Chapter 6. Promoting farm/non-farm linkages: a case study of French bean processing in Kenya - Lydia Neema Kimenye

INTRODUCTION

With over 50 percent of the population living below the poverty line, poverty alleviation has necessarily been a primary goal of Kenya’s development strategy. Given that over 80 percent of the population and the majority of the poor are based in rural areas, promotion of employment and other income-generating activities in rural areas is crucial to alleviating poverty. The rural areas of Kenya have already been singled out as having “acted as a very effective ‘sponge’, absorbing and retaining the population and labour, so that rural-urban migration has not been as serious as it could have been, given the high rates of population growth and the limited availability of cultivable land” (World Bank, 1994). What is urgently needed is identification and promotion of economic activities that offer the greatest potential for employment and income generation at grassroots level, so that rural areas can continue to absorb population growth.

Most of the poverty alleviation strategies employed in the past have focused on increased commercialization of smallholder farming, particularly through cultivation of tea and coffee and recently through production of horticultural items for export. Although the farm sector is considered to be the backbone of the economy and a major source of food and income for most Kenyans, recent evidence shows that rural household income in Kenya is increasingly diversified, with a substantial share coming from sources outside the farm. A recent review of several field studies on rural household income diversification in Africa shows the importance of non-farm earnings in rural household income (Reardon, 1997). It is estimated that in Kenya as much as 60 percent of rural household income is gained from non-farm sources (World Bank, 1994); the major sources include non-farm wage employment in rural areas, such as working in agroprocessing enterprises, and profits from small-scale enterprises in the non-farm informal sector.

Evidence from past studies suggests that promotion of farm/non-farm linkages, especially those focusing on commercialization of smallholders, has enormous potential to create employment and to further diversify income. Many smallholders are becoming increasingly commercialized by growing high-value non-traditional crops such as fruits and vegetables for the fresh export and processing markets. Vegetable production is currently the most important commercial horticultural enterprise among smallholders, especially those with very small farms of less than 2 ha. About 80 percent of fresh export vegetables are grown by smallholders, who sell them to exporters through contracts or brokers (Kimenye, 1995). Some of the vegetables are processed and sold in domestic or export markets. Besides providing income directly to farm households that cultivate the crops, commercialization of farm subsectors can generate farm employment and stimulate greater rural non-farm economic activities (Haggblade and Hazell, 1989). Kenyan policy makers should as a matter of importance be informed about the nature of linkages emanating from or towards vibrant smallholder subsectors such as French bean processing, and about ways in which such linkages could be exploited to enhance rural incomes and overall economic growth.

Haggblade and Hazell (1989) have shown that rural non-farm activities across Africa account for about 14 percent of full time employment and between 25 percent and 30 percent of income. Statistics show that in Kenya, at a time when employment opportunities in the public sector are shrinking and the formal sector is growing at only a modest rate, small and microenterprises account for over 50 percent of employment outside agriculture. The informal sector has an annual employment growth rate of over 10 percent (Ministry of Agriculture, Livestock Development and Marketing, 1996). Horticulture, especially fruit and vegetables, maize and dairy commodities were the subsectors identified by government and development agencies as areas with the greatest potential for promoting equity-enhancing small and microenterprises (Ministry of Agriculture, Livestock Development and Marketing, 1996; TechnoServe, 1997). Most microenterprises supply consumers in towns, but it is becoming widely recognized that rural microenterprises are gaining considerable importance across much of Africa as sources of employment and incomes (Jaffee and Morton, 1994) and that they have the potential to generate a variety of linkages within the rural economy.

The objective of this chapter is to characterize the nature of farm/non-farm linkages in Kenya, to explore the ways in which the geographic location of processing firms influences linkage effects, and to identify constraints to the expansion of linkages. These issues need to be addressed in order to enhance employment opportunities and income diversification in rural Kenyan households. The analysis is based on a detailed case study focusing on the French bean processing subsector, chosen because of its growing importance for the fresh export and processing markets and its potential to benefit smallholders producing French beans and households linked to the subsector. Detailed information is provided on the households that participate in this market and the processing firms.

The remainder of this chapter is organized as follows. Section 2 provides background information for the study and details of the analytical approach. Section 3 examines farm households involved in producing French beans and ways in which production of French beans is linked to the rest of the rural economy. Section 4 investigates the firms processing French beans and their linkages to other sectors. Section 5 contains conclusions.

BACKGROUND

The choice of French beans for analysis of intersectoral linkages is based on the growing importance of horticulture in general and the importance of the French bean in particular. Horticulture is the fastest growing and commercializing agricultural subsector, especially with regard to production for the fresh export and processing markets. Statistics from the Horticultural Crops Development Authority (HCDA) show that exports of fresh horticultural produce grew dramatically from about 1 476 mt in 1968 to 57 363 mt by 1992, worth about KSh2.5 billion or approximately US$48.0 million. In 1996, the export volume of fresh horticultural exports had reached 84 824 mt according to HCDA export statistics. The horticultural processing subsector also recorded a remarkable increase, particularly in terms of processed products destined for the export market. Between 1985 and 1992, for example, the volume of processed horticultural produce rose by 50 percent from 54 465 mt to 81 551 mt worth KSh1.8 billion, or about US$34.0 million. Kenya’s processed horticultural products include frozen products, notably French beans, snow peas and juice concentrates, canned products, mainly French beans, baby corn, juices, fruit slices, jams and marmalade, and dehydrated products such as cabbages, carrots, leeks and onions. The dynamism of this sector makes it of special interest and ideal for examination of the development of linkages.

Two types of data were collected for the case study. First, two farm household surveys were conducted, one in Njoro division of Nakuru district and the other in Maragwa division of Maragwa district. The two areas differ in one major respect: in Njoro, processing firms are located in a rural bean-growing area, whereas in Maragwa processing firms are located in Ruaraka, near Nairobi. The household interviews were based on a structured questionnaire. Second, the processing firms supplied by the surveyed households were studied. Informants in the French bean processing firms in Njoro and those based near Nairobi were interviewed through a semi-structured questionnaire.

