CASE STUDIES OF CONTRACT FARMlING

IN THE HORTICULTURAL SECTOR OF KENYA

by
Steven Jaffee

March 1987

IDA Working Paper No. 83

Contract Farning in Africa Project
Working Paper No. 7

Contract Farming in Africa,

Prepared for the research project, Clark
AID, and carried out by the
funded by the Africa Bureau, Cooperative
Anthropology
University/Institute for Development
Resource Systems Analysis (SARSA)
Agreement on Settlement and
 PREFACE

This Working Paper is part of a larger research project on

Contract Farming in sub-Saharan Africa conducted by the Clark
University/Institute for Development Anthropology Cooperative
Agreement on Settlement and Resource Systems Analysis (SARSA) for
the Africa Bureau of the US Agency for International Development
(AID).

For purposes of this study, contract farming is defined by

three fundamental characteristics: (i) a futures or forward
market in which a buyer or processor commits in advance to
purchase a crop acreage or volume; (ii) the linkage of product
and factor markets insofar as purchase rests on specific grower
practices or production routines and input and/or service
provision by buyer-processors; and (iii) the differential
allocation of production and marketing risk embodied in the
contract itself. Contract farming includes, therefore, the
large-scale nucleus-estate/outgrower schemes associated with, for
example, palm oil in West Africa and sugar production in Kenya;
the parastatal, export-oriented smallholder achemes associated
with tea, tobacco, and coffee in Central and East Africa; and a
multitude of private schemes producing fresh fruits and
vegetables for canning, drying, and direct export to
international markets.

Contract farming in a variety of institutional forms has

been present in North America since the 1930s, but it has more
recently become of increasing importance in Third World states,
particularly throughout much of Africa. The objective of this
study is to assess the form, organization, and impact of a
diversity of contracting arrangements in sub-Saharan Africa,
based on both secondary literature and field research in seven
countries (Gambia, Nigeria, Ivory Coast, Ghana, Kenya, Malawi,
and Senegal). The case studies have been carefully selected to
represent the primary commodities and diversity of institutional
forms of contract farming. A final report, based in part on the
representative case studies, will indicate the conditions under
which contract farming emerges; assess the distribution of costs
and benefits to the principal actors, including growers; and
evaluate the role of contract farming with respect to donor and
host-government policies, technology transfer, and institutional
development.

Michael Watts and Peter Little

 TABLE OF CONTENTS

INTRODUCTION .............. ......................... 1

THE FRENCH BEAN CONNECTION . ......... ........ 1

Introduction .......... ...................... 15

French Beans .......... ...................... 16
Njoro Canners .......... ..................... . 23
Saupiquet............ ......................... 25
The Origins of the Kenyan Project .... ............ 26
Basic Organizational Structure and Components ....... 38
General Performance Indicators ...... ............. 44

Impact of the Project ........... ..................... 48

Evolution of Performance and Institutional Arrangements . 49

Concluding Remarks ......... ................... 60

THE DEVELOPMENT OF THE KENYA/UNITED GDOM TRADE IN "ASIAN

VEGETABLES" WITH PARTICULAR REFEREN 0 PROBLEMS OF
COORDINATING PRODUCTION AND EXPORT MAHAET.... ........... 67

The Origin of the Trade ........ ................ 70

The Market for Asian Vegetables ..... ............. 71
The Asian-Vegetable Marketing System .... ........... 74
The Asian-Vegetable System in Kenya ..... ........... 81
A Contractual Scheme for Asian Vegetables .. ........ 96
Concluding Remarks ......... ................... 110

CONTRACT FARMIhG, MARKET CONDITIONS, AND THE VEGETABLE

DEHYDRATION INDUSTRY IN KERYAP 1964-1982.... ........... 117

Introduction .......... ...................... 119

1964-1972: Subsidized Trial and Error .... ............ 121

1973-1974: Enter the Experts....... ................. 126

1975-1982: Pan African Vegetable Products, Ltd... ........ 129

Concluding Remarks ....... ................. 146

INTRODUCTION

The Study of Contract Farming: A Note of Caution

Contract farming is an institutional form whereby

agricultural production is carried out according to an agreement
between farmers and a buyer which specifies certain production
and marketing arrangements. For many years contract farming has
played a prominent role in certain agricultural sub-sectors in
North America and Western Europe. Production contracts are the
dominant form of farmer-buyer coordination in these regions for
such commodities as poultry, seed crops, processing vegetables,
sugarbeets, and fluid-grade milk. In recent decades, contract
farming has become a more prominent feature of African
agriculture. Crop-specific contractual schemes have been
developed by both private firms and specializea government
agencies, sometimes with concessional funding provided by
international development agencies.

Contract farming in developing countries has been viewed

from two distinct perspectives. One perspective sees contract
farming as an institutional innovation developed to increase
agricultural productivity and specialization and to improve
coordination between production and marketing. It views the
development of specialized contractual schemes as a compensating
responae to imperfections in factor and product markets and as an
attempt to fill an organizational vacuum left by a poorly
functioning public agricultural administration. These
contractual arrangements are seen as offering a series of
potential advantages to both farmers and buyers, relative to
either dealing strictly in spot markets or developing integrated
production/mnrketing operations. Contract farming is viewed as a
potentially useful vehicle for improving small farmer
productivity and increasing rural incomes.

Critics of contract farming tend to view it as an

institutional innovation developed by powerful economic and
political groups to increase agricultural productivity and
specialization, to appropriate the gains from these improvements,
and to pass on the relevant costs and risks to farmers or third
parties. This perspective sees contractual arrangements being
designed to create or strengthen market imperfections so that
private interests gain at the expense of social misallocations of
resources. This perspective posits a zero-sum process of the
following nature. The contracting firm benefits by gaining
greater control over a crop than possible under spot market
conditions, yet without incurring most of the costs and risks of
actual investment in production. These buyers are placed in a
monopsnistic position, able not only to dictate prices, but also
to manipulate quality standards to make adjustments for raw

I
material and market imbalances. Farmers, on the other hand, are
seen as getting locked into a dependent relationship with the
buyer, made more vulnerable by their increased crop
specialization and use of material inputs. Farmerc may begin
losing their autonomy as the contractor begins controlling many
agronomic decisions. Writers representing this position reject
the possibility of small-scale farmers benefiting from contract
farming.

Contract farming is a highly complex subject requiring

analysis of a range of technical, economic, and sociopolitical
factors. The institution exhibits wide variations in structure,
participants, operating arrangements, and impacts. For this
reason caution is necessary in making general comments about the
past record of contract farming, let alone about its wider
potential and limitations as a vehicle for development of
agriculture and agro-nased industries. Making generalizations
from individual case studies may thus be hazardous. One's
insights into a particular form of organization and contracting
procedures and into a particular commodity sector of one country
may be quite powerful, but the strength of one's arguments
dissipates as one moves across organizational, commodity,
country, and temporal space. This is frequently not acknowledged
by either the outspoken proponents or critics of contract
farming.

Contract Farming in Kenya

Within Africa, contract farming has been most extensively

developed in Kenya. Production contracts have been extended to
both small-scale and large-scale farmers. Contract farming plays
an important role in the Kenyan tea, sugar, tobacco, oilseed,
horti.culture, poultry, ard beer-making industries. Raw materials
produced under con.ract are thus used in both export and
import-substitution industries.

Within Kenya, interest in contract farming as an institution

of development has appropriately focused on schemes incorporating
primarily small-scale farmers. Since colonial times, the
administration of agricultural support and marketing in Kenya has
had a larqe-farmer bias. Large-scale farmers in Kenya have
typically had greater access to inputs, credit, extension and
research advice, market information, and alternative distribution
channels than has been the case for smallholders. Given
constraints in land availability, prevailing demographic
patterns, and the political risks associated with highly unequal
distributions of wealth and income, agricultural development in
Kenya must be oriented toward greater intensification of
production and improvements in the producti.ity and incomes of

2
small-scale farmers. In c-ertain circumstances, these goals may
be approached through the mechanism of contract farming.

Within Kenya, there are a number of crops for which

smallholder contracting has been fairly significant. These crops
include: tea, sugar, tobacco, sesame seed, sunflower seed, French
beans and other vegetables, and horticultural seed. The extent
of smallholder participation in contractual schemes is estimated
in the chart below:
------------------------------------------------------------
Smallholder Contract Farming in Kenya

Crop(s) Firm(s) Farmers

----------------------------------------------------------------
Tea KTDA 150,500(a)
Sugar MSC; others(b) 35,000(c)
Oilseeds OCD(d): Ufuta(e) 34,000(l)
Horticulture (g) Njoro Canners; others(h) 21,500(i)
Tobacco BAT 10,0(0

Total 251,000

-------------------------------------------------------
(a) The number of licensed growers under KTDA in 1986 was
150,414. However, there is evidence that additional farmers are
growing tea without a license.

(b) Includes Associated Sugar Company, Muhoroni, and others.

(c) This is an estimate. During 1985-86 MSC and ASC contracted a

combined total of 29,000 smallholders. We do not have data for
the other firms.

(d) Oil Crops Development Ltd.. East African Industries holds

45%, CDC holds 35%, and the IFC holds 20%. The project was
initiated in 1984 for sunflower and rape seed production under
contract.

(e) Ufuta Ltd. is a subsidiary of Kenya National Hills Ltd and a

sister company of Elianto Kenya Ltd. The latter had an
unsuccessful sunflower contracting project in the late 1970s.
This project is oriented toward sesame seed production at the
coast.

(Y) The OCD project intends to incorporate 20,000 smallholders by

1988 while Ufuta's target is 14,000. We have no data on the
number of farmers actually under contract in 1986.

3
(g) Includes fruits, vegetables, and flowers, although most
contracting is for vegetables for processing or export.

(h) Includes seed companies such as Kenya Seed Company, Regina

Seed Company, and Hortitech and exporters such as Kenya
Horticultural Exporters and Makindu Growers and Packers.

(i) A conservative estimate based on 15,500 farmers for NJoro

Canners, 3500 farmers for the various seed companies, and 2500
farmers with contractual links to other firms.

If the oilseed projects do reach their proposed scale in the

late 1980s and if the other schemes simply maintain their
participation rates, then up to a quarter million smallholder
farmers will be producing under contract in the late 1980s. This
represent approximately 16.7% of the 1.5 million smallholder
families in Kenya. The proportion of contracted households is
probably somewhat less than this figure as some farmers may grow
more than one crop under contract. For example, several
contracted horticultural farmers also grow tobacco or tea.

Numerous large-scale farmers also operate under production

contract in Kenya. BAT Kenya Developments Ltd. has contractual
arrangements with a limited number of poultry growers. Several
hundred medium- to large-scale fruit and vegetable growers
produce under contracts with processors and fresh produce
exporters. Kenya Breweries Ltd. has a total of 17,500 ha of
malting barley being grown for it under contract with large
farmers. Oil Crops Development Ltd. intends to have 5000
largeholders producing sunflower and rape seed on 60,000 acrc by
1988.

Looking across the different agricultural sub-sectors, one

finds that the majority of existing contract farming schemes are
linked to a processing operation. Many schemes also feature the
participation of a European company, either as owner/managers of
a scheme or through management and/or marketing contracts with
locally owned firms. Many schemes are joint venture investments
involving private management and Kenyan Government equity
participation.

Literature Review

There is a sizeable literature on contract farming in Kenya.

(See page 11.) This literature provides insight into a range of
issues, including: the problem of incentives and controls for

4
staff and farmers, the participation of the contractor in the
production process, the transfer of technology, the generation
and uses of income, the impact on labor and land markets, and the
potentially central role of the State. However, this literature
deals almost exclusively with three schemes: i.e., KTDA's
smallholder tea project, Mumias Sugar Company, and the BAT
tobacco project. The large schemes of KTDA and Mumias have
received by far the most attention.

On the other hand, there has been no in-depth research and

liLtle reference to smEller or less formal schemes, to schemes
that failed or were associated with unsuccessful companies, or to
schemes that did not have considerable government backing. There
has also been no research on the considerable number of contract
farming schemes developed for horticultural or oilseed crops.
Thus, while the volume and quality of research on contract
farming in Kenya is arguably the best in Africa (or even amongst
developing countries generally), this literature provides
extremely few generalizable propositions and little or no insight
into several potentially important dimensions of contract
farming.

The literature on the KTDA, Mumias, and BAT schemes does

feature a consensus on a few issues. First, there is evidence
from all three schemes that contract farming leads to an increase
in cash incomes. Contract smallholders are economically better
off than non-contracted smallholders in their area and the
difference can at least be partially attributed to participation
in the scheme.

Second, there is evidence that the income stream generated

from contract fai ing is unevenly distributed. This has
contributed to increased socioeconomic differentiation in the
contracted areas. The differential stream of benefits relates
substantially to the prescheme landholdings of participants and
nonparticipants as well as to the availability of alternative
sources of income and employment for households. As contractors
have set minimum landholding and production scale requirements,
the very poor have generally been excluded from such schemes
other than through wage labor opportunities on contracted farms.

Third, the literature strongly suggests that the impact of

contract farming will vary with organizational and production
structure as well as with preexisting conditions and simultaneous
socioeconomic changes. For example, while active farmer
participation in the production processes for tea and tobacco has
led to real "learning effects" which have "overspilled" into iood
production, this has not been the case for sugar where the farmer
is more passive in the production process. While landholding
sizes and the economies of scale in mechanical plowing and

5
harvesting have resulted in land competition between sugar and
food crops, such competition has not generally been important in
the tea and tobacco areas due to previous landholding patterns
and the smaller scale of contracted crop plantings. Crop and
trade diversification has been common in tea areas, while the
sugar zone resembles a monoculture economy.

Fourth, there is fairly wide consensus that smallholder

farmers are not adequately represented or protected by
intermediary organizations. In the case of tea, the grower
committees and the factory boards tend to be controlled by larger
and more prosperous farmers. In the case of sugar, the Mumias
Outgrower Company has not been an effective intermediary. Local
MPs typically emerge as the "voice" of farmers.

