, India, 2-6 November 1993, pp. 410-432. Bangkok, CIAT.

Jebuni, C., Asuming-Brempong, S. & Fosu, K.Y. 1990. The impact of economic recovery programmes on agriculture in Ghana. Accra, USAID. (Report.)

Kreamer, R.G. 1986. Gari processing in Ghana: a study of entrepreneurship and technical change in tropical Africa. Ithaca, NY, USA, Cornell University Department of Agricultural Economics. (Cornell International Agricultural Economics Study.)

Ministry of Agriculture. 1990. Medium-term agricultural development programme. Accra.

Ministry of Food and Agriculture. 1995. Proceedings of workshop on cassava as a substitute in the feeding of poultry and livestock, Kumasi, 11-12 September 1995. Accra.

Ministry of Food and Agriculture. 1996. Agriculture in Ghana: facts and figures. Accra.

Nweke, F.I. 1996. Cassava: a cash crop in Africa. Ibadan, Nigeria, International Institute of Tropical Agriculture. (COSCA Working Paper No. 4.)

Seini, A.W., Nyanteng, V.K. & van den Boom, G.J.M. 1997. Income and expenditure profiles and poverty in Ghana. In W.K. Asenso-Okyere, G. Benneh & W. Tims, eds., Sustainable food security in West Africa, pp. 55-78. Boston, Kluwer Academic Publishers.

Stryker, J. & Dumenu, E. 1986. A comparative study of the political economy of agricultural pricing policies: the case of Ghana. Washington DC, World Bank (unpublished).

TechnoServe Ghana. 1994. Feasibility report on cassava chips enterprise in Ghana. Accra.

World Bank. 1993. Ghana 2000 and beyond: setting the stage for accelerated growth and poverty reduction. Washington DC, Africa Regional Office, Western Africa Department.