Maragwa division, situated in Maragwa district in Central Province, covers an area of 20 000 ha; about 15 000 ha of this is suitable for agricultural production. The division has an estimated population of 109 000 people and 17 690 farm families. The average size of farm holding is 1 0.4 ha. In the recently settled Samar and Maragwa Ridge, however, farm sizes reach 80 ha. The main crops grown in Maragwa are maize, beans, bananas, potatoes, coffee, French beans, tomatoes, cabbages and kale; other crops include mangoes, avocadoes, cotton and soybeans. The Ministry of Agriculture estimates that there are currently about 40 ha of land under French beans every season.

Njoro division, a high-potential agricultural area in Nakuru district, covers an area of 774 km2, all of which is suitable for agricultural production; it has a population of 160 607 people. The size of farm holdings varies from 0.4 ha to 4 000 ha; farms of 8 ha or less are small-scale farms, those between 8 ha and 20 ha are medium-scale and those over 20 ha are large-scale. The major crops are maize, wheat, barley, potatoes, cabbages, kale and pyrethrum. Dairy production is an important enterprise in the division. Large and medium-scale farmers usually have access to irrigated land for production; water usually comes from boreholes. A few smallholders use water from roof catchments to irrigate vegetables. Only a few farmers are producing French beans, because of market limitations.

FARM HOUSEHOLDS AND FRENCH BEAN PRODUCTION

Household characteristics in the study region

In Maragwa, 75 farm households were interviewed; 31 percent of the households in the sample were headed by women. The division has three main trading centres - Maragwa, Gakoigo and Sabasaba - where farm households can obtain production inputs, consumer goods and services, or sell their produce. The distances from farms to the nearest trading centre range from 0.5 km to 10 km, but 72 percent of farmers are within 2 km of a trading centre; Maragwa town is the closest centre for 43 percent of the farmers interviewed.

Most of the French bean growers in Maragwa are young; nearly 60 percent of farmers in the sample are 35 years old or less. According to the Maragwa division agricultural extension officer, French bean production has attracted many of the unemployed young people in the area.

Farm holdings in Maragwa are fairly small: the farmers interviewed had between 0.05 ha and 5.67 ha; average farm size was 0.75 ha. About 5 percent of the respondents did not own land, but rented it to grow beans. Some of the young people grow French beans on portions of land owned by their parents. The survey data indicates that 70 percent of farmers had rented land during 1996; 84 percent of them used it for French bean cultivation. Average hired land per household was 0.18 ha per season; the rental rate was KSh2 400/ha per season; the average amount of land hired per household for the purpose of growing French beans in 1996 was 0.08 ha per season.

The practice of hiring land to grow French beans is more prevalent in Maragwa than in Njoro. Farmers in Maragwa hire additional land to plant French beans, because the amount of land they own is very small and devoted to food crops, and in order to get access to irrigation water. Farmers whose land is at a high elevation and therefore difficult to irrigate often hire portions of land along rivers, on valley floors or near shallow wells for French bean production. The undulating topography of Maragwa has more rivers, valley floors and shallow wells than Njoro. With access to irrigation water, it is possible to grow up to four crops of French beans per year.

In Njoro, 47 farm households were interviewed, of which 51 percent were headed by women. The major trading centre frequented by farmers is Nakuru town, approximately 20 km from Njoro; other local market centres are Njoro town and Egerton centre. Njoro farmers are older than those in Maragwa - 55 percent are over 35 years of age - and have larger farm holdings; average farm size is 1.37 ha. About 79 percent of the farmers in the Njoro sample had hired land from other farmers in the division during 1996; the average size of land hired per household was 0.75 ha. Only 15 percent of the farmers had hired land to plant French beans. The cost of hiring land in Njoro during 1996 was KSh800/ha per season.

French bean production and returns

Two types of French beans are produced in Kenya, one for the fresh export market and the other for processing. Maragwa farmers produce both kinds; Njoro farmers produce only the processing variety. This may be because of Maragwa’s proximity to Nairobi, where the fresh-produce exporters operate, and to Jomo Kenyatta International Airport, the main point of export. Maragwa is about 100 km from Nairobi; Njoro is about 250 km away.

Although Maragwa produces both types of bean, the survey data indicates that production for the fresh export market has declined substantially in the last three years, probably as a result of increased competition from the processing subsector and the breakdown of farming contracts between farmers and exporters. Processing firms buy beans from Maragwa under contractual arrangements with smallholders, under which processors usually provide farmers with input credit and farmers are assured of a market. Most fresh-produce exporters do not provide farmers with input credit, however, so farmers are frequently unsure whether a buyer will collect their produce or what the price will be.

The survey indicated that 47 percent of smallholders in Maragwa had been growing French beans for the fresh market in 1994, but that by 1997 only 21 percent were doing so. The average area under French beans for fresh export per farm household has remained roughly the same: 0.08 ha in 1994 and 0.09 ha in 1997.

French bean production for processing has increased substantially in the last three years. Many smallholder farmers, particularly the young or unemployed, have turned to French bean cultivation to earn their livelihood or augment income from traditional crops. Of the 75 farmers randomly sampled in Maragwa, 85 percent were growing French beans for processing in 1997; 67 percent of them started in 1997, 19 percent in 1996, 11 percent in 1995 and 3 percent in 1993. None of the farmers in the sample had been growing French beans for processing five years previously. Two main reasons were given by the farmers as to why they started to grow French beans for processing: about 46 percent started to grow the beans to augment farm income; 44 percent were motivated by the existence of a more reliable market, Frig-O-Ken Ltd. The reliability of the market encouraged 34 percent of the farmers to increase the area planted per season; however, 42 percent of farmers are unable to increase the area under beans because of lack of sufficient labour, especially at harvest, 21 percent have inadequate water for irrigation, and some are not allowed by Frig-O-Ken to increase the area.