Fifth, it is a common finding that in male-dominated

societies a contractual scheme may adversely affect the position
of women. In both the sugar and tobacco schemes it has been
observed that men typically gain control over income while the
women are relegated to perform difficult and unpaid routine work,
such as weeding.

While the literature on contract farming does provide

insight into several important issues, the literature features a
sample that is biased in the direction of large, state-supported,
formal, and successful schemes. As a result several dimensions
of contract farming are given little or no attention.

For example, the existing literature frequently leaves the

impression that contract farming arrangements are monolithic
structures, stable over time. In fact, contractual arrangements
may evolve gradually as managers, staff, and farmers adjust their
behavior arid formal structures to counter inefficiencies and
pu.sue new opportunities. The exclusive locus on highly formal
contract schemes has led to limited analysis of the possible
transitions that occur in production/marketing arrangements
between contractual and quasi-contractual links. The need for
formal contracting may be related partly to the absence or
presence of trust between farmers and buyers. Many contract
farming schemes arc- not "greenfield investments" involving new
crops, new farmers, and new buyers. Contract farming may involve
farmers with prior experience with the crop, entering into a more
inx.ensive, multifaceted relationship with an existing or new
buyer.

Also, the existing contract-farming literature in Kenya

describes contract enforcement problems large-ly in relation to
quality control and to credit recovery by thE firm. In each case
examined, the contracting firm has had a de facto monopoly over

6
the purchase of the crop. Alternative market outlets for farmers
either do not exist or are not remunerative.

Contract enforcement is a more general problem. It is

problemaLic where one or both contracting parties benefits from
acting opportunistically and where such behavior is difficult to
detect. Such opportunistic behavior may relate to direction of
sales/purchases, quality manipulation, and qtantity cheating. In
many cases of contract farming the "leakage" of raw material out
of the project and into alternative distribution channels may be
a major problem. The relative merits of sales through
alternative outlets will vary, depending on seasonal market
changes, the physical location of farmers vis-6-vis the
alternative outlets, and the tiervices provided by competing
marketino agents. The development by the contractor of measures
to guard igainst leakage may be a key dimens-ion of a contract
scheme. Both farmers and buyers may breach contractual terms
related to the quality of the product.

Due to unforeseen circumstances (i.e., weather change), poor

production practices (i.e., careless harvesting), and/or deceit
(i.e., hiding subquality produce on the bottom of a carton), the
quality of a farmer's crop may be below standard. This may or
may not be detected by the firm. In some cases the firi, will
chose to ignore the quality problems. In other cases it will
make price deductions or reject the crop entirely. Farmers may
be able to connive with contractor staff to allow subquality
produce to go unnoticed. On the other hand, the contractor may
be able to use quality control procedures to adjust quantity
imbalances. Particularly where quality is difficu.t to measure
and grading and sorting are performed by company staff, farmers
may be surprised by produce inspection results. FPrmers and
contractors (or their staffs) may attempt to cheat one another
with regard to the quantity of the contracted crop. Farmers may
obtain seed or other inputs outside of the contract and then sell
the extra crop with the contracted crop. Company staff may be
given incenives by farmers to overweigh their crop.
Alternatively, staff acting on their own or under company orders,
may underweigh farmer deliveries.

Further, the literature on contract farming in Kenya notes

that changes in product market conditions affect the
profitability of schemeZ and the level of benefits accroing to
farmers, but there may be cases where such market changes may
undermine the viability of the contracting scheme itself.
Adverse market conditions may undermine the contractor's
financial position, preventing it from raising producer prices in
line with production costs or redu'cing the scope of its services.
Highly favorable market conditions may lead to the emergence of
competing contractors or marketing agents offerin g farmers terms

7
that the original contractor is unable or unwilling to match.
Some market changes may undermine the comparative advantage of
the entire venture and lead to closure even when the contract­
farming component was performing adequately.

Contract Farming in Kenyan Horticulture

These three dimensions of contract farming schemes--their

evolving organizational structure, their vulnerability to
opportunistic behavior by one or both parties, and their critical
links to the downstream market--are all readily apparent in
several of the contractual schemes which have been attempted in
the horticultural sector of Kenya. Horticulture has been one of
the most dynamic sectors in the Kenyan economy in recent years.
It has been driven by a growing export trade, together with rapid
rates of increase in domestic trade and consumption.
Horticultural exports, comprising fresh and processed fruit and
vegetables as oell as flowers, are now the country's third
largest source of foreign exchange after coffee and tea. The
sector features a wide range of organizational structures and
mixtures of private and public investment. Large integrated
production/marketing operations have played an important role in
the development of the sector and these organizational forms
remain dominant for flowers, pineapples, and strawberries.

However, for several horticultural crops and commodities

there have been numerous attempts at organizing small- and
medium-scale production under contract. For different
horticultural crops there have been as many as twenty different
contract farming schemes proposed or attempted over the past two
decades. In the past decade alone, there have probably been at
least ten dilferent schemes developed to have farmers grow French
beans under contract for processing or fresh export. Many of
these schemes failed or had only short-term success. At present,
there are at least four schemes which feature small and
medium-scele farmers growing vegetable and flower seed under
contract. Since 1980 there have been at least three attempts at
having smallholders grow "Asian vegetables" under contract for
exporters through the intermediation of cooperatives. Since the
late 1970s there have been several attempts to organize
smallholder flower production under contract.

In each of the attempts at contract farming in horticulture

the reLationship between buyer and farmers has gone well beyond a
strictly marketing agreement. In some cases the involvement of
buyers in the production process has been substantial. In most
of the cases farmers had experience growing the crop prior to the
development of the contracting scheme. However, inefficiencies
in product and input markets made production contracts attractive

8
to farmers. In many cases the buyer faced competition for the
crop and contracting was seen as a method of lowering
uncertainties about rpw material supplies. Still, leakage of
produce and poaching by competing firms have typically been
problematic. In contrast to the very large contract farming
schemes, several horticultural contractors have lacked
substanLial staffs or access to seconded governmental staff.
They have thus had to rely more substantially on local stafi or
agents or on existing cooperative societies. Most of these
schemes have involved no government funding and limited
government involvement.

We have chosen three horticultural contracting schemes for

in-depth case study anolysis. One case concerns the vegetable
dehydrating company, Pan African Vegetable Products Ltd. (PVP).
This is the first case of smallholders growing under contract
with an agricultural processing firm in Kenya. The project was
initiated in 1964 and, with numerous changes in ownership and
management, carried on until 1982. The smallholder contracting
scheme of PVP was largely successful, yet the project experienced
continuous financial losses as a result of processing and
marketing problems and the insufficiency of large-farmer supplies
of raw materials.

Our second case deals with "Asian vegetable" production and

marketing and the contractual scheme attempted by Kenya
Horticultural Exporters (KHE). KHE has been Kenya's leading
exporter of fresh fruit and vegetables for nearly two decades and
has on several occasions entered into production contracts with
small and medium-scale farmers. The company's scheme for
contracting smallholder "Asian vegetable" producers was
successful for a few years, but the project was not sustainable
due to the larger competitive environment for "Asian vegetable"
production and marketing in Kenya. The scheme contributed to
substantial increases in smallholder production which the
contracting company was only temporarily able to benefit from.

Our third case is the most formal horticultural contracting

scheme. It is that of Njoro Canners, a processor of French beans
which has production contracts with over 15,000 smallholder
farmers in western Kenya. The Njoro Canners project was
initiated in 1982 in the wake of numerous unsuccessful prior
attempts at contracting western Kenya farmers to grow French
beans for processing. Seventy percent of the farmers
participating in this scheme are women, growing French beans on
only 1/20th of an acre. While experiencing numerous technical,
organizational, and political problems, this project has managed
to survive, produce a high-quality export product, and provide
additional sources of income and employment in an economically
deprived area.

9
 A review of the literature on the tea, tobacco, and sugar
schemes provides insight into the forms of contract farming and
its potential impact. The more "high profile" schemes exhibit
substantial variation in the nature of the production process anc
sales arrangements. For example, tobacco production is carried
out under a "supervision-intensive" regime and based solely on
outgrowers. BAT's comprehensive extension service is responsiblE
for instructing farmers and monitoring their behavior throughout
the growing and curing processes. All necessary inputs are
provided on credit. However, the tobacco farmer is responsible
for carrying out all tasks. Hired labor is uncommon. Farmers
are paid cash on the day of delivery according to quantity and a
diverse grading scale.

In contrast, sugar production is done both on estates and or

outgrower farms. Even with the outg-3vers, the company carries
out many production tasks either mechanically or through the use
of work gangs. The farmer's main task is weeding and even this
may be carried out by hired labor. Farmers have no post-harvest
role and payment is based strictly on volume.

Various researchers see three strata of farm households

emerGing in the contract farming areae. The top stratum is that
of the "capitalist farmers" who have relatively high income,
derived partly (or largely) from trade and salaries. They rely
heavily on hired labor on their farms. These farmers can use the
additional income from the contracted crop to invest in shops,
taxis or production inputs. The second stratum, the "middle
peasants," derive income from contracted as well as other crops.
They use both family and hired labor. The income generated by
the cash crop is used for school fees, housing improvements, and
consumer goods. The third stratum consists of very poor
households with small holdings and relying solely on family
labor. Casual wage labor may be their sole source of cash
income. They may have to reduce their holdings to obtain
required cash. These farmers can produce cash crops under
contract only at the expense of lood production, thus increasing
their vulnerability. As minimum landholdings and/or production
scales are set by the contractors, these poor farmers may be
excluded from the projects even if they wished to participate.
Sources include:

Allen, G. (1983) "The Development of the Mumias Sugar

Corporation" Oxford Agrarian Studies.

Barclay, A. H. (1977) The Mumias Sugar Project. A Study of

Rural Development in Western Kenya. Ph.D. Thesis, Colombia
University.

Buch-Hansen, M. (1980) "Agro-Industrial Production and

Socio-Economic Development: A Case Study of KTDA Smallholder Tea
Production in Buret, Western Kenya." Working Paper No. 11,
Institute of Geography, Roskilde University Centre.

Buch-Hansen, M. and H. Marcussen (1982) "Contract Farming and

the Peasantry: Cases from Western Kenya. Review of African
Political Economy, (Jan/Feb).

Buch-Hansen, M. and J. Kieler (1983) The Development of

Capitalism and the Transformation of the Peasantry in Kenya"
Rural Africana. (Wint-Spr).

Mulaa, J. (1981) "The Politics of a Changing Society: Mumias"

in Review of African Political Economy.

'Nyong'o, P. (1981) "The Development of a Middle Peasantry in

Nyanza." Review of African Political Economy.

Shipton, P. (1985) Land, Credit and Crop Transitions in Kenya:

The Luo Response to Directed Development in Nyanza Province.
Ph.D. Thesis, Department of Anthropology, Cambridge University.

Williams, S. (1985) "The Mumias Sugar Company: A Nuclear Estate

in Kenya" in Agribusiness and the Small-Scale Farmer: A Dynamic
Partnership for Development by Williams and R. Karen. Westview
Press.

ii
 THE FRENCH BEAN CONNECTION:
FRAGILE SUCCESS OF A SMALLHOLDER CONTRACT FARMING PROJECT
IN WESTERN KENYA

Contract Farming in Africa----- Kenya Case Study #1

13
Introduction

This report examines several features of a privately managed

production and marketing operation that has linked up to 15,000
smallholder farmers of a relatively deprived and very densely
populated area of Kenya into an international market for a
specialized high-quality agricultural product. Njoro Canners,
Ltd., is a locally owned firm acting through a marketing,
management, and technical assistance contract with the French
company Saupiquet and producing a top-grade canned French-bean
product. The study most closely examines the raw-material­
procurement dimension of the project. This is a contract-farming
scheme with emallholder farmers. We explain the rationale for
this contract-farming scheme and trace its historical background
and organizational features. We also examine various aspects of
the project's performance and impact. The analysis of the Njoro
Canners project is set within the context of the West European
market for French beans and the wider development of French bean
production and marleting in Kenya.

The case of Njoro Canners is one of fragile success. The

project followed upon several relatively unsuccessful attempts in
Kenya to have farmers produce French beans under contract for
processing. In the first year and a half of the project it
appeared that low farmer productivity and weak capacity to
enforce contracts would doom it. Several important technical and
institutional adjustments saved it, enabled it to expand, and put
the contract-farming element on a sounder economic footing. At
this juncture the projecL was shaken by internal and external
efforts (both legal and illegal) to redistribute project eara­
ings. Adjustments were made to reduce further risks of this
nature and the project has continued to expand its sales,
employment, income generation, and farmer participation. The
future of the project is uncertain, noL only due to the fragility
of the company's organizational structure or its potential
competition from other Kenyan firms, but also possibly due to
technical developments in Europe that could virtually negate part
of Kenya's comparative advantage in French-bean production.

The study is organized as follows: We begin by making some

general comments about French beans, their European market, and
the overall pattern of French-bean production and marketing in
Kenya. Moving on to the Njoro Caariers case s':dy, we first
examine the French market for canLed green beans and provide some
background information on Saupiquet. Next we trace the origin of
the project by discussing Saupiquet's prior experience with
French-bean contracting in Morocco, its trade ties to Kenya prior
to the Njoro Canners project, and its feasibility study for the
Kenya project. The next section outlines the physical and

15
socioeconomic characteristics of the site for the contracting
scheme. This is followed by a broader analysis setting out the
rationale for this type of organization for raw material produc­
tion. The project's actual organizational and contractual
structure are then discussed. Next comes an overview of project
performance according to a range of indicators. This review of
basic organizational features and performance reveals that there
have been considerable variations over time. One then needs to
explain these performance variations and see whether they were
linked to structural changes within the scheme. This we do in
the next section where we view the processes of project develop­
ment. We close witn some final comments about the future
prospects for the project and some lessons that the project
suggests.

French Beans

The French bean (Phaseolus vulgaris) is one of the names given to

the pods of the plant species Pheseolus. Other names commonly
given to these pods are green beans, snap beans, string beans,
bobby beans, and haricot beans. There are many hundreds of
different varieties of Phaseolus vulqaris. These varieties may
have different characteristics with regard to their production
and their quality. Important differences may relate to the
following:

Production: Quality:

Color and application rate of Length of uod

seed

Size and shape of plant Width of pod

Color of leaves and flowers Curvature of pod

Length of time to maturity Texture of pod's skin

Tolerancy zo bean rust and String development in pod

halo blight

Rate of seed development Color of pod

Rate and pattern of yield

The French bean is thus potentially a highly heterogeneous

product. Varietel selection may be a complicated process.
First, it involves a ma-ching of quality characteristics with
consumer preferences, or the requirements for processing or
effective distriLution. French beans are consumed in various
forms, including fresh, canned, frozen, or dehydrated. Certain

16
varieties have characteristics entirely unsuitable for some forms
of consumption or processing. Even within categories for
consumption, there are grading schedules outlining quality
specifications for individual pods as well as acceptable quality
tolerances (i.e., variances in quality).

Varietal selection will also need to relate varietal

production characteristics with the ecological conditions,
agronomic practices, and even the socioeconomic features of the
area where the crop will be grown. For example, in recent years
varietal development in Western Europe has concentrated on
cultivars that are amenable to mechanical harvesting. Many of
the older French-bean varieties have a yielding pattern consist­
ing of several dispersed flushes over a period of three to six
weeks. For these varieties mechanical harvesting is not economi­
cal. The harvesting machine acts as a comb, pulling the plant
completely from the ground. Using a mechanical harvester for the
multiple-flush varizties would result in very low yields.
Harvesting of these older varieties must be done by hand, and
labor requirements per acre of beans are very high. European
producers, faced with rising labor costs, required single-flush
bean varieties that could be mechanically harvested.

Fortunately for countries with relatively low labor costs,

the single flush mechanically harvested French beans are typi­
cally larger in length and width than the pods of the "old"
varieties, and frequently have a rougher ukin surface. While
these characteristics may be suitable for some forms of process­
ing or meet the preferences of certain consumers, they may not be
suitable for other uses ar market segments. Certain consumers or
institutional users of Fiench beans have retainer a preference
for small and smoothly textured varieties. For this market
segment there exists premium demand for particular varieties and
certain quality characteristics. Probably the most important
specificationr by these consumers relates to the width of the bean
pods. Certain groups offer premium prices for "extra-fine"
beans, i.e., those with a width of 6.5 mm or less. Other groups
may have preference for "fine" beans, i.e., those with a width
less than 9 mm but more than 6.5 mm. Beans of this size cannot
be mechanically harvested.

French-Bean Production an Marketing in Kenya

Over tne past two decades the French bean has become an
important crop in Kenya. While grown for both fresh sale and
processing, the main ±mpetus for production has been an expanding
export market. Since the early 1960s, Kenya hps exported "fine"
and "extra-fine" Frenc- beans to Western Europe. While this
trade was initially targeted toward high-class catererr and
department stores, over the years the air-freighted Kenyan beans

I1
have become en item distributed by supermarket chains and
purchased by middle class consumers. Kenyan exports are concen­
trated in the October-May period when European production of
French beans is limited by adverse weather conditions. Market
prices for French beans during this period are substantially
higher than during the European summer when local supplies are
plentiful. Still, a certain level of demand for the Kenyan
product is retained during the summer months by caterers and "up­
market" greengrocers.

Kenyan French-bean exports have been aimed largely for sale

to France, Belgium, and the United Kingdom. Consumers in the
first two countries have a preference for "exLra-fine" beans
while U.K. consumers prefer "fine" beans. Consumers in the
Netherlands and West Germany prefer the larger bobby beans, which
can be widely and cheaply procured from Spain or Egypt. Kenya's
main competition for the "off-season" markets for "fine" and
"extra-fine" beans comes from seversl West African countries that
have long-standing trade ties with France.

The growth in Kenya's exports of French beans can be seen in

the following figures:

Table 1: Kenyan French-Bean Exports

Year Tons

1968 109
1972 642
1976 2324
1980 4965
1982 6306
1983 6447
1984 7094
1985 6558

Source: HCDA trade figures

French beans have also been grown in Kenya for processing.

For many years several firms have been canning them for sale in
both export and local markets. Generally the high levels of
protection in the domestic food-processing industry together with
high production costs have made the canned products uncompetitive
in world markets buL highly profitable on the local market.
Canning companies have generally purchased beans from wholesalers
or directly from farmers in times of market surplus. In addition
to canning, French beans have been processed in Kenya through
dehydration. During the 1960s and 1970s a dehydration factory at
Naivesha processed French beans for export to Western Europe.
This firm entered into loose production contracts with farmers.

18
Both the dehydration factory and each of the canning companies
have experienced considerable problems in obtaining sufficient
quantities of raw material. The prices and other terms that they
have offered French-bean growers have frequently been
uncompetitive with those offered by the fresh market. The Njoro
Canners project contrasts with these other processing operations.

Kenya's comparative advantage in French-bean production

rests on two main factors: its ecology and its relatively low
labor costs. Limited seasonal variations in temperature and day
length allow French-bean production to be extended throughout
most of the year in Kenya. French beans cannot survive frost and
thus can be grown only under controlled-temperature conditions in
most parts of Western Europe during the winter. They are grown
in areas of Kenya with altitudes ranging from 1000 to 2000
meters, which are only rErely subjected to frost conditions.
Various areas of Kenya have soils that are highly suitable for
French-bean production. Furthermore, the presence of trees or
bushes on many farms provides natural wind-breaks for the French­
bean plants.

Production of "firie" and particularly "extra-fine" French

beans is not economically viable in most parts of Western Europe,
given the high labor costs that wouiJ be incurred in harvesting.
Harvesting of an acre of French beans may require 15-20 people
over a period of three to six weeks. Once pods are formed they
grow at a rapid rate. To obtain "extra-fine" beans, picking must
be done every day. The result is that harvesting costs will make
up a high proportion of overall production costs Yor French
beans. Where labor costs are relatively low, one may still
obtain an economic return on a crop even when such labor time is
allocated. In Kenya the daily wage for French-bean pickers
ranges from Ksh 10 to 22, equivalent at the present rate of
exchange to $0.63-1.38 per day.

French beans are produced in Kenya by both small-scale

farmers under rain-fed conditions and l~rcjer commercial farmers
under Irrigation. In recent years 4000-6(00 smallholders have
been engaged in French-bean production for the fresh export
market alone. These farmers typically grow 1/2 to I acre of
French beans as part of a mixed-farming pattern including maize,
dry beans, dairy cows, and other crops. Such smallholders are
based in Athi River and in various sites in Central Province.
Larger scale producers for the fresh export market may number
100-150. These farmers may have up to 20 acres of beans under
production with harvesting being done on 4-5 acres at a single
time. These farmers typically grow French beans to supplement
incomes from salaried employment or to improve the cash flow
position of farms oriented primarily to tea or coffee production.
Some larger farmers are specialist horticultural growers. Larger

19
French-bean farmers are common at Lake Naivasha, Thika, and Athi
River.

A wide range of institutional arrangements exists for

farmers in the marketing of their output, from essentially market
transactions, through quasi-contractual and contractual sales,
and on to vertically integrated operatiors. A full examination
of these different marketing arrangements is not possible here.
We merely summarize the main features of three alternative
channels.

Several thousand smallholders in the Gatundu and Makuyu

areas of Central Province are engaged in French-bean production.
Some farmers have grown this crop since the early 1970s. More
than a dozen exporters are fairly regular bean buyers in the
area. Most of these firms recruit local people to act as
intermediaries recruiting farmers and organizing collection and
farmer payments. Some intermediaries work with more than one
exporter. Exporters may send their trucks to the area three or
four times per week during the main export season. Prices are
set for the season with one price for "extra-fine" beans and
another price for "fine" beans. These "fixed" prices may be
subject to short-term adjustment as a result of changing market
conditions. Exporters provide no seeds or other inputs and are
not in a position to provide any technical advice. The
intermediaries distribute cartons I.o farmers and arrange the days
for the farmers to deliver filled cartons to a store or stall.
For his efforts the intermediary will take a few shillings per
carton commission. Payments to farmers are made fortnightly.
Farmers may deal with several different intermediaries (and thus
exporters), shifting their sales in light of short-term higher
price offers being made by competing exporters.

Kenya's largest exporter of fresh fruit and vegetables is a

lirm called Kenya Horticultural Exporters (KHE). In recent years
the company has exported up to 2500 tons of beans annually. The
bulk of its supplies are obtained on contract from large and
small growers. In 1986 KHE had 150 farmers growing beans under
contract. KHE provides seeds and chemicals on credit to be
deducted against the delivered crop. The company has two
experienced horticul-turists who can advise farmers on production
problems, and it employs several people who assist farmers with
proper grading and packing. Farmers are paid a fixed price for
the full export season. Prices are changed only in exceptional
circumstances. During the peak export season KHE trucks may
collect produce five or six days a week. Farmers are paid
whenever they want. Some receive payment weekly, others
fortnightly or monthly.

20
 While KHE may directly contract with only 150 bean farihers,
operating under its bean procurement "umbrella6 are probably 500
or more farmers. Several of KHE's farmers have theLr own
subcontractors. One contract farmer in Mwee has developed a
procurement network of over 200 small-scale farmers in the area.
The subcontractors, most of whom are women and many of whom grow
the beans on plots provided by the National Irrigation Board,
typically have 1/4 to 1/2 acre under beans. The KHE contract
farmer provides seed, fertilizers, and chemicals on credit to
"loyal" subcontractors. He maintains the collection stations
where KHE trucks pick up supplies. The contract farmer takes a
margin of 5-10 percent of KHE's contract price.

The export company Homegrown presents another method of raw­

material procurement. Homegrown actually has two separate
systems. With twenty large-scale farmers he maintains seasonal
contracts. He pays premium prices over those offered by
competitors, but his quality standards are far more rigid. He
employs fifty graders who are actually brought to the contracted
farms during harvesting. These graders go through the fields
advising and monitoring the pickers. They check the quality and
weights of cartons before they leave the farm. The contract
farmers receive seeds and some chemicals on credit. Homegrown's
manager, an engineer by training, has designed small-scale dams
for ten of his farmers. Producer payments are made weekly.

Homegrown simultaneously operates a different system for raw

material procurement from smallholders. He maintains two
collection centers in Mwea. Small-scale farmers bring their crop
in bags, grade them, and sell them in bulk form to the company.
Depending upon the regularity of a farmer's sales to the company,
she may be paid cash on the spot or else paid weekly. No inputs
or technical advice are provided. Transporters take these beans
to a company packing/cold-storage unit where the beans are
rechecked for quality and packed into cartons.

The production and marketing of French beans has had a

number of beneficial impacts. One immediate benefit is the
generation of foreign-exchange earnings. In the early 1980s the
foreign-exchange earnings for fresh French-bean exports have been
the following:

Table 2: French-Bean Export Earnings

1981 Ksh 59.8 million

1982 63.1
1983 70.9
1984 78.0
1985 72.1

21
Source: HCDA Export Data

These figures are actually "minimum" export values,

calculated by taking the Government's minimum export prices and
multiplying them by the volume of sales. Actual foreign-exchange
earnings are probably 10-15 percent above these minimum values.

A second major benefit has been the generation of cash

income opportunities for small-scale farmers. French beans are
an ideal smallholder crop given their labor intensivity, their
short production cycle (i.e., three months from planting until
completed harvest), and the small planted acreages needed to
obtain a good supplemental cash income. Based on exoorter
reports about their bean procurement systems, we estimate that 60
percent of the beans that are exported are produced by small­
scale farmers. If one assumes for 1987 that Kenya will export
7000 tons of French beans and one takes a rough average producer
price of 10.4 shillings/kg and deducts I sh./kg for the
middlemen, then smallholder gross income for beans this year will
be Ksh 37.11 million.

French beans have also been a lucrative source of income for

many large-scale farmers and have helped coffee farmers to
overcome cash flow problems associated with delayed payments for
that crop. Even when using conservative estimates for yields and
producer prices, large growers can obtain a gross income of Ksh
19,600 per acre against production costs (not including
depreciation on equipment) of about Ksh 10,653. This net income
of Xsh 9000 is for only a three month crop. At least three
separate crops per year can be grown.

A third major benefit of French-bean production has been its

generation of employment opportunities. Bean production on small
farms is undertaken by family members, although a few local
people may be hired to assist in picking. Bean production on
larger farms is carried out almost entirely by hired labor. The
picking and grading of beans is performed almost exclusively by
women. Some women may reside permanently on the farms, while
others come from nearby villages and work on a seasonal basis. A
long-distance migrant flow has also been observed with women from
Western Province coming to pick beans in areas such as Athi River
and Naivasha.

French-bean production is considerably more labor intensive

than most crops grown in Kenya. Compare the figures below:

22
Table 3: Work Days Needed Per Crop Per Ha

French beans 554 days

Coffee 294
Cotton 235
Hybrid Maize 152

Source: Hormann and Thuo (1979)

Having examined some general characteristics of French beans

and the production and marketing of French beans in Kenya, wp
move now to discuss the case of Njoro Canners.

Nioro Canners

The Market

Njoro Canners produces and exports canned French beans of

the "extra-fine" quality. Its market orienxtation is exclusively
the French market. The market for canned green beans in France
is segmented into three quality levels--"extra fine," "trifine"
(or simply "fine"), and bobby bean. Annual French consumption of
canned extra-fine beans is 30-35 million cans of A 2 1/2 size
(approximately I kilo). (1) This level of demand has been stable
over several years, and the French market for canned vegetables
generally is essentially saturated. Demand for the canned
product is seasonal with reductions during the summer months when
fresh green beans are available in abundance.

French production of canned French beans has declined since

the mid-1970s as seen below:

Table 4: Production of Canned French Beans in France (tons °000)

1975 35.2 1978 41.0 1981 31.7 1984 16.2

197'2 34.1 1979 35.1 1982 36.0 1985 21.2
1977 33.9 1980 28.1 1983 22.1

Soiirces: Marketing In Europe (April 1981), (April 1986), (October

1986)

High labor costs have rendered French production of this

labor-intensive quality product uneconomical. Consumer demand is
being met by increasing levels of imports. Examine the following
figures for French imports of green beans (including French beans
and mange-tout):

23
Table 5: French Imports of Green Beans (tons; Francs '000)

Year Volume Value

1980 9.6 39.7

1983 20.1 101.4
1984 19.9 137.5
1985 25.2 193.1

5 2
Source: Marketing in Europe, Oct. 1986, p.