Table 1 shows a breakdown of bean growers in Maragwa Division during 1996 by age and farm size. The results confirm the reports of the Maragwa district agricultural office that introduction of French bean production for processing during the early 1990s made farming, and particularly bean cultivation, attractive to young people, many of whom turned to it as a source of employment and income. Table 1 indicates that about 39 percent of bean growers in Maragwa are between 20 and 30 years of age; about 41 percent own no more than 0.4 ha of land. In 1996, the mean area under French beans for processing was 0.1 ha per household per cropping season. Of those growing beans, 22 percent were women, who planted an average of 0.06 ha each per season; 78 percent were men, who planted about 0.11 ha each per season.

TABLE 1
Age and farm size characteristics of French bean growers in maragwa division, Maragwa district, 1996

¹These growers either hired land or borrowed some portion of land from their parents.
Source: Author’s calculations from household survey data.

Production of French beans for processing in Njoro has not increased as remarkably as in Maragwa. Smallholder French bean production had virtually stopped during the time of the survey in 1997. None of the 47 smallholder farmers interviewed were growing French beans in 1997. About 70 percent of the farmers in the sample were growing French beans in 1994, but by 1996 only about 30 percent were doing so. In 1994, the average area under French beans per farm household was 0.18 ha; by 1996 it had declined to only 0.09 ha. Reasons for the decline and eventual disappearance of French bean cultivation include an acute shortage of labour, particularly for harvesting, lack of a reliable market and low prices. As is indicated in the section on farm incomes, growers in Njoro incurred huge losses from the bean crop in 1996, because they lacked a market and because of low prices. A significant proportion of farmers did not harvest their 1996 crop.

The problem of market failure is at first glance perplexing, because there are two processing firms, Njoro Canning Factory and Kokoto Factory, in Njoro division. Njoro Canning Factory is the oldest French bean processing firm in Kenya, but it obtains most of its French beans from contract growers in Vihiga district, about 190 km away, and Kericho district, about 170 km away, primarily because of the availability of water for irrigation in these regions. The factory has its own nucleus farm in Vihiga. Njoro Canning Factory buys 5 percent of its beans from farmers in Njoro Division on an irregular basis.

Kokoto factory was at one time the main market for smallholder farmers in Njoro Division. According to the Njoro divisional agricultural extension officer, a substantial number of smallholder farmers started to grow French beans around 1994 under contract to supply Kokoto factory, which began processing operations in May 1995. The contract growers were tenants on the firm’s 23 ha farm; they were supplied with irrigation water from boreholes. In 1996, however, the company experienced a serious cash-flow problem after a large consignment of its processed French beans, worth KSh28.0 million, was rejected by customers in Europe because of poor quality. The firm was therefore unable to pay its contracted growers and suspended the contracts. The Kokoto factory currently produces all of its beans itself.

Production of French beans is highly intensive in terms of labour, fertilizer and agrochemicals. Input credit in kind, and in cash to pay for hired labour at harvest, are critical because the majority of smallholder farmers are resource-poor. In the Maragwa sample, 88 percent of farmers were producing French beans under irrigation; in Njoro, production was rainfed. This is because there are more and cheaper sources of water for irrigation in Maragwa, such as rivers - 85 percent of farmers irrigating French beans get water from rivers - and shallow wells; in Njoro the main source is boreholes. Access to irrigation water enabled farmers in Maragwa to produce up to four crops a year; in Njoro there was only one crop. The most common irrigation system in Maragwa is by bucket: farmers use watering cans to collect the water irrigate the crop manually.

Virtually all smallholder growers in Maragwa sell their produce to Frig-O-Ken under a contract arrangement that specifies the area to be planted per season, planting dates, dates and number of times to spray against pests and diseases, the date to start and finish harvest and the price. Farmers are paid at the end of each season. Although some farmers had formed farmers’ self help groups to facilitate negotiation of prices, the final contract is drawn between the individual grower and the company. Under the contract, the grower is supplied with seed, fertilizer and agrochemicals on credit.

During harvest, each grower transports beans daily to the nearest Frig-O-Ken collection point, which is usually about 2 km away. They carry the beans on foot or on bicycles to collection points, where the beans are inspected and weighed; weights are recorded against growers’ contract numbers. During 1996, prices offered by Frig-O-Ken were KSh25/kg in the first season, KSh26/kg in the second, KSh27/kg in the third and KSh30/kg in the fourth. In contrast, the prices offered for beans in the fresh-export market showed a higher average over the year, but fluctuated between KSh10/kg and KSh100/kg.

When farmers started to plant French beans in Njoro in 1994-1995, they had contracts with the Kokoto factory. Under the contract, farmers were supplied with input credit in terms of seed, fertilizer and chemicals and each farmer was allocated between 0.1 ha and 0.4 ha on the company farm and supplied with irrigation water. The produce was collected from farmers’ plots by the company tractor and taken to the adjacent factory. Under the contract, farmers should be paid at the end of the season, but in 1996 the contracts were suspended or not honoured by the company. Most farmers had no alternative market for the crop. Njoro Canning Factory bought beans from a few farmers, but the price offered was too low and many farmers chose not to harvest. The average price in 1996 was KSh23/kg.

In general, smallholder bean growers seem to be earning significantly lower net returns than at the beginning of the 1990s, despite the current boom in the bean-processing subsector. In 1992, smallholder farmers in Mwea division in Kirinyaga district were earning an average of between KSh6 000/ha per season for non-contract growers and KSh11 200/ha per season for contract growers (Kimenye, 1995). Average net earnings for smallholder farmers in Maragwa in 1996 were KSh2 200 from an average holding of 0.1 ha per year.