French imports of canned green beans carry no tariffs for

EEC and ACP countries, but carry a 20 percent custom for other
countries of origin. The product must be labeled in the French
language and conform to specifications related to weight, size,
and quality.

Since the mid-1970s, Morocco has been the leading supplier

of canned green beans to France. In 1985 it provided 56 percent
of France's imports, sending 13,998 tons. The second largest
share was taken by Belgium/Luxembourg, sending 4061 tons and
accounting for 16 percent of imports. Kenya was the third most
important supplier, sending 3714 tons for a 15 percent share. It
should be noted that supplies of extra-fine beans are coming
almost exclusively from African countries--i.e., Morocco, Kenya,
and the Cameroon.

The French vegetable-canning industry comprises 143

enterprises, but sales are concentrated in a few firms. Five
manufacturers account for 63 percent of the industry's turnover
and three national brands account for over a third of canned
vegetable sales through the grocery trade. These three brands
are Cassegrain (for Scupiquet), D'Aucy (for Compagnie G~n~rale de
Conserve [CGC]), and Bonduelle (for Bonduelle). (2) These are
also the three largest firms and brands for the trade in canned
French beans. Saupiquet and CGC each supply approxlmately 8
million cans/year while Bonduelle supplies 3-4 million cans/year.
Maiy smaller firms supply the balance. (3)

While the manufacterers formerly distributed their products

to individual supermarket chains, in recent years a half dozen
central food-distribution firms have emerged that deliver a large
range of foodstuffs to supermarket chains. The major
manufacturers now sell through these organizations. While
Saupiquet sells its products almost exclusively under its
Cassegrain brand, the other leading firms sell under both their
own brands and the labels of the retail chain. Heavy competition
has sharply reduced margins, and price premiums for prominent
brands have been reduced. At the retail level canned fine mange­

24
tout beans sell for approximately one-half the price of extra­
fine beans. (4)

Saupiquet

The company involved in the Kenya project is Saupiquet. It

is a public company with shares traded in the Paris stock
exchange. There are a few major shareholders, including
Compagnie Navigacion hixte who hold 20-25 percent. The firm
dates from the late 19th century and has always been a canning
company. The present company is a result of a long series of
mergers which, beginning in 1955, have incorporated twenty family
businesses. The group consists of a parent company, five French
subsidiaries, two European subsidiaries, and two African
subsidiaries. Unlike its two leading competitors, it has not
operated its own farms in France. Also unlike its leading
competitors, it supplies canned vegetables only to the household
market, not to the institutional sector. (5)

Sixty percent of the firm's turnover derives from fish

(mainly tuna) obtained from the Guinea Gulf and the Seychelles
with nearly a quarter of fish requirements coming from the firm's
own boats. Ten percent of the firm's turnover comes from ready­
made meals. For this it imports meat from Argentina, Austrailia,
and New Zealand. The remaining 30 percent of turnover is derived
from sales of canned vegetables. It produces in France canned
bobby and French beans, carrots, sweet corn, celery, peas, and
mixed vegetables, while importing canned red pepper from Eastern
Europe, sweet corn from the U.S., Canada, and Israel, and French
beans from Morocco and Kenya. In France it ranks #1 in fish and
#2 in vegetables and ready-made meals in terms of sales. It is
one of Europe's five largest canners. Saupiquet had a 1985
turnover of French francs 1.63 billion and employed 3437
people. (6)

Saupiquet's attraction to Kenya rests on the two aspects of

comparative advantage discussed earlier: ecology and low labor
costs. Since the early 1970s Kenya had been supplying fresh
"extra-fine" beans to the Paris Rungis market und had begun to
develop a reputation for quality. The Kenyan product was
available all year long, in contrast to local French production
which was limited to the summe-r months. Local production
patterns forced canners to process green beans during a short
period and to maintain costly stocks for the remainder of the
year.

The most important factor, however, was rising agricultural

labor costs in France rendering it uneconomical to harvest and
process "fine" and "extra-fine" French beans. Still, the French
consumer was willing to pay a premium price for supplies of the

25
high quality product. Saupiquet needed to source this product
from areas with relatively low labor costs. Moat important is
the identification of areas with low cost but productive
agricultural labor forces. The cost of harvesting the raw
material is the most important cost in the processing of French
beans. Even in Kenya where taxes, tariffs, and imperfect
competition render the costs of fuel, cans, and equipment
considerably higher than in France, the beans themselves have
comprised the largest component of total pruduction costs
coverinc an average of 37.7 percent of total costs over the 1983­
85 perio (7)

The Origins of the Kenyan Project

In the mid-1970s, witnessing increased competition in the

French market and continuously rising agricultural labor costs in
Europe, Saupiquet began to examine the possibility of sourcing
canned French beans from Africa. Initial efforts were made in
Morocco and Kenya. Both of these efforts would contribute to the
later development of the Njoro Canning project in the 1980s.

Saupiquet in Morocco(8)

Prior to Moroccan independence, Saupiquet had operated fish

canning factories in that country. When these were nationalized
with minimal sompensation, the firm adopted a policy of not
making further capital investments there. However, Morocco had
become an important supplier of fresh "off-season" French beans
to the French market and one of Saupiquet's leading competitors
was obtaining canned beans from that country. Contacts between
the Vice President of France and a top official in the Moroccan
Ministry of Agriculture led to a fact-finding mission to explore
the scope for processing beans for Saupiquet.

An agreement was reached with a Moroccan businessman who

owned a small processing factory (producing paprika for export to
the United States) whereby the local businessman would provide
the finance and Saupiquet would provide technical assistance and
management, ensure the marketing of the canned product, and
guarantee a minimum profit level. Saupiquet sent Mr. Gilbert
Bintein, a manager of one of its European factory operations, to
manage the project. The local businessman invested one million
French francs to build a new factory site (1.2 tons/hour
capacity) and provict d 300,000 French francs toward the initial
raw material production operations. Production began in 1976.

The project manager knew that they could not base raw
macerial procurement on a large-scale estate. Due to the crop's
labor-intensivity and the problem of supervising a large labor
force, he figured it unlikely that they could obtain an "extra­

26
fine" or "fine" product from large-scale production. However, if
they could not obtain such a quality level there was no point
operating in Morocco. Low cost bobby bean supplies could be
obtained in Europe. An experienced production specialist from
Saupiquet recommended that the factory obtain raw material by
procedures similar to those used in France: i.e., the company
should provide production contracts to farmers to grow plots of
2-5 ha of beans. Another adviser, a man who had just completed
work on a rice project in Madagascar, suggested that better
results could be obtained by focusing on smaller units of
production. The latter strategy was eventually adopted.

The production area chosen was a Spanish- and Arabic­

speaking area near Tangiers with sandy soils but with good
ground-water rescurces. Local farmers were growing cereals and
vegetables fcr home consu: ption. Most did not know what a French
been was, as the area w~s about 250 km from any major bean
growing area. The company began with demonstration plots and
initially convinced 50 farmers to grow the crop. By 1980 nearly
4000 farmers were participating in the project.

Participating farmers had a minimum holding of 6 ha with

some farmers having 15-40 ha French beans were generally grown on
1/4 to 1/2 ha plots although some farmers had up to 2 ha of beans
growing at any one time. Farmers grew French beans under
contract throughout the year. Initially the crop was collected
and brought to the factory for weighing and sorting. Farmers
were suspicious about this quality control and weight reporting
system, so a system was developed to purchase the beans at a
village-level collection center using a company representative
with a scale. The company wanted to reduce the risk of loss due
to theft or improprieties surrounding cash payments, so it
instituted a system of provicing farmers ticket receipts for
their deliveries for a lump-sum payment at the end of the crop.
Initially this practice was resisted, but as an "ambiance of
trust" was built up, the farmers gave their support. Groups of
farmers elected leaders to act as intermediaries between them and
the company.

Saupiquet's (i.e., Bintein's) experience in Morocco over the

1976-1980 period had an important influence over the design and
functioning of the Njoro Canners project, especially in its early
development. The knowledge gained and the lessons learned would
have both positive and negative influences on the Kenya project.
This issue will be explored below.

Saupiquet Imports from Kenya

Since the early 1970s French companies had been importing

Kenyan fresh French beans to supply the local catering and "up­

27
market" consumer trade. Saupiquet was interested in finding
someone to expand Kenyan production and to process extra-fline
beans. At the time Kaba~i Canners was the only firm actually
processing green beanq to supply the small and highly protected
Kenyan market. Kabazi was Jointly owned by a local businessman
and Brooke Bond (K). Kabazi began supplying small quantities to
Saupiquet in 1976. Kabazi was not interested in getting involved
in supporting French-bean production, but agreed to increase
processing output if provided additional raw material. One
French importer who was in contact with Saupiquet suggested that
the latter contact his fresh French-beans supplier, a firm called
Corner Shop Ltd., to see whether that firm would be interested in
organizing raw material supplies for Kabazi. Corner Shop's
manager, Mr. Wadhwa, was amenable to thia arrangement. (9)

Between 1977 and 1981, Mr. Wadhwa, using technical or

financial support from the Ministry of Agriculture and from
several foreign donor agencies, initiated a number of Frenich-bean
production schemes in Western Kenya. Together with an American
partner he leased a 1000 acre farm in Nanyuki to grow potatoes
and French beans each on 100 acres. The potato seeds that he was
given by a government agency proved to be defective and that crop
was lost. With the beans they were unable to organize sufficient
labor to do the weeding and harvesting of such a large planted
area. That effort was also written of:. (10)

In areas such as Kitale, Eldoret, and Bungoma, Wadhwa

attempted to encourage large scale farmers to grow a few acres of
French beans. Rather than deal directly with the farmers, Wadhwa
provided inputs and crop payments through local cooperative
societies that had been handling other crops. By 1979 Corner
Shop had 1500-2000 farmers growing beans under this system. The
firm was not sufficiently able to supervise input distribution,
production, and collection, given the scattered pattern of the
farms, and was dependent on the effective functioning of the
local cooperatives. Cooperative mismanagement and
entrepreneurial pursuits an the part of managers undermined the
system. Many participating farmers became disillusioned with
growing French beans for processing, given the heavy labor
demands and the low price offered them relative to what was being
offered by exporters of the fresh product. The seeds provided by
Wadhva were of the Monel variety, the same variety preferred on
the fresh market.

Wadhwa continued to search lor new areas. A staff member of

the Bungoma Horticultural Cooperative recommended that Wadhwa try
his home area, Vihiga Division of Kakamega District, because of
its suitable ecological conditions and the absence of
satisfactory cash crop options in the area. In 1979 Wadnwa
started operating in Vihiga. Corner Shop operated through the

28
Manyatibu Cooperative Union, which had previously dealt with
locally produced dry beans, honey, tomatoes, and poultry. Corner
Shop would provide inputs to the Union on credit to be deducted
against the future crop. The Union in turn was to deal with
three primary societies. These societies would issue seed,
collect the crop at collection stations, and serve as bases for
local staff appointed by Corner Shop who would do chemical
spraying of fields and supervise grading at the collection
stations. Corner Shop appointed two field supervisors to go on
motor bikes to advise farmers. (Il)

While the effort was based on good intentions and there was
initial enthusiasm about the project, the operation was neither
technically nor organizationally sound and eventually brought
finanacial loss and farmer disappointment. (12) Neither Corner
Shop nor the cooperative leaders knew what inputs and cultural
practices would be necessary to grow French beans successfully
under Vihiga conditions. Field research was not undertaken
locally. Rather, technical advice was based on field research
conducted at government research stations ±n Thika and Nakuru,
each under significantly different ecological conditions.
"Advice" provided by chemical company salesmen proved to be
misguided. Farmers were encouraged to grow continuously, even
though rainfall was insufficient over 4-6 months to get a
profitable crop.

The performance, both of the cooperatives and of the

farmers, proved to be disappointing. Cooperative staff
frequently sold chemicals and fertilizers, and some farmer
receipts went "missing." The cooperative union delayed its
payments to project workers and farmers, sometimes over three
months atter the time when Corner Shop paid the union. The
deductions taken by the cooperatives were excessive given the
level o± services provided. In 1980 Corner Shop paid sh.2.50/kg
but farmers were paid only sh.l.75, the cooperatives having taken
30 percent.

Farmer yields were very low, averaging 30-40 kgs per kilo of
seed provided. This would be the equivalent of 600-800
kilos/acre, which is one-third to one-half the norm in Kenya for
French beans. Thirty percent of the value of the input loans was
not recovered by Corner Shop. Lacking adequate advice and
supervision, farmers preferred to keep pods on the plants for
additional time to get a heavier crop. The weight difference
between an "extra-fine" and "fine" bean is approximately 40
percent. FarmerE could thus considerably increase the weight of
their crop by picking every other day rather than every day. The
company had thua to take and process fine as well as extra-fine
beans, selling the canned fine bean product on the local market.

29
Operating at a loss, Corner Shop's operations in VJhiga drew to a
virtual halt in 1981.

Project Establishment

In the fall of 1981 Gilbert Bintein came to Kenya to examine

the potential for expanding processed French-bean exports to
Fronce. Bintein's attention was focused on identifying a
suitable location for establishing a contract-farming scheme. He
looked for an area with 1) high population (and farm) density,
2) temperatures in the range of 20-25 degrees celsius,
3) relatively high and evenly spaced rainfall patterns, and
4) natural wind breaks. Visits were made to Kitale, Kisii,
Kericho, Njoro, Thika, the Coast, Eldoret, and Vihiga. He
examined existing French-bean producLion for processing or
expoi t, noting the insufficient collaboration between farmers and
buyers and inadequate use of fertilizers and chemicals. (13)

Bintein gave little consideration to the prospect of

establishing a large estate to grow French beans. Labor
recruitment and supervisifn problems ruled out this option. On
larger horticultural farm3 in Kenya nearly all harvest rs of
French beans are migrant vomen, many of whom are single. The
social problems accompanying large-scale deployment of such a
labor force have proven to be large. (14)

There was hope, however, that medium-scale farmers would

provide the factory part of their output. The prospect of
getting such farmers to grow exclusively for the factory was
rather grim as many such farmers were being sought after by
exporters of fresh French beans who offered 2 1/2 to 3 times the
price that the factory would offer. Past efforts by Wadhwa to
recruit medium-scale farmers to grow beans for processing had
proven unsuccessful.

The only group of farmers for whom growing beans for

proceasing would appear highly attractive would be smallholders
with limited cash crcp options and with sufficient family labor
to carry out the necessary husbandry-intensive techniques for
high quality French beans on a very small scale. This issue is
further discussed in the section below entitled "smallholder
participation."

Indeed, Bintein decided that the most appropriate area for

production would be Vihiga in Kakamega District, Western
Province. This area not only possessed the physical and
socioeconomic characteristics noted above, but it also lay a
considerable distance from any important French-bean market, thus
reducing the risk of "leakage" of beans onto alternative markets.
Approval to operate in the area was sought from the District

30
Permanant _mmissioner, the District Agricultural Officer, and
the local government chiefs and subchiefs.

One of the few individuals to assist Bintein during his

fact-finding trip was Wadhwa, and this led Bintein to incorporate
Wadhwa into the project being developed. Wadhwa would be
responsible for financing a Vihiga-based production control unit
called Hortiequip Ltd. and would share in the profits of the
overall Kenyan operation. Kabazi Canners showed little interest
in working further with Wadhwa or in expanding their capacity to
process French-beans. An alternative partner was identified. A
prominent Nakuru-based businessman (dealing in building
supplies), T. K. PaLel, had acquired a small canning factory in
Njoro in 1978. It was operating periodically employing 20-40
people, canning peas and beans in tomato sauce for the local
market.

In December 1981 an agreement was signed between Saupiquet

and Patel whereby Patel would finance capital investment in an
expanded factory and cover the operating costs of the factory.
Saupiquet would provide technical assistance in remodeling the
factory, manage the factory and the raw material production
operation, marke. all factory output, and guarantee Patel a
minimum return on his investment. (15)

Project Location

Kakamega District is divided into ten administrative

divisions. The Fre:ich-bean project has operated in three of
these--Vihiga, Hamksi, and Ikolomani. The District (and the
divisions where the project operates) is characterized by three
main features: 1) high agricultural potintial, 2) high population
density, and 3) high rate of labor out-m'gration.

Kakamega District lies in a zone of high agricultural

potential. Of its total 3520 sq km, about 3250 sq km are arable.
Rainfall varies between 1250 and 2000 mm with a less than 10
percent probability of obtaining less than 750 mm of rain in a
year. Rainfall is generally adequately distributed with no major
dry season. Rainfall maxima come in April/nay and August/
September. (16) The area's geography und climate are thus highly
suitable for growing vegetables. (17) A Ministry of Agriculture
report warns, )hovever, that the high rainfall pattern provides a
breeding ground for pests and diseases and that hail is a hazard
in the area. (18)

The populationi density of the District was 295 per sq km in

1979 and estimated at 349 per sq km in 1983. The divisions with
the three highest population densities are those where the
project is based. In 1979 the population densities were 692 per

31
sq km in Vihiga, 612 per sq km in Hamisi, and 402 per sq k:m in
Ikolomani. (19) Martin (1985) notes that according to colonial
officials, the "Fihig area already had a population density of
450 per sq mile in 1919. One group of resevrchers claims that
the population density of Vihiga is probably as high an any rural
location in eastern and southern Africa. (20)

From the early part of the coloniel period this region has
served as a labor reserve. Martin argues that this pattern arose
from a combination of the following factors:
i) the colonial ban on African export crop production;
2) increasing land pressure;
3) neglect of agriculture by the colonial government during
the 1930s and 1940s; and
4) an anti-capitalist ethic engir;eered by Quaker
missionaries based in the area.

Referring in 1960 to the area where Vihiga lies, Elspeth

Huxley stated that "Maragoli has become a sort of dormitory area
for places as distant as Mombassa and its communities return for
a month or two every year efter harvest to drink millet beer and
produce a new crop of babies."(21)

The high rate of out-migration has created anomolies in the

local labor market. In the 1984-88 Kakamega District Plan it was
estimated that out of a workforce of 482,484 in 1983, 276,293 or
57.2 percent were outmigrants. A large proportion of migrants
are male, leaving the majority of productive labor in the
District to be provided by wcinen, children, and older people.
While for Kenya generally the sex ratio for the population 15-49
years of age is 105 females per 100 males, for Kakamega it was
134-100 in 1979 and 126-100 in 1983.

Martin argues that "agriculture has ceased to be a

sufficient source of income and households have become more and
more dependent upon income from wage labor." While agriculture
has been poorly developed it has "been a cushion against the
vagaries of labor demands" and thus prevented the marginalization
of the population. (22) Martin presents survey results showing
that the proportion of household income in Maragoli deriving from
off-farm activities rose from 77.5 percent in 1969 to 84.5
percent in 1977.

The survey results did show considerable differences among

sub-groups, with those households with more Lhan 7 acres getting
91 percent of income from off-farm activities compared to 78
percent for those with less than 3 acres and 58 percent for those
with 3 to 7 acres. While larger landholders tend to find off­
farm employment in teaching or the civil service or else operate
their own small business, off-farm income for smaller farmers

32
tends to come from employment in the Mumias Sugar scheme, the
Webuya Paper Mills, the Nandi Hills tea estates, or work in
Nairobi or Nakuru. (23)

Smallholder Participation

The number of farmers participating in the project has

expanded significantly since its initiation, as seen in Table 6
below. The company has sought to control farmer participation,
firstly through the specific sublocations where it establishes
collection centers, and secondly through endeavoring to achieve
maximum control over the distribution of production inputs. The
locating of collection centers is critical as the cost and
availability of motor transport limits farmers to delivery points
only within close proximity to their farms. Control of inputs
begins with the distribution of seed with exact seed allocations
made to indiviJual collection centers based on the number of
farmers whom the center's extension agent (i.e., the "control
clerk") has registered. When seeds are distributed the farmer
signs a contract with the company, her name and ID number is
recorded, and a "farmer card" is irsued on which subsequent input
and crop transactions are recorded. The French-bean variety used
is Vernandon. This variety is not commonly used in Kenya and
thus there are few alternative sources of seed. Farmers need not
show a land title when obtaining a contract. (24)

Table 6: Farmer Participation

Year/Season Number of Farmers

1982 1,000-1500
1983 (first season) 3,290
(second season) 3,397
1984 (first season) 10,359
(second season) 12,686
1985 (first season) 13,526
(second season) 15,765
1986 (first season) 12,078
Kisii area 3,466

Source: Hortiequip Ltd.

Approximately 70 percent of the farmers participating in the

project are women. This is perhaps not surprising given the
incidence of male out-migration and the significant number of
farms that are managed by women. While in the early years of the
project many of the farmer contracts were signed by the husbands,
more recently women themselves have signed for the contract and
their ID number is noted on the farmer card. (25) This change is

33
significant since payment is made to the person whose ID number
is on the contract.

The widespread participation of women in a production system

involving extension and credit is significant and a departure
from past patterns in the Vihiga area. Staudt (1977) found that
there was a severe bias against women in Vihiga in the delivery
of government agricultural services. This bias held even when
controlling for economic standing, size of landholding, and
demonstrated interest in adopting agricultural innovations. She
found that 98 percent of government agricultural field staff were
men and that communications between women farmers and male
extension staff who are not related by kinehip frequently aroused
suspicion, especially when the husbands were absent. She found
that 49 percent of female-managed farms were never visited by
extension staff while 28 percent of jointly-managed farms were
not visited. Attendance by women at demonstration sessions and
training courses was also considerably lower than for men. She
found that 99 percent of women on female-managed farms knew
nothing about the procedures for a loan application even though
an Agricultural Finance Corporation program had been active in
the area for three years prior to the time of her survey. Women
felt that since they lacked a regular salary and since they
themselves did not hold the land-title deed, agricultural credit
was not open to them. Staudt summarizes that "a large part of
the bureaucracy's clientele, who are women, are in effect
ignored."(p.2)

To establish a brief profile of the Vihiga-area farmer

participating in the project we have drawn from results of
surveys carried out by Moock (1971) and Staudt (1977), and we
carried out a survey of 21 participating farmers. The farmers
interviewed in our survey were drawn from five different
sublocations that vary in 1) their length of time in the project,
2) their level of farmer yields, 3) their location, and 4) their
proximity to major roads. Farmers selected for interview also
represent a cross sample based on relative yields for the in-
progress 1986 second season. Farmers were drawn from categories
of "high," "medium," and "low" performance for the season.

Both Moock and Staudt found median landholdings per

household to be 2.5 acres. Moock found that 39 percent of
households had 2 acres or less and 44 percent had between 2 and 5
acres. The farmers in our survey had the following landholdings:

34
Table 7: .andholdings in Vihiga Survey

Area Number of Farmers

I acre or less 8
Between I and 2 acres 8
Between 2 and 4 acres 3
More than 4 acres 2

Multiple fragmented holdings have been common in this area.

In Moock's survey 38 percent of fariu consisted of more than one
piece. In our survey only 2 of 21 households had more than one
plot, but several farmers did report having sold plots in the
past five years. Most Vihiga farms have a considerable number of
people living on them. Moock found that 56 percent of farms had
7 or more people. This actually may be difficult to access as
one commonly finds holdings where parents and the families of
their sons are resident with the land being divided up amongst
the "households" but with children and family labor "migrating"
throughout the holding.

A common finding of investigators of the Vihiga scene is the

paradoxical condition that in an area with extreme population
density, there remains considerable uncleared arable land. Moock
estimated that 12 percent of Vihiga farmland was uncleared, 80
percent of which was arable. It is generally argued that labor,
not land availability is the prime determinant of cropping
acreage. (26)

Maize and local dry beans are the most important crops, with
subsistence requirements taken first and surpluses sold in local
markets. Hybrid maize has been widely adopted. Cash crops
generally consist of small plantings of coffee, tea, sunflower,
cotton, cooking bananas, and vegetables. In our survey 8 of the
21 farmers also grew vegetables (cabbages, onions, kale) for
sale, followed in incidence by coffee (7 farmers), bananas (5),
and tea (3). Five of the farmers grew no other cash crops than
French beans. These farmers had an average holding of only 1.1
acres. Those with some coffee and/or tea tended to have slightly
larger holdings than the average, with coffee growers having an
average of 2.86 acres and Lea growers 4 acres. Until the mid­
1960s, farmers with less than 7 acres of land were not permitted
to grow coffee. (27) Flucuating prices and delays in payment have
restricted smallholder interest in coffee with 1982 Kakamega
District production of the crop being less than two-thirds of its
level for 1969.

Small acreages and the considerable extent of hilliness and

rockiness limit the scope for mechanizing farm practices. All
activities from land preparation through planting, weeding, and

35
harvesting are done by hand. Small acreages, cash constraints,
and the availability of family labor (generally women, children,
and older people) limit the incidence of hiring agricultural
labor. Moock found that only 18 percent of households have paid
part-time labor and only 8 percent have paid full-time labor. In
our survey 8 of the ' farmers hire workers part-time with work
focusing on the pickL g of French beans, coffee, and tea. Most
of the women interviewed said that their husbands were working on
tea estates in Nandi Hills or Kericho or that they were resident
in Nairobi. Casual empiricism suggests that many of the men
participating in the project are either not in the general labor
force (i.e., over 60 years or less than 18) or are in the process
of making a transition between obtaining income through seasonal
work elsewhere and settling on the farm and perhaps using some
savings to establish a local business.

Past efforts to organize vegetable production under contract

for processing proved unsuccessful in Vihiga. Kabazi Canners
attempted to obtain tomatoes from Vihiga smallholders in the mid­
60s, but local market prices sometimes reached 5 times that
offered by Kabazi, and these opportunities outweighed the
consideration of a guaranteed market outlet. (28) A local factory
that extracts papain from papaya has been unable to organize
consistent supplies of raw materials and has relied primarily on
seasonal surpluses that then render the factory's price
competitive with the local fresh market for papaya. (29)

Experience prior to and after the initiation of the project

suggests that only farmers with extremely small landholdings,
with available family labor, and with limited cash-crop and wage­
labor options would find the growing of French beans for
processing economically interesting. The income earned would
barely cover the labor costs of a commercial or sma-lholder
farmer using hired labor. Only where farmers do not value family
labor at the market rate does the production prove economically
interesting. This can be seen below where we calculate the
implicit labor cost for growing 1 kilo of French-bean seed during
a season and then compare this with average net earnings from the
project. Estimations for labor input, length of work day, and
the cost of hired labor are drawn from farmer survey rexponses.
This estimation is rather crude, as considerable variations in
effort (particularly in harvesting) are observed.

36
Table 8: Estimated Labor Input and Implicit Labor Costs for
French Beans in Vihiga

Activity Quantity Cost

(10 sh per 7-hour
work day)

Land Clearing I day Ksh 10

Ridging I day 10
Cleaning/Planting 4 hours 5.71
Weeding 12 hours
(3 by 4 hours) 17.14
Fertilizer Application 4 1/2 hours
(3 by 1 1/2) 6.43
Stick Support I day 10
Bean Picking 72 hours
(36 days by 2 hours) 102.86
Transporting/Sorting 18 hours
(36 days by 30 min.) 25.71

Total Ksh 187.85

Thus, we find that the implicit labor cost for growing

French beans over a three month season is Ksh 187.85. It is
important to note, however, that a majority of farmers do not yet
perform the practice of setting up a stick support system for the
beans. When comparing net earnings with labor costs we shall
deduct the Ksh 10 for this activity. Picking is by far the most
important item in the above costing. Two hours per day was the
most commonly reported level of effort, although picking time may
vary between 1 hour and 3 hours per day depending on the
development of the crop. The time spent carrying beans to
collection stations, sorting the beans, and having them weighed
and receipted is again an average figure with actual timing
depending on distance travelled, the number of farmers at the
collection station, and even the d9gree of trust between a
particular farmer and the center's quality inspector. The
intensity of quality inspection varies from farmer to farmer.