Mean gross earnings for bean growers in the sample were KSh14 440 per year; total expenses were KSh8 941, which gave an annual net return of KSh5 499 per farm household. Although low, the net return from French beans was the highest among all farm enterprises. The other major crops were maize, beans and potatoes for food, which showed negative net returns resulting from low yields and low prices. Table 2 shows the distribution of net farm earnings among farmers, based on age and gender. Women farmers obtained higher net returns per year than men, partly because they do a better job of harvesting.

TABLE 2
Average net returns from French beans, by gender and age group of farmer, Maragwa division, 1996

Source: Author’s calculations from household survey data

Average gross returns on French beans in the Njoro sample during 1996 were significantly lower than in Maragwa, primarily because most of the growers in Njoro failed to harvest or sell the crop because there was no market. Average net returns per household in Njoro were negative at KSh-2 421. Average expenses were about KSh6 338. Table 3 shows the distribution of net returns for bean growers in Njoro division; the figures indicate that most obtained negative returns from French beans. Unlike smallholder farmers in Maragwa, Njoro farmers have larger holdings and can cultivate maize, wheat, barley and potatoes for commercial purposes and cabbages, carrots and other vegetables on a larger scale for the domestic market.

TABLE 3
Average net returns from French beans, by gender and age group of farmer, Njoro division, 1996

Source: Author’s calculations from household survey data.

Marketing problems in terms of lack of a reliable market outlet and low prices are the most serious difficulties confronting growers in the Njoro sample. Lack of water for irrigation and shortage of labour have discouraged Njoro farmers from growing French beans. Farm workers demand higher wage rates for working in the French bean plots, especially to pick the beans. Because of the low price and the unreliable market, many growers were unable to pay higher labour wages and left the crop to seed. Several of the farmers interviewed had a stock of French bean seed, which they did not know what to do with; the seed has a strange taste and is not consumed locally.

The main constraints facing growers in Maragwa Division include lack of labour, insufficient water for irrigation and the high cost of agrochemicals and fertilizer relative to the price they get for beans. A few farmers stated that shortage of land had prevented them from increasing French bean production. Among the problems, the high cost of inputs was stated by 45 percent of farmers to be the most critical constraint, which made for low net returns; 30 percent of farmers stated that their main constraint was was inadequate water for irrigation. Fewer than 10 percent complained about the payments being made at the end of each harvest; they argued that low prices left them unable to hire labour to complete harvests.

Comparison of farm and non-farm income

This section presents a comparison of the contribution of farm and non-farm income to the net income of rural households; the results demonstrate the increasingly important role of the non-farm element. The sources of non-farm income are described, such as wage employment, profit from business and the share of income from French bean production.

Non-farm income is crucial to rural farm households, because it helps to smooth the flow of farm income over the cropping cycle and it stabilizes income by spreading risk through diversification (Lanjouw and Lanjouw, 1995). For smallholders in areas where agricultural output varies greatly over a year or years because of unpredictable weather conditions, the seasonal smoothing and risk diversification obtained through non-farm income sources can be very important.

Table 4 shows the composition of net household income for rural households in Maragwa. The results are fairly consistent with findings from elsewhere showing that non-farm income contributes a significant share to rural household income. The present study estimates that rural households in Maragwa obtain about 63 percent of their income from non-farm sources. The magnitude of non-farm income is small, however. Net farm income was negative mainly because of poor performance of food crops; net returns for French beans were positive.

Sources of non-farm income include wage employment, operating a food kiosk, selling charcoal and retailing maize grain. Only 15 percent of farm households in Maragwa reported owning a non-farm business. The low level of involvement of rural farm households in non-farm business observed in this study is consistent with evidence drawn by Reardon (1997) from past studies. Lack of start-up capital was the main reason preventing smallholders from venturing into non-farm enterprises or expanding the ones they already have.

TABLE 4
The composition of net household income for rural households in Maragwa, 1996; net income per year and proportion of non-farm income in household income

Source: Author’s calculations from household survey data.

A third of farm households in Maragwa earned non-farm income from wage employment; of these, 72 percent were working as farm workers for their neighbours. The rest of the farmers who were employed off-farm were working in restaurants, food kiosks and canteens, in shops and as civil servants. Of the ten households that reported having a non-farm business, the business belonged to the husband in six cases, and to the wife in four cases. Businesses owned by women were food kiosks, tailoring and charcoal retailing, all within the division; men were involved in the maize trade as middlemen and in selling building sand in Nairobi.

Virtually all the non-farm businesses were started between 1994 and 1996. In general, start-up funds consisted of savings from farm income; none of the owners indicated that the money came strictly from French beans, however. Given the poor performance of other crops, it is nevertheless plausible to assume that income from French bean production could have contributed to start-up funds.

The level of employment generation from non-farm enterprises owned by farm households was very low. Hired labour was used sparingly, because the enterprises were usually run by the owner with the assistance of one full-time worker and one or two casual workers.

TABLE 5
The composition of net household income for rural households in Njoro, 1996; net income per year and proportion of non-farm in household income

Source: Author’s calculations from household survey data

Table 5 shows the composition of rural household income in Njoro, where farm income plays a much larger role than in Maragwa. This may be partly because farm holdings in Njoro are substantially larger than those in Maragwa; there is also greater diversity from more profitable farm enterprises. The proportion of non-farm income in net household income is thus only 27 percent. Although this is low, the estimate obtained for the Njoro sample still falls within the range found across African households (Reardon, 1997). Among farmers between 20 and 30 years old, the proportion of income from non-farm sources is considerably higher at 48 percent, primarily because the majority had off-farm work.

The proportion of farmers owning non-farm business is higher in the Njoro sample than in Maragwa; 23 percent of farm households in Njoro reported that they were self-employed, selling food or general provisions from kiosks in 55 percent of the cases, trading in maize or working as carpenters. All except one belonged to men. With the exception of one maize dealer, all the businesses were located in Njoro Division.

As in Maragwa, non-farm enterprises run by rural households in Njoro do not generate significant rural non-farm employment. The owners run virtually all of them, with assistance from family members on a part-time or full-time basis. Most of the enterprises in Njoro were started between 1994 and 1997. Over 70 percent of the owners said they obtained start-up funds from savings generated from farm income; start-up funds averaged about KSh5 500. French beans were not significant in the generation of start-up capital.

Farm/non-farm linkages at the household level

Anticipated farm/non-farm linkages at the farm level include backward and forward production linkages and consumption linkages. Employment linkages associated with non-farm activities are apparently non-existent because of low participation by smallholders in non-farm enterprises. Because the focus of French bean production at local level is on small-scale producers, production and consumption linkages are especially important in terms of stimulating national economic growth and because they can have profound effects on poverty alleviation and spatial growth patterns.

One of the premises underlying this case study is that the growth of the French bean subsector and the increase in smallholder farmers’ production of beans for the processing market had reached a level sufficient to induce supply of non-farm production inputs from within the local area (backward linkages). The hypothesis is that as more and more smallholder farmers cultivate French beans for the processing subsector, a significant proportion of their demand for non-farm inputs such as fertilizers and repair services for farm tools and equipment will be supplied by local sources. The theory on linkages and experience from past studies point to the fact that the type and extent of development of backward linkages is determined by factors such as the farm size, type of crop and agricultural technology, and whether production is rainfed or not. In this study, the size of farm holdings in Maragwa is significantly smaller than those in Njoro division. The technology for producing French beans appears to be the same in both areas, but in Maragwa production is irrigated. These characteristics have in theory different implications regarding the type and magnitude of backward linkages. With large farms in Njoro division, for example, linkages associated with the demand for tractors or ox ploughing and related services would be expected, whereas in Maragwa linkages associated with demand for irrigation equipment would be expected.

As expected, the results show that 51 percent of farm households in Njoro rent tractors from other farmers in the region; none of the farmers in Maragwa reported demand for tractor services or any other farm machinery. None of those who were renting out tractors was captured in our sample, however, perhaps because the farmers who own tractors in Njoro are large-scale farmers who were excluded from the survey. The rental rate for tractors varies from KSh320/ha to KSh480/ha, depending on whether it is used to plough or to harrow; the mean is KSh400/ha.

In Maragwa, it was found that because production of French beans is irrigated, 87 percent of smallholder farmers had bought at least two or three watering cans; nearly 90 percent of farmers obtained them in Maragwa: 66 percent bought from Maragwa trading centre and 23 percent from Sabasaba trading centre. A local entrepreneur has started manufacturing watering cans in Maragwa trading centre and supplying them to farmers in Maragwa and in nearby Murang’a district. This is an excellent example of how the booming French bean subsector has generated backward linkage into the rural informal non-farm economy.

Another input related to French bean production and the horticulture sector in general is knapsack sprayers, for which there is substantial demand. Farmers use them to spray agrochemicals on to vegetables. In Maragwa, however, the sprayers were used to spray agrochemicals on to cabbages, tomatoes, kale and potatoes, and were rarely used in French bean production. This is because Frig-O-Ken employs people to spray crops. But 46 percent of smallholder farmers in Maragwa purchased knapsack sprayers, over 50 percent locally in Maragwa or Sabasaba; the rest bought them from outside the area, 29 percent from Thika and 17 percent from Nairobi. In Njoro, 57 percent of farmers in the sample purchased knapsack sprayers, mainly from Nakuru town. A few farmers rent sprayers from other farmers in the region.

In terms of backward linkages associated with other traditional production inputs, fewer than 15 percent of the farmers in Maragwa sought production inputs for French bean production from local suppliers, because they obtained supplies from Frig-O-Ken under the terms of the farming contract. Although farmers benefit from input credit given by the processor, the practice limits the spread of backward linkages into the rural small-scale non-farm economy. Other production inputs such as seed maize, dry bean seed, fertilizer and agrochemicals for other crops were obtained locally from a Maragwa trading centre or Sabasaba trading centre. In Njoro, virtually all production inputs used by farmers for major crops during 1996 were obtained from nearby Nakuru, mostly from medium-scale input suppliers. None of the farmers reported buying production inputs in local trading centres such as Njoro, Egerton centre or Pivie market, probably because of the close proximity of Nakuru.

Forward linkages at the farm level are insignificant. The only meaningful forward linkage is the supply of beans to the processors. After production, the beans are transported to the factory or sale point by the farmers, or they are collected from farms by Kokoto factory in Njoro.

In terms of consumption linkages, the results of the case study reflect spending patterns of rural households across Africa, in that a large proportion is spent on food items and less on rurally produced non-food goods such as furniture. This implies that current consumption profiles of rural households do not induce growth linkages into the rural non-farm economy.

FRENCH BEAN PROCESSING FIRMS AND LINKAGES

Evolution of the processing subsector

The first bean-processing firm, founded in the 1970s, was called Njoro Canning Factory; it still exists, but under a different name and management since1980. It was started as a joint venture between a French company and a local firm. The firm developed a nucleus farm in Vihiga district in what was then Kakamega district and introduced French bean production to smallholders in the area through an outgrower contract-farming scheme. The beans from the company farm and from contracted growers were transported to Njoro for processing and then exported to France. Vihiga is approximately 190 km from Njoro, and its factory still obtains over 50 percent of its raw beans from Vihiga District.

Until the early 1990s, Njoro Canning Factory was the only firm processing French beans for export. Other firms processing horticultural crops focused on fruit products such as juices, slices and jams, and vegetable products such as tomatoes, haricot beans and dehydrated leeks, cabbages, onions and carrots. The information available from export volume and value data and the number and size of processing firms indicates that the turning point for the bean processing subsector occurred in the 1990s. The largest French bean processing firm in Kenya, Frig-O-Ken, was established in 1994. Two other bean-processing firms were established at about the same time: Highland Canners in 1990 and Kokoto Factory in 1995. During the same period, Njoro Canners increased their processing plants at Njoro from one to three; in 1997 the firm established a new French bean canning factory at Juja in Thika district, about 50 km from Nairobi. At the time of the survey in December, 1997 another firm was being set up at Ruiru in Thika district.