Let us now compare this implicit cost of labor (or cost for
having hired labor work on the beans) with the average income for
farmers participating in the project. To obtain average income
we made the following calculation:

(Price x Average Yield Per Kilo of Seed) - Value of Inputs Loan

For 1985 and 1986 we use a rate of 10sh/day for the cost of
labor while for the three preceeding years we use 7.5sh/day. For

37
labor costs we have deducted the cost of constructing stick
supports. Labor cost totals are thus Ksh 133.4 for years 1982-84
and Ksh 177.9 for 1985-86.

Table 9: Average Income Versus Implicit Labor Costs

Year Average Income Implicit Labor Cost

1982 -0.23 133.4

1983(1st season) 17.96 133.4
(2nd season) 106.17 133.4
1984(1st season) 155.1 133.4
(2nd season) 235.7 133.4
1985(1st season) 137.3 177.9
(2nd season) 190.5 177.9
1986(1st season) 162.8 177.9

Of course, labor costs will var, with harvesting effort that

in turn will influence yields. Thus, implicit labor costs may be
lower than average for those getting ponr yields and higher for
those with superior yields. However, taking our crude estimation
for illustrative purposes we find that farmers obtained a cash
income exceeding the implicit cost of their labor only in three
of the eight seasons or years in which the project has operated.
This suggests the economic infeasibility of hiring labor solely
for work on the French beans for processing. When calculating
for the different seasons the yield required for a farmer to
cover riot only the value of the inputs loan but also her implicit
labor cost, we find a range of 71.3 to 88.2 kilos per kilo of
seed. On an acreage basis this would be 1426 to 1764 kilos. The
latter figures are not far below the average yields for French
beans in Kenya and generally higher if one deducts the output of
fine beans and takes only the output of extra-fine beans from a
plot of French beans. Thus, larger farmers who will generally
have higher labor costs than the 10 sh./day rate in Vihiga and
will have labor supervision costs are unlikely to find growing
French beans for processing economically interesting.

Basic OrQanizational Structure and Components

Here we discuss the basic structure of Hortiequip's contract

fe ning system. Its organizational structure considerably
matches that which was developed at Saupiquet's operation in
Morocco. In the early stages of the project many of the
company's policies also matched those adopted in Morocco.
Certain cultural practices, the terms of company-farmer
contracts, and the technical package comprising seeds,
fertilizers, and chemicals were all transferred largely intact.
Even today the overall organizational structure remains virtually

38
the same. However the operation of the system has undergone
considerable chanige since the project was initiated with the
company adjusting its package of incentives and its control
mechanisms for fh:m'nrs and for otaff. These adjustments were
necessary as the comLeany found that it was not adequately in
control of its organization and not generating the expected
farmer-productivity results. The company also found that it was
unable to enforce the terms of its contracts and unable to
prevent costly "leakages" out of the system. While the
transaction costs inherent in an organization incorporating large
numbers of smallholder farmers are necessarily high, unexpected
transactior costs arose that necessitated a company response. In
this section we outline the basic components of Hortiequip's
system. In the subsequent section, where we discuss the
performance of the project, we will identify institutional
changes made by the company.

a) Function---
The prime function of Hortiequip is to meet the raw material
requirements of the Njoro factory both in terms of quantity and
quality and to minimize the costs of raw material procurement.
Hortiequip is not expected to earn a profit on its own
operations. The strategy adopted by Hortiequip is to disperse
supply risks and spread project benefits by incorporating large
numbers of smallholder farmers.

b) Form of Transactions--­
Hortiequip's prime mode of transaction is contractual
relations based on a season or year. The company enters into a
contract with each farmer, staff member, and transporter
individually. Formal contracts are supposed not only to assign
rights and responsibilities, but to engender a perception of
continuity and common interest and effort. Rather than seen as
an alternative to trust, contracts are viewed by the company as
the frameworks in which to develop relationships based on trust.

c) Method of Organization---
The basic structure of the Hortiequip operation is that of a
pyramid with information, inputs, and harvested product flowing
through a hierarchical system, with quality-control points being
located at several levels in the hierarchy. The structure of the
pyramid is as follows:

39
 General Manager (1)

Field Manager (1)

S u p e r v i s o r s (4)

C a n t r 0 1 C 1 e r k s (60-80)

F a r m e r s (12,000-16,000)

Farmers receive general information about the project in

barazas called by their local Chief. All subsequent information
will be provided by a control clerk who acts as technical
adviser, inputs supplier, and general on-farm production overseer
for the farmer and company. Each control clerk recruits and is
responsible for approximately 200 farmers (plus or minus 50).
Each control clerk operates out of a specific collection center
to which all their farme-s come for inputs and French-bean
deliveries. At each collection center there are individual staff
members responsible for a) sorting inspection, b) weighing beans,
and c) issuing ticketed receipts to farmers.

Control clerks are to instruct farmers how to prepare their

land for the planting of French beans. Company specifications
are particular, i.e.:

170 sq meters of land well dug anI properly cleaned with a

fence of 5 rows of maize around it. The plot should have
ridges 20 cms. high, 30 cms. wide ind 80 cms. apart. There
should be no rocks, trees or any other crop or plants in the
plot.

The control clerk is to inspect the farm before issuing seed

and having the farmer sign the contract. Farmers are told when
to plant. Control clerks are issued a top dressing fertilizer
(C.A.N.), and they must instruct farmers in its application.
Urea is supplied to the control clerks in three installments and
this must be distributed to farmers and its use explained. Four
chemical sprayings sre undertaken during each crop by hired
workers under the supervision of the control clerks. When the
beans are ready for harvesting the control clerk is responsible
for ensuring that harvesting is done every day and that pods of
the proper size and quality are picked. Thus, the control clerks
play a vitnl role in the Hortiequip system, not only filtering

40
inputs and information down to the farmers, but also feeding
information upward in the hierarchy. The proper execution of the
control clerk'Lj job is thus vital to individual farmer
productivity and the overall performance of the project. (30)

Initially Hortiequip trained 60 local people to be control

clerks. Many of these were people recommended by chiefs and
subchiefs. Most had some secondary school education and had no
past record of crime. Most were 18-20 years of age. Trainees
were taught the basic stages in the production of French beans,
warning signals for plant disease, and the standard operating
procedures of Hortiequip. In subsequent years new control clerks
generally have worked in some other capacity for the company
(i.e., as chemical sprayers) for perhaps two seasons and have
been recommended by e local authority figure. These are the
methods of "screening" potential staff for responsible positions.
An important unresolved issue within the project concerns who is
actually responsible for the behavior of control clerks. Is
Hortiequip responsible as the clerks are its employees, or are
the local political officials who recommended them responsible?
Where a control clerk has committed a crime (O.e., sold spraying
equipment belonging to the company) can the company fire the
individual and take them to the police or is the political
official's consent required? A difficulty arises in that when a
local staff person commits some crime or fraudulent act and a
local political authority is considering taking disciplinary
action, typically strong local and family pressures are applied
to tha official not to take action. This typ(. of case reduces
the overall deterrent value of company policies to minimize staff
abuses of the Horti'quip system.

The ratio of farmers to contrcl clerks has increased during

the course of the project, but appears now to be near the level
of 200 farmers per control clerk, which the company considers
optimal. Changes in this ratio can be seen below:

Table 10: Farmers per Control Clerk

1982 56-83
1983 (first season) 110
(second season) 92
1984 (first season) 148
(second sea& on) 249
1985 (first season) 218
(second season) 188
1986 (first season) 183 (193 at Xisii)

Source: Calculated from Hortiequip Records

4±
 Supervisors are responsible for an area that will
incorporate 20-30 collection centers. Based on the number of
farmers each control clerk has, the supervisor will request the
necessary quantity of seed and other inputs, and this is
delivered to the collection centers. Supervisors visit each of
their control clerks each day and issue daily reports to the
field manager indicating problems, actions taken on prior
problems, and various indicators of farmer and staff performance.
The field manager assesses general patterns and p.-oblems in
production and may target additional supervision or other
remedies to areas experiencing problems. The field manager
together with the general manager carefully monitor the quality
of the delivered beans and act on quality-related problems as
identified at collection stations, at the Hortiequip main center,
or at the factory. The general manager oversees the activities
of the Hortiequip farmer-accounts unit, the inputs-supply unit,
the local transport arrangements for beans collection, and the
dispatch of beans from the Hortiequip to the factory. The
general manager is in steady contact with the overall project
manager, Mr. Bintein.

Absence of Intermediaries

No intermediary organizations are involved between the

farmer and the company. Neither cooperatives nor traders come
between the farmers and the company for input supply or product
marketing. The company has sought to minimize the extent of
government involvement in the project, fearing that such
involvement would reduce the flexibility of decision-making and
the performance-based orientation of the company. The company
has required the support and sometimes the assistance of the
district agricultural officer and the local chiefs. Assistance
frnm chiefs has been needed in disciplining negligent farmers,
fraud-committing staff, and various opportunists trying to
undermine th project. While initially official extension
officers were used to assist, inaccuracies in advice and requests
for remuneration led the company to decide to utilize strictly
its own hired staff.

The a sence of any intermediary organization between company

and farmers has several implications. Farmers have no institu­
tionalized channel to render their grievances other than through
their control clerk. Within the confines of the project, farmers
have no capacity to influence company decision-making; individual
farmers have no bargaining power. The information that they pass
on to control clerks is likely to have a high disipation rate
before reaching senior staff members. This is especially the
case if the information relates to the behavior of the control
clerk. (31) Control clerks are not supposed to represent the
farmers in the sense of presenting farmer positions and bargain­

42
ing with the company over the issues. Control clerks acting in
such a way are in danger of being perceived by the company as
being "trouble-makers."

Lack of institutionalized representation has led farmers to

make greater use of political channels to voice their complaintd.
At barazas called by local government authorities and KANU party
officials, farmers will discuss problems they have related to the
project. In this manner one event or one problem that a few
farmers have faced may become blown up into a larger issue
between the politicians and the company.

The absence of a farmer representative body is also likely

to reduce the company's capacity to enforce its contracts with
farmers and staff. For the company to sanction negligent farmers
or negligent or fraudalent staff it generally must go through
political and then police channels. There is no institutional
mechanism to bring social pressure on the offending party from
within the project. While the company has been able to instill
in participants some feeling of joint effort and cooperation,
this attitude has not been nurtured in the direction of mutual
self-government of the project.

Planting Seasons and Input Loans

Over the past four years the production of French beans has
taken place over two distinct seasons per year. With the short
rains in March comes the first planting for harvesting from May
to early July. The second planting is to accompany the long
rains in September for harvesting in October and November. Both
the cost of inputs and the producer price are set at the begin­
ning of the year and carry through for both seasons. An input
package accompanies each one kilo of seed and is costed on such a
basis. While the company does maintain stocks of certain inputs
(largely due to uncertainty of their timely availability), the
company still must bear the risk of changes in the procurement
cost of fertilizers and chemicals throughout the year. Table 11
breaks down the inputs loan for 1985.

General Performance Indicators

In this section we present data depicting various dimensions

of project performance. The drta relate to such results as
company sales and earnings, emnployment, farmer yields and income,
producer prices, and loan recovery. The prime causes of varia­
tions in performance by year or season are discussed in the
subsequent section where we examine changes in the project
chronologically.

43
Table 11: Input Loan (1985)

Input Quantity Cost

Seed 1 kg Ksh 51.00

N.P.K. 5 kgs 27.65
D.A.P. 1.2 kgs 8.06
Furadan 330 gms 11.40
C.A.N. 1.2 kgs 4.99
UREA 2.4 kgs 13.80
Chemicals 4 sprays 36.15

Total for 1 kilo seed 153.05

Rounded off to 153.0

Sales and Earnings

One indicator of performance is the growth in company sales.

For Njoro Canners all sales are exports to Saupiquet. In the
table below we give both the Kenyan Shilling value and the US$
equivalent of export sales. The dollar value is given so that
the effects of the Kenyan Shilling devaluation since 1982 are not
'Idden.

Table 12: Company Sales

Year Sales US$ Equiv.

(Ksh mills.) (Mills.)
1982 6.1 0.56
1983 14.0 1.05
1984 27.0 1.87
1985 40.3 2.45

Source: Njoro Canners

From this table one can see the steady expansion in sales
recorded by the project, which provided added foreign-exchange
earnings for the country. On the other hand, on account of
capital investments of nearly Ksh 31.8 million over the 1982-85
period and subsequent deductions for depreciation, t)B company
has registered operating losses in each year. Thus corporate tax
was not paid over the 1982-85 period. However, these "account­
ing" losses do not threaten the financial viability of Njoro
Canners. The company's owner is guaranteed by Saupiquet an
income equivalent to a certain percentage of f.o.b. sales
volume. This sum more than adequately covers the company's
"accounting" losses.

Employment

Another indicator of company performance concerns employ­

ment. The data available do not provide a breakdown between

44
full- and part-time staff. Most of the field staff work between
6 and 8 months/year. The date do indicate considerable growth in
employment. The location of employment in Njoro and Vihiga is of
major importance given the relative absence of salaried employ­
ment in both of these areas.

Table 13: Company Employment

Year Factory Staff Field Staff

1982 100 50
1983 250 100
1984 800 300
1985 850 350

Source: Njoro Canners

Farmer Productivity

A third set of indicators of project performance concerns

trends in farmer productivity and the level of productivity cf
participating farmers relative to French-bean growers elsewhere
in Kenya. Data for average farmer yields are presented in the
table below:

Table 14: Project Farmer Yields

Year/Season Yield Yield

(per kilo of seed) (on acre basis)

1982 28.17 Kgs 563 Kgs

1983 (1st season) 42.64 853
(2nd season) 69.37 1387
1984 (1st season) 83.43 1669
(2nd season) 106.46 2129
1985 (lot season) 77.4 1548
(2nd season) 91.6 1832
1986 (1st season) 79.44 1589
Kis:.i 61.0* 1220

Source: Calculated from Hortiequip records.