By 1997, there were four French bean processing firms in operation. Frig-O-Ken was the largest in terms of estimated volume of beans processed per day, number of permanent employees and casual workers and number of contracted growers; this information is based on 1995 estimates of Frig-O-Ken’s operational capacity obtained from field staff and survey data of the three other process-ing firms. In 1995, Frig-O-Ken had more than 300 full-time employees and over 1 000 casual workers at the factory, compared with 70-100 full-time staff and 100-1 000 casual workers in the other firms. Frig-O-Ken handles over 50 mt of raw beans per day compared with about 2-12 mt per day in the case of the other processing firms (see Table 6). In 1995, Frig-O-Ken had about 3 500 contract growers and was expecting to increase the number to over 8 000 within two years. Information about the firm’s current processing situation at the factory and in the field was unfortunately not available, because the company declined to be interviewed.

TABLE 6
A comparison of the processing capacity of rural farms in Njoro and urban farms in the Nairobi area

¹ Highland Canners used to process about 7 mt of raw materials when the company started processing beans in 1990. The decline in the amount of beans processed results from a problem of supply.

Source: Author interviews.

Among the main reasons for the proliferation of bean-processing firms in and around Nairobi, an urban area where beans were not previously grown, are that the city is close to the main French bean growing areas of Murang’a, Maragwa, Thika, Kiambu, Kirinyaga and Nyeri in Central Province and Embu, Meru, Machakos and Makueni in Eastern Province, and close to other firms that supply important secondary inputs used in food processing, such as containers for raw and finished products, and firms that supply equipment and spare parts. Bean processing firms use large quantities of cardboard cartons and plastic crates when transporting French beans from farms to the processing factories and during sorting, grading and handling of the beans prior to processing. Other inputs used are glass jars and cans for packaging the finished products. Most of the firms that manufacture handling and packaging materials are found in Nairobi and Thika town. Metal Box Kenya, for example, is the largest manufacture of cans in Kenya; its head office is in Nairobi and the factory is in Thika town. Most food processing firms that use cans for packaging of the finished products obtain their cans from the Metal Box factory. Proximity to Nairobi is also convenient for the firms in terms of following and obtaining quick administration of official matters such as renewal of licences and permits, exemptions or payments of duty and resolving income tax issues. It is often necessary in Kenya to follow up official bureaucratic processes personally in order to speed up the paperwork. Being closer to the government offices can make a difference in terms of time and money. Road conditionss in Kenya are very poor, so Nairobi’s central location and direct access to the main gateways for export of horticultural produce - Jomo Kenyatta international airport and the port of Mombasa - make it a convenient location for the business. Some of the produce, especially frozen French beans, are transported in refrigerated trucks or containers to the export exit points.

The nature of linkages associated with the processing firms

The emergence of French bean processing firms has stimulated local smallholder cultivation. Contrary to the expectation from theory whereby a firm invests in a processing factory or increases its capacity in response to increased production of raw material that has attained a certain critical mass, experience in Kenya suggests that bean firms emerged in response to an export demand for processed beans. In turn, the bean firms established their own farms in order to supply some of the beans and then bought the rest from smallholders through contract-farming arrangements. The establishment of contract farming that often includes provision of input credit is major factor in motivating smallholders, especially those with very small farms, to concentrate on French beans for the processing market. In spite of the low levels of net return from French beans because of the low prices offered by the processors, the enterprise is still the major source of income for smallholders in a region with high population density and without alternative cash crops such as tea or coffee.

The data from the survey of bean firms indicate that in general the processing firms obtain between 50 percent and 75 percent of raw materials from smallholders, mainly in the central, eastern, western and Rift Valley provinces. Njoro Canning Factory obtains 50 percent of the beans it processes from its own farm, 45 percent from contract growers and 5 percent from non-contract growers. The number of contract growers is currently about 200 in Vihiga and 30 in Kericho. Highland Canners gets 25 percent from its own farm of about 137 ha in Embu district and 75 percent from smallholders in Murang’a, Maragwa, Mwea and Thika districts.

Bean processing firms generate backward production linkages to traditional production-input dealers. Virtually all the processing firms own a nucleus farm to produce a proportion of their beans. Where the processing firm has established farming contracts with smallholder bean growers, the firm supplies farmers with production inputs in credit form. All the firms stated that they buy inputs such as fertilizer and agrochemicals in bulk from dealers, in Nakuru in the case of Njoro Canning Factory and Kokoto Factory and Nairobi in the case of Highland Canners. This means that none of the business associated with the provision of production inputs is injected into the local community.

Another production linkage deals with transport of beans from production areas to processing factories. The available information suggests that transport of raw materials from farms to the processing factories does not offer significant opportunities for development of linkages between the bean firms and truckers. Only Highland Canners reported use of middlemen for transport of beans sourced from smallholders. This may be because Highland Canners does not have farming contracts with smallholders and relies on middlemen to buy beans from farmers. Highland Canners uses its own trucks to transport beans from its own farm, accounting for 25 percent of the beans it processes; the rest are transported by middlemen locally referred to as dealers or brokers. The firm uses three to five dealers per week during peak seasons.

The other aspect of the transport linkage is moving finished products from the factory to the point of export. The available data indicates that development of this type of linkage is limited because most bean processing firms use their own trucks for transport to the point of export. This is because the products, particularly frozen beans, are often transported in special refrigerated trucks owned by the firms. Njoro Canning Factory indicated that about 20 percent of its finished product is transported to Mombasa in hired transport. Every week in the peak season, about ten 20-ton refrigerated container trucks transport produce to Mombasa; two or three of them are hired.

The case study identified other important auxiliary activities which arise indirectly from or to the bean-processing firms, including activities such as providing special secondary inputs such as packaging and services such as repair and maintenance of factory equipment and vehicles, clearing and forwarding and legal and medical services. Most bean firms employ a resident mechanic and have a workshop for minor maintenance work, but they also use servicing firms from nearby towns. The firms contract dealers, usually from Nairobi, to conduct annual safety inspections of the plants.

The most promising spin-off activity is provision of packaging materials, for example glass jars, cans, plastic crates and cardboard cartons, and uniforms for factory workers. Processing firms used to import the glass jars, but they can now be obtained locally in Nairobi. The data from Njoro Canning Factory, which was the most complete in terms of annual demand for packaging materials, indicate that a medium-scale processing firm would have the following annual requirements for packaging materials:

• about 1 300 plastic crates every 15 months bought from a manufacturer in Nairobi; at the price of about KSh100 per item in 1997, this amounts to KSh1.3 million;

• between 150 000 and 300 000 cans every month, obtained from Nairobi or Thika and costing between KSh2.4million and KSh4.8 million every month at peak season;

• uniforms for about 1 000 factory workers, two complete sets per year obtained at the closest major town, usually consisting of dust coat, cap and gum boots for the canning factory and pullovers, woollen caps and gum boots for the freezing plant.

Other spin-off linkages include supply of medical care to permanent employees and occasional medical service to casual workers in the case of factory accidents. Njoro Canning Factory hires the services of a private clinic in Njoro to treat employees. A spin-off service identified in Njoro and Nairobi was provision of lunch and tea for factory workers. At Njoro Canning Factory, the management invited a local resident to set up a canteen in the factory where lunches and teas are served. Under an arrangement between the management and the canteen operator, the workers can have their meals at the canteen on credit, with the amount consumed deducted from wages. In Nairobi, women come to the factory gates to sell hot food, usually a mixture of maize and beans, and porridge, tea and fruit, usually bananas.

Table 7 shows the type and amount of direct employment generated by the Kokoto factory at Pivie in Njoro division, a small-scale processing factory with a capacity of about 2 mt of raw beans per day at peak seasons. It obtains beans from its own adjacent farm; apart from the senior managers, virtually all its factory and field employees come from the community surrounding the factory. Some of its casual workers come from families that used to grow beans for the company in 1995, before the contractual arrangements with smallholders broke down. Kokoto Factory is about 15 km from Njoro Canning Factory.

TABLE 7
Type and amount of employment from a small-scale bean-processing firm in a rural area

Source: Author interviews.

As shown in Table 7, a single small-scale processing firm such as Kokoto Factory does not generate much direct employment; nearly 60 percent of those employed are casual labour, working in on-farm jobs. Of the work generated on-farm, 65 percent is for women, because more labour is required for harvesting than for other activities; women are preferred to men because of their gentle handling of the beans. The data obtained does not include a gender breakdown of those employed at the factory.

Although small, the number of workers employed in the field increased in the three years following establishment of the firm. In 1995, the firm had 2 permanent and 70 casual employees working in the field, advising and assisting the contract growers. At the time of writing it had 7 permanent employees and 142 casual workers, an increase of over 100 percent. Such growth in employment would be significant if there were more firms in the area.

Direct on-farm employment potential by other processing firms, including the Njoro Canning Factory and Frig-O-Ken, cannot be determined because the field production manager for Njoro Canning Factory was on leave at the time of the survey and Frig-O-Ken refused to be interviewed. Evidence from Kokoto Factory regarding the type and number of workers and the growth rate in employment over the last three years indicate that the direct on-farm employment creation potential for a medium-scale bean-processing firm with a nucleus farm and smallholder contracts is significant.

The data used to show direct non-farm employment generation potential and to illustrate the employment leakage resulting from the location of processing firms outside the rural bean-producing areas are drawn from information collected from Njoro Canning Factory and Highland Canners. Njoro Canning Factory is not located in a bean-producing area. Table 8 shows a breakdown of the type and amount of labour employed at Njoro Canning Factory and Highland Canners.

TABLE 8
Type and amount of employment from medium-sized bean-processing firms in a non-bean-producing area

Source: Author interviews.

A large percentage of the employment generated is for women casual workers: 87 percent of the employees at the factories are casual and 79 percent are women. The processing firms prefer casual labour because they do not operate at full capacity throughout the year. Women are preferred to men at the factory because most of the work is delicate, such as nipping off the tips of the beans and packing. At Njoro Canning Factory, about 50 percent of factory employees come from the surrounding community within 10 km of the factory. Because it is in a rural area, the factory contributes to reducing rural unemployment. The workers at Highland Canners in Nairobi are from the city and the nearby town of Thika.

Of the three processing firms interviewed, two indicated that they had reinvested some of their profits in the local area. Kokoto Factory reinvested some of its profits locally by establishing a bakery that supplies bread in Njoro and Nakuru and has diversified its activities into fruit processing; the firm processes about 3 000 pineapples per day and the produce is exported. Njoro Canning Factory has made two major reinvestments since 1980. It recently invested in a medium-scale processing factory at Juja in Thika district. The new factory is situated approximately 300 kilometres away, however, and so has no impact on employment and income of the residents in Njoro Division. The other reinvestment was the purchase of the Pan-African Vegetable Dehydration Plant at Naivasha and its re-establishment in Njoro town. Although highly capital-intensive and therefore generating only a small amount of non-farm employment, the dehydration plant has benefited local farmers because it is an alternative market for their cabbages, carrots, onions and leeks. In 1996, the company lost its major European market, Germany, when one of the German partners in the dehydration business left the company. The firm then dehydrated vegetables on a very small scale for the domestic market, mainly East Africa Industries, which uses the dehydrated and ground vegetables as a base for Royco Muchuzi Mix, a popular spice in the Kenyan market. The company identified another client in the export market, however, and resumed large-scale vegetable dehydration in 1998.