*Kisii yields are per farmer, not per kilo of eeed.

We should note here that the output of both small-scale

(less than I acre) and medium-scale (2 to 10 acres) grouers of
French beans for fresh export has been largely within the range
of 1620 to 2160 kgs per acre in recent years. This, however, is

45
the yield of fine and extra-fine beans combined. A harvest of
beans from one acre may consist of 60 percent extra-fine beans
and 40 percent fine beans. If we ignore the considerable weight
difference between fine and extra fine (i.e., I fine bean = 1.67
extra-fine beans) and simply take 60 percent of this yield range
for extra-fine French beans by the leading exporter, we find an
increase over this period of nearly 78 percent, with actual
prices as follows:

Table 16: Producer Prices for Bean Exports

1982 Ksh 6.7/kg

1983 8.1
1984 8.9
1985 10.4
1986 11.9

Source: KHE Ltd. larmer vouchers

It is important to point out that Vihiga farmers are well

beyond the range of fresh French-bean procurement systems, which
are generally within a 150 km radius of Nairobi's international
airport.

While we have already presented data showing the average

incom warned by participating farmers, we have yet to provide an
indication of the total cash earnings of Vihiga farmers from the
project. This is shown in the table below:

Table 17: Cash Income to Farmers in Vihiga

Year Amount(Ksh)

1982 400,000
1983 800,000
1984 4,700,000
1985 5,750,000

Source: NJoro Canners

The table shows that it was really not until three years
into the project that a substantial amount of additional income
was injected into the Vihiga economy. As we showed earlier, this
is due to the low yields obtained in 1982 and 1983.

Inputs Loan Recovery

During both 1982 and 1983 a high proportion of farmers had

output levels inadequate even to cover the input loan value.
While we do not have the exact data, it is very likely that more

46
than 50 percent of participating farmers had an outstanding
inputs balance during th fIircL two years of the project. What
data we do have for these years looks at the total outstanding
inputs balance as a proportion of the value of inputs loaned for
different seasons. This can be seen in Table 18.

Table 18: Outstanding 1puts Balance Data

Year Outstanding Number Percent

Balance of Farmers of Farmers
Total Inputs with out- with out­
(percent) standing standing
balance balance

1982 25.4
1983 (lst) 32.9
(2nd) 18.1
1984 (1st) 11.7 2173 20.9
(2nd) 3.5 1041 8.2
1985 (1st) 6.0 2127 15.7
(2nd) 9.5 2787 17.7
1986 (Ist) 10.6 2263 18.7
Kisii 21.4

Source: Calculated from Hortiequip Data

The table shows that during 1982 and 1983 approximately one­
fourth of the value of inputs loaned was not recovered by the
company. Only for those farmers shown to have misused their
inputs (i.e., sold them) would the company have attempted to
enforce loan repayment. The actual number of these cases was
small. Results for 1984-86 show that while there was a consider­
able decline in the proportion of total loan value left outstand­
ing, performance has been somewhat unsteady.

More interesting is the sustained (or even rising) propor­

tion of farmers who do not produce enough to earn any cash
income. This is seen in the last column. This represents a
measure of risk for participating farmers. While farmers new to
the project have a higher rate of failure in meeting the break­
even production point, other factors are also important. While
variations in yiell generally will arise from such factors as
ecology, labor availability, farmer attention to the crop, and
the effectiveness of control clerks, the experience of a crop
failure or harvest of a very low yield are usually a result of
climatic factors. During several planting seasons hailstorms
have badly affected some production areas with the impact
depending on the stage in the crop cycle. Hail that hits before
actual picking begins may wipe out the entire crop. Lack or
abundance of rainfall has also played an important contributing

47
role in some crop failures. The company staggers planting times
to expand the length of the processing season. This necessarily
puts some farmers at greater risk as, rather than planting
exactly with the onset of rain, their planting time may be
scheduled too early or late to take advantage of the rains.

One farm visited clearly illustrated this weather-linked

risk. The family has several members with their own plots of
French beans, although for a variety of reasons (e.g., illness,
absence of family member, etc.) their timing of planting dif­
fered. Those who planted when first provided seed were obtaining
good results with yields well above 80 kgs, but two family
members who delayed planting for 7-14 days had virtually no
yield. The two unfortunate members planted their seeds in soil
dry from an absence of rain for over a week, and the plants were
more affected by a hail storm that hit the area just before
picking was to begin.

Impact of the Project

Certainly the most important impacts of the project are its

injection of additional cash income into the Vihiga economy and
its creation of several hundred full-time jobs both in Vihiga and
at the factory. The project has also had secondary impacts in a
number of areas. It has generated some technical overspill from
the cultivation of French beans to the cultivation of maize,
local beans, and vegetables. Participating farmers have in­
creased their awareness of the positive impact of fertilizer and
chemical use for crop yields, particularly for maize. They
learned this through direct application of urea (the company's)
on maize as well as through their rotation of the French beans
with maize. More farmers are now applying manure or compost to
their food crops.

Success in growing French beans in rows with ridges and with

proper spacing has led many farmers to experiment growing the
local dry beans as well as several vegetable crops with such
methods. Results have generally been positive. An interesting
side effect noted by several farmers is that while they may have
had only limited contact with the official extension service in
the past, their participation in the project has taught them ahow
to ask for advice" from extension workers.

The project has had some social impacts as well. By

providing women with their own source of income, the project has
increased the influence of many women over the handling and
allocation of family financial resources. Increased sums have
gone toward children's clothes and school fees. Several success­
ful primary school building drives have been based on earnings
from French-bean production. Some people argue that household

48
conflicts over the use of income have been reduced because of the
women's direct access to cash. Another impact of the project is
that it has kept a number of people in the area who might
otherwise had gone off to find temporary work elsewhere. Several
chiefs report that the project has contributed to greater peace
in their areas as people are kept busier and have less time to
get into trouble.

The project's impacts have been broader than changes within

its own confines. Njoro Canners has obtained permission from a
European seed breeder for a local firm to multiply Vernandon bean
seed in Kenya. This local firm has contracts with several dozen
small- and medium- scale farmers in the Lotokitok area tn
multiply French-bean seed. In 1985 that firm had contracted for
nearly 500 acres of French-bean seed. Al+hough the production
process for seed is not as labor intensive as that for fresh
beans, this scheme certainly generated at least temporary
employment for several thousand local people.

Income and employment spin-offs from the Njoro Canners

operation also derive from the factory's purchase of French beans
from both exporters and Lake Naivasha medium-scale growers. When
the European market for fresh beans is oversupplied or when air
cargo space limitations create an excess supply condition, both
exporters and larger farmers can sell beans to the factory at
prices that can off-set the labor and overhead costs for these
farmers and part of the procurement costs of exporters. This
reduces the heavy risk of producing or exporting during the
European summer as the farmers or firms will generally have a
buyer of last resort. The maintenance cif some level of "off­
season" production has generated additi,3nal employment during
this period.

Evolution of Performance and Institutional Arrangem nts

In this section we retrace the development of the contract

farming scheme through a series of formative stages. This
enables us to provide explanations for some of the variations in
project performance over time and to discuss how the project's
institutional arrangements have evolved.

Establishment

Hortiequip's contract farming scheme began in 1982. Results

in that year would be nothing less than disastrous. Hortiequip
faced unexpected weather and crop disease problems, lacked
effective supervision over a staff and a group of farmers
familiar with neither French beans nor contract farming, and
struggled to implement a technical and organizational package
borrowed from Saupiquet's Moroccan project but not fully appro­

49
priate in Vihiga. Borrowed from Morocco was a particular
fertilizer and chemical "package" to be provided with each kilo
of seed distributed. Also borrowed was the policy that farmers
would be loaned as many kilos of seed as they thought they could
manage. Plantings would take place at approximately fortnightly
intervals in order to obtain a crop continuously over the year.
For the first planting some farmers took as many as 15 kilos of
seed, enough for about three-quarters of an acre of production.

Early plantings, involving several hundred farmers, were hit

by a leaf rust disease that spread rapidly in some of the growing
areas. Hortiequip was late in gauging the extent of the rust
disease outbreak. The official agricultural establishment could
not provide advice on how to control the spread of the disease.
Dutch agronomists working on a legume research project at Thika
helped diagnose the problem but advised Hortiequip to have
farmers uproot the entire first two French-bean plantings. The
company feared that this would cause farmers to lose interest in
the project as it would leave them with no income at all from
their cultivating efforts. The crop was left in and a minimal
yield was recorded. (32)

Throughout much of 1982 Hortiequip was focusing on organiz­

ing its physical facilities, its system of record-keeping and
contracts, and its arrangements with local and other transporters
to collect and then delilier beans to the factory. Production
supervision and information feed-back were not yet sufficiently
developed to enable the company to know the causes and extent of
the disease problem. The activities of control clerks and
chemical sprayers were not closely monitored. Area supervisors
were acting on their own initiative and were not yet following
any standard operating procedures for problem evaluation and
reporting. Staff were being paid standard salarieh without any
built-in incentive system based on measurable performance.

The outbreak of disease and the occurrence of certain pests

suggested to the company that either the chemical spraying staff
were not performing their job or that the chemicals (or their
particular strengths) were not appropriate for growing conditions
in Vihiga. Questions also began to be raised about the appropri­
ateness of the fertilizer regime that was based on the Moroccan
experience. It was becoming clear that the company would need to
initiate its own local-level research program in order to
establish the soundness of its inputs package and to distinguish
a technical problem from a problem of human negligence.

Not only was there an outbreak in disease, hut it was slowly

becoming apparent that farmers did not understand the heavy labor
demands of growing French beans and that Saupiquet's experience
with farmers in Morocco led it to misjudge the appropriate scale

50
of production in Vihiga. Hortiequip was providing farmers with
quantities of seed far in excess of what they could possibly
manage. Some farmers began selling excess seed to others. As
Hortiequip identified this problem it began to limit the quantity
of seed to be loaned to each farmer for each planting. The first
limit set was 6 kilos. This was later reduced to 3 kilos.

Farmers were provided with an input package of seeds,

fertilizers, and chemical sprayi.ng. At the then prevailing
inputs cost and producer price the farmers needed to produce
28.24 kgs of beans per kilo of seed simply in order to cover
their loan. They would receive cash for yields over and above
this level. What transpired was that many farmers did not
deliver enough to cover the first input loan. Still they
expected some payment, either as an advance for the second
planting or to cover their labor input. Many farmers did not
really understand the nature of the contract. The contract was
explained to farmers at baratas and then by the control clerk in
their area, bjut uncertainty remained. The contract was written
only in English and some farmers flatly refused to sign it. They
feared that the paper they were signing would lead to the loss of
their land. This had happened to several local farmers who had
obtained loans from the Agricultural Finance Corporation but were
unable to repay.

Farmers who had taken more seed than they could manage
themselves had hired laborers to harvest the crop. These farmers
thus had a cash deficit from their early bean crop. In order to
prevent farmer disillusionment the company adopted a policy to
have half the value of the farmer's delivered crop go toward
recovery of the loan while the other half would be paid to the
farmer in cash. Many farmers had their crop badly affected by
the rust disease and then later in 1982 by a fungus arising from
rapid bacteria growth during heavy rains. The level of -'ejected
beans at the collection centers was thus high. In order to
provide some incomes to farmers Hortiequip sometimes accepted
non-processable beans and hen provided these free to Kisumu area
institutions (i.e., schools and hospitals). Actual enforcement
of the contract's quality-related provisions was impossible for
the company if it wanted to remain in operation. Debt collection
would have been difficult and would certainly have led to farmer
withdrawal.

For the year of 1982 (which included at least nine plant­

ings) overall performance was poor. The average yield per kilo
of seed supplied was only 28.17 kgs of beans, slightly below the
figure needed merely to recover the input loan. Had the company
not adopted the policy of paying the farmer for half of her
deliveries, the average net income per kilo of seed would have
been a credit note of Ksh 0.23. During the year Hortiequip

51
provided inputs costed at Ksh 1,226,700 and at the end of the
year the outstanding inputs balance was Ksh 311,195, amounting to
25.4 percent. During the year 12.3 tons of seed had been
distributed with the company estimating that virtually no yield
was obtained from 8 tons from this total. (33)

Reconstruction

1983 was a year of adjustments for Hortiequip. Several

major policies were altered. Incentives and controls for staff
were changed. Farmers with low productivity were either dropped
by Hortiequip or exited4 on their own accord. One important
decision that was ade was that the project would operate only
during two distinct seasons accompanying the short and long
rains. Rainfall between these two seasons was not reliable
enough to expect income-generating yields for farmers, while
attempts at encouraging small-scale irrigation activities were
still in their infancy. To provide some dispersion of raw
material supplies to the factory, each season would consist of
two plantings staggered according to sub-area.

A second policy change related to an attempt to gain

increased control over the distribution and application of
inputs. Farmers would be restricted to a maximum of two kilos of
seed per season, and most farmers would be given only one kilo of
seed. During the first season of 1983 the average quantLty of
seed taken by farmers was 1.51 kilos. For the second season this
dropped to 1.09 kilos. Control clerks would be provided only the
quantity of seed needed for the farmers, which they had regis­
tered before the start of the season. Rather than provide
farmers the total allocation of urea at one time, it was decided
to subdivide the provision into three smaller lots so as to
increase the proportion of urea actually going to the French
beans rather than to the farmers' maize or vegetables crops.
Staged urea distribution would also prevent the practice of
applying urea all at one time rather than spaced over various
points in the bean growth cycle.

Uncertainty over the actual performance of chemical spraying

led the firm to adopt a practice whereby both the control clerk
and the farmer had to sign the farmer's card at the time of each
of the four chemical sprayings. An incilent arose where the
company was accused of using dangerous chemicals after a sprayer
had apparently sold some insecticide that was subsequently
sprayed on cows. The cows died.