The extent to which the processing firms develop or promote linkages that enhance the employment and income of rural people is not related to their locations: neither the Nairobi-based nor the Njoro-based firm has generated significant on-farm linkages with rural bean-producing areas. Virtually all the non-farm growth linkages generate auxiliary business in large towns, particularly Nairobi. This is because the type of linkage activities that generate the most lucrative business seem to be capital-intensive, such as operating refrigerated container trucks and producing glass jars and cans. These capital-intensive, large-scale input supplying enterprises are owned by big firms or dealers in large urban areas.

Factors that have constrained or enhanced the growth of processing firms

None of the French bean processing firms cited any form of government regulation or policy on horticultural production, processing and marketing that has constrained them. The government has in fact played a significant role as a facilitator for the private sector, recognizing the immense contribution of the horticultural sector in generating employment, income and foreign exchange. HCDA provides advisory services to farmers, processors and exporters, particularly in marketing extension and initiation and organization of farmers’ groups to meet the needs of processors and exporters.

Another factor that has enhanced the growth of the French bean subsector is the ready export market for processed beans. Virtually all the firms interviewed, however, stated that identification of clients in the export market was easier if the firm had a partner or associate from the destination market who could ensure that quality requirements are met. All the firms visited had a manager from Europe, usually France or Germany, in charge of quality control at the factory.

Among the factors that have severely constrained the subsector is the poor and deteriorating condition of the highway from Nairobi to Mombasa and the roads in rural bean growing areas and urban centres, including Nairobi. The problem of poor roads needs to be addressed urgently if the growth rates of the processing subsector and the whole horticultural sector are to be sustained.

All the firms visited complained of problems of insufficient supply of water for irrigation, which made supply of raw materials less dependable, especially from smallholders. With development of more sources of irrigation water, such as boreholes, production could be increased in the current bean-growing areas and new areas could be opened up for growing and possibly local processing.

CONCLUSIONS AND RECOMMENDATIONS

The French bean processing subsector is one of the most rapidly growing subsectors, after cut flowers, in Kenya’s agricultural sector. The case study revealed that smallholders, particularly those in densely populated areas with very small farms, rely on French bean production as a major income-generating enterprise. The results indicate that many young unemployed people consider French bean cultivation as a source of employment, income and livelihood. Smallholders in the densely populated Maragwa division, whose farm holdings average 0.4 ha, have rapidly adopted French bean production in recent years.

A major factor in the rapid adoption of bean production in Maragwa is the availability of an accessible and reliable market outlet, the processing market. Although it is based in Nairobi, the processing firm has established French bean buying centres in the area. All growers have farming contracts with the processor, which assures them of a market and input credit.

The results of the study imply that the growth of the French bean processing subsector over the last three years has not contributed significantly to the development and spread of farm/non-farm linkages in the rural economy. It made no difference whether the processing firms were located in rural or urban areas: the effect on the development of non-farm linkages in the rural areas was insignificant in both situations. Most of the non-farm linkages generated as a result of growth in the subsector were experienced in the urban area, particularly Nairobi and its environs and to a limited extent in Nakuru town. These linkage activities stemmed largely from the spin-off linkage activities, particularly auxiliary firms that supply the agro-industry with packaging materials and services.

Another potential avenue for development of non-farm linkages in rural areas is backward production linkages related to the supply of production inputs such as agrochemicals and fertilizers. The demand for these inputs, particularly in Maragwa, appears to be high and probably capable of supporting informal small-scale suppliers in the area. The processing firms procure these inputs from Nairobi, however, and distribute them to smallholders on credit.

The poor condition of Kenyan roads is a major factor contributing to the firms’ preference for large urban areas such as Nairobi in terms of setting up their factories and associating with urban based input and service suppliers. Bad roads in most rural areas increase transportation costs and introduce high risks of failing to deliver supplies and materials on time. This is an issue that needs to be urgently addressed by policy-makers.

The findings of the study show the role of irrigation in increasing output and incomes and in terms of stimulating non-farm linkages in rural areas. It was found that in Maragwa, where French beans are produced under irrigation, farmers are able to grow four crops per year; in Njoro, where production is largely rainfed, growers can only produce one crop per year. Because of the high demand for watering cans that resulted from the increase in irrigated French bean production in Maragwa, an informal rural non-farm business has emerged to supply the cans from within the area. The availability of irrigated production also influences the processing firms selection of the regions from which to procure their raw materials. It is therefore crucial that efforts to encourage smallholders to produce French beans or other horticultural crops for export and processing markets, as currently being tried in parts of Rift Valley Province, should seriously investigate the availability of water for irrigation and determine ways to improve water access for smallholders in a cost-effective way.

REFERENCES

Haggblade, S. & Hazell, P. 1989. Agricultural technology and farm/non-farm growth linkages. Agricultural Economics, 3: 345-364.

Jaffee, S. & Morton, J. 1994. Africa’s agro-entrepreneurs: private-sector processing and marketing of high-value foods. Washington DC, World Bank Environmentally Sustainable Development Division. (AFTES Working Paper No.15.)

Kimenye, L.N. 1995. Kenya’s experience in promoting smallholder production of flowers and vegetables for European markets. African Rural and Urban Studies, 2(2/3): 121-141.

Lanjouw, J.O. & Lanjouw, P. 1995. Rural non-farm employment: a survey. Washington DC, World Bank. (Policy Research Working Paper No. 1 463.)

Ministry of Agriculture, Livestock Development and Marketing. 1996. Kenya: agriculture strategy for agricultural growth. Nairobi.

Reardon, T. 1997. Using evidence of household income diversification to inform study of the rural non-farm labour market in Africa. World Development, 25(5): 735-747.

TechnoServe. 1997. Delivery of non-financial services to micro and small enterprises in Kenya. Nairobi. (Paper for workshop at Barclays Bank Staff Training Centre, Karen, Nairobi, Kenya. May.)

World Bank. 1994. Kenya: employment growth for poverty alleviation. Washington DC.