Getting the technical package right was also a focus in

1983. Trials with different seed varieties and different
chemical and fertilizer applications were developed on farmer and
demonstration plots. Assistance was sought from the Dutch

52
advisors at the National Horticultural Research station as well
as technical advisors from Saupiquet. it was not until the end
of the year that the company had become confident in its inputs
package. (34)

Efforts were also made to improve quality monitoring

throughout the system and to more firmly base staff salaries on
performance. Delivered beans were to be examined for quality
throughout the chain to the factory on the basis of collection
center code number. In this way quality problems could be
countered by location-specific remedies. Remuneration of control
clerks was changed from a basic guaranteed salary to a system
with a basic salary together with flexible (and rather signifi­
cant) bonuse& and deductions according to individual behavior and
farmer performance.

For the first season of 1983 15 new sub-areas were added,

and three low-performance areas from 1982 were dropped. The
number of participating farmers more than doubled over the 1982
levels. Farmer performance during the season was generally poor.
In fact 18 of the 30 centers had average yields below the 37.2
kiloo needed merely to cover the loan. The overall seasonal
average yield was 42.64 kilos, bringing an average net income of
a paltry Ksh 17.96. At the end of the season the outstanding
input loan balance was 32.9 percent of the total loan value.
Unexpectedly, new entrants into the project performed better than
those who had participated during 1982. Each of the four highest
yielding staticns were new for 1983.

There is some evidence that the staggered planting system

adversely affected certain areas. Collection centers were
divided into two regions with each region planting at slightly
different times. Region "A" recorded an average yield of 61 kgs
while region "B" recorded an average yield of only 25.4 kgs.
Since new and old collection stations were included in both
regions and since there is no clear geographical or ecological
divide between the two regions, one can only conclude that
rainfall patterns were such that the scheduled planting time for
region "B" was either too early or too late.

Between planting seasons of 1983 a considerable "shaking

out" occurred in the participants in the project. Six collection
centers were dropped and thirteen new centers added. Several of
the dropped centers had actually performed rather well in terms
of factor yields. Problems of an "attitudinal" nature were
encountered either in the form of control clerk drunkenness or
fraud, or disagreements between the company and local authori­
ties. Examining the 18 collection centers that had average
yields below the figure necessary to cover the loan, one finds a
drastic reduction in farmer participation during the second

53
seas-n. Three of these centers were dropped completely while in
some centers there were as little as 1/10 the participants in the
second as in the first season. Farmer participation in these 18
sub-areas dropped from 1514 farmers to 510 farmers during 1983.
From this information, one might estimate that 1/3 of the
project's participants exited during 1983. The vast majority of
these farmers were those who received no income during the firsu
season and may have held an outstanding loan balance.

Performance during the second season of 1983 improved

considerably and provided the first sign that the organizational
structure of Hortiequip could generate results with smallholder
farmers. Average yields per kilo of seed were 69.37 kilos, and
foir of 37 collection centers had total averages exceeding 100
kgs per kilo of seed. The magnitude of outstanding loans showed
a major decline and represented 18.1 percent of the total value
of loans.

Expansion and "Migration"

Having built confidence both in the functioning of its

organization and the technical package it was offering farmers,
Hortiequip moved over the 1984-1986 period to expand the size of
the project considerably, to diversify its operating areas and to
raise overall productivity. It obtained at least qualified
success in each of these objectives. Thcough adaitional invest­
ment, the processing capacity of Njoro Canners was expanded.
Greater effort was thuis put into expanding the period of raw­
material supply and maximizing the actual quantity of raw
material that would be processed and canned.

During the first season of 1984, 33 new collection centers

were added and the level of farmer participation was tripled to
over 10,000. By the end of 1985 a further 50 percent rise in the
number of farmers had taken place to reach a level of nearly
16,000. In addition to new collection centers in Vihiga and
Hamisi Divisions, operations began in Ikolomani Division of
Kakamega District. An elfort was made in late 1985 to expand the
project to the Bahati area in Rift Vqlley Province, but this
proved unsuccessful and was subsequently dropped. In 1986 the
project initiated an operation in the Kisii area, contracting
over 3000 farmers there; however, a consolidation of the Vihiga
operations of Hortiequip reduced farmer participation numbers
there and left total participating farmers at slightly below the
1985 maximum.

The first season of 1984 featured a tremendous productivity

improvement over the prior season. Average farmer yields were
83.43 kilos and 21 of 70 collection centers registered average
yields in excess of 100 kgs per kilo of seed. The outstanding

54
input balance fell to 11.69 percent of the total loan value. The
vast majority of participating farmers earned a reasonable
income. Officials from several locations requested that the
company establish a collection center in their area.

Despite the improved performance, the company was becoming

worried by a pattern of deviations between the weight of beans as
recorded and receipted at the collection centers and the weight
of the beans as recorded at the Hortiequip central office. As
the company is responsible for paying farmers iccording to the
receipted weights, this leakage would be a cost horne directly by
the company. The scale of the problem would take on greater
magnitude during the second season.

The second season of 1984 brought the appearance of the

highest level oi productivity yet recorded for the project.
Between seasons the company had dropped centers that were
performing poorly for either ecological or "attitudinal" reasons.
The number of farmers linked to control clerks with superior
performance was increased. During the second season average
yields at several centers exceeded 150 kgs. hearly hall the
control clerks had groups of farmers without a single shilling of
outstanding inputs balance. Over Ksh 3 million was paid out to
farmers during this season.

While farmer _roductivizy had undoubtedly improved signifi­

cantly during the season, this result was perhaps overshadowed by
the tremendous discrepancy found between farmer-receipted yield
and actual deliveries of beans. The receipts farmers were
obtaining from collection center staff were showing a higher
number of kilos than the farmer was actually bringing to the
center. Sometimes the total discrepancy between the weight as
recorded at the centers and as checked at the Hortiequip base
office would be 5 percent while at other times it might be as
high as 10 percent. For 1984 as a whole fore than 120 tons of
produce was overrecorded by collection c nter staff. This
equalled 5.4 percent of total deliveries and cost the company Ksh
420,000 or over I percent of its total operating costs for that
year. (36) This large payment for beans never delivered led
Hortiequip again to operate at a loss despite L ibstantial farmer
productivity gains.

Naturally this issue is highly sensitive, and participants

are not prepared to discuss it, but it is necessary to speculate
on the factors that led many farmers and staff to collude in an
effort to extract additional income from the company. One fairly
weak hypothesis is linked to the 1984 drought which affected
several major agricultural areas in Kenya. The suggestion is
that numerous farmers in the project had family members who
experienced a decline in their migrant wage earnings, and this

55
created an increased demand for cash-crop income within Vihiga.
"Beating" the Hortiequip system appeared to be the easiest
method.

A more plausible hypothesis relates the fraud to the

changing of local attitudes toward the project. The considerable
expansion and improved performance of the project in 1984 was
providing participants and other local people with the perception
that the company was earning substantial profits. Several local
individuals including people in "high places" were voicing the
opinion that the company was "exploiting" participating farmers,
paying them an inadequate price for their beans. As some of
Hortiequip's senior management staff were Asians, Hortiequip was
increasingly being described as a typical "middleman" operation
profiting "on the backs" of farmers. Most people did not
understand that Hortiequip staff are merely employees of Njoro
Canners.

An attitude of suspicion was adopted by an increasing

number of farmers. Farmers complained that Hortiequip was taking
their rejected beans and then selling them at high prices in
Kisumu. As a result, the company had to stop its practice of
distributing beans free to local institutions. Some farmers and
staff must have decided that they could effectively redistribute
company earnings through their own initiative. This form of
income redistribution may not have appeared too devious as, after
all, the company was being approached by many officials to donate
sums of money to social and political causes (or provide jobs to
certain people), and why shouldn't those actually gLnerating the
wealth be better remunerated. Rather than acting on behalf of
Hctiequip, some staff formed a quasi-alliance with farmers in
order to extract additional income.

The weight overrecording was the most graphic although

certainly not the sole method by which farmers sought to beat
Hortiequip's system. Farmer attempts to add rocks or weeds to
their bags of beans to increase weight were certainly not rare.
A less devious and more common pi-actice has been for farmers to
retain a certain proportion of bean pods on the plants in order
to produce their own seed for use in subsequent seasons. Whether
planting additional seed actuall) brings the farmer higher yields
is uncertain. The company's chemical sprays and fertilizers are
calculated on the basis of one kilo of seed, so these inputs will
be required to do "extra work" on a field larger than 170 sq
eters. The risk of pest or disease attack probably increases.
Farmer seed multiplication may be one of the most important
factors contributing -to the greater incidence of seed-borne
diseases in the project over the past several seasons.

56
 Despite the losses incurred by Hortiquip and recorded in
the annual financial statement of Njoro Canners, the expanded
volume of production during 1984 increased the turnover and
profits of the overall production and marketing operation
(including distribution in France).

Several changes in the scope of the project occurred in

1985. Capital investment of over Ksh 14 million was made in
expanding the capacity of the factory and in putting in a canning
line for celery hearts. This celery line would later be dropped
due to its unprofitability. Late in the year an attempt was made
to encourage medium-scule growers at Bahati to grow French beans
so as to obtain a crop for the factory for several weeks after
the end of the Vihiga second season. Over 2000 farmers were
contracted, some with up to 4 acres under French beans. While
the ecological conditions proved appropriate, inadequate labor
was available for picking. As a result farmer yields and income
from French beans could not compete with alternative crops.

During 1985 Njoro added a product line for fine beans

because at the beginning and end of each planting season the
company was getting a significant proportion of beans that were
not extra fine. Over the course of an entire season perhaps 15
to 20 percent of beans delivered to the factory from Hortiequip
are not extra fine. This raw material has to be utilized to
lower wastage costs, hence the development of the fine bean line.

In 1985 the company announced a policy that if there were

further discrepancies between receipted produce weights and
actual deliveries, then the company would deduct an equivalent
amount the following day from the offending collection center.
The company repeatedly warned that the practice of overrecording
deliveries could undermine the existence of the project. While
this policy wes never actually implemented, it did serve its
deterrent role. Weight differentie's totaled only approximately
10 tons in 1985. The cost of this level of discrepancy was less
than the level of company donations to local political functions
that year.

During 1986 a number of initiatives were made. In an effort

to improve management supervision and lower transport costs, the
Vihiga operation was consolidated by dropping 18 control­
clerk/farmer groups and by increasing the number of farmers
reporting to each collection center. Various experiments were
carried out in an effort to increase productivity and lower
costs. Experiments were conducted with a climbing variety of
French beans whose yield (but also production cost) per area was
expected to be considerably higher. Experiments were also
conducted using compost (made up of rejected French beans and

57
sawdust) instead of urea in an attempt to save the farmer the
cost of the latter.

The factory began a more systematic analysis of bean

deliveries in an effort to even out the peaks and troughs of raw
material supplies and to carefully monitor the quality of beans
on a sub-area basis. While control clerk remuneration was based
in part on the quantity of beans that their farmers delivered, a
refinement of the incentive system to link pay with various
quality characteristics was beginning to be developed.

Probably the most important initiative of 1986 was the

start-up (and then termination) of a new Hortiequip operation in
Kisii, contracting 3466 farmers at 18 collection centers. Since
1982, Kisii had been viewed as a potentially suitable locale for
the project; its ecology and high population density were seen as
suitable. It was felt, however, that area diversification could
not take place until the company had confidence in its organiza­
tional system and technical package. Another reason for moving
into Kisii in 1986 was to reduce political activism in Vihiga by
sending the people there a message that they are not the only
people who can grow French beans. A third reason for area
diversification had to do with staff considerations. The
production manager at Vihiga had been working in that position
since 1982 and decided that without a new challenge she would
probably quit. She was made general manager of the new Kisii
scheme.

Along with a few senior staff members, a group of the local

staff members of the Vihiga operation were brought to Kisii to
train local people and to serve as supervisors. Farmers in the
Kisii area generally have 2-3 acres of land and more significant
cash-crop earnings than Vihiga farmers. Tea is widely adopted
here, and here are 5 local tea factories. Banana production for
sale is widespread. Many local farmers decided to try French
beans, however, because of the shorter production cycle and to
spread overall risk.

Engendering farmer interest had not been the most important

problem of the new project in Kisii. The Kisii Hortiequip
management reported that the main problems stemmed from staff
dishonesty and the uncertainty of local political support.
During the first season the company received a large number of
fake receipts from its collection-center staff. Even where the
cases of fraud could be proved, local pressure on the political
authorities prevented sanctions being applied against the
offender. As a response to this situation, the company adopted a
new system whereby receipt books would no longer be held at the
collection centers. Instead, collection center staff merely
recorded the names and weight deliveries of farmers each day and

58
submitted summary papers to the local Hortiequip office. At the
office the beans were weighed and receipts were written out.
When a farmer had made 10 deliveries he would receive that number
of receipts from the office. In this way the company was
responsible for payment only for the quantity of beans actually
delivered to its office. (37)

The Kisii project seemed to be encountering more problems of

a political nature than were faced in the establishment oI the
Vihiga operation. In late August a speech was made by a leading
government official claiming that a "businessman ... has
introduced a new crop to Kisii farmers and is failing to pay for
the product delivered to him.... Nobody should be left to feed on
others' sweat without working for it." The official's descrip­
tion of the offending "businessman's" operation suggested that he
was referring to the Hortiequip project. Neither the owner of
Njoro Canners nor the Hortiequip staff were contacted, however,
and when an inquiry was made, it was neither confirmed nor
disclaimed that the official's comments were directed at the
French-beans pro)ect.

During the 1986 2nd season conflicts between Hortiequip

management, staff, and local politicians increased. At one point
the staff actually went on strike to protest against their
treatment by management. Complaints were being made against
Hortiequip by both the Kisii District Commissioner and individual
chiefs. An investigation by the Njoro Canners' project manager
revealed that the Hortiequip management in Kisii had been acting
in a dictatorial fashion, delegating little authority to staff,
limiting information flows to downward orders and upward reports,
and generally rejecting a priori potentially legitimate com­
plaints by staff and farmers. The tension that was building up
between the company and the local people was making productive
results impossible. The Kisii project was closed at the end of
1986, although local officials, staff, and farmers were told that
it was possible that Hortiequip would return to the area at some
future date under different management.

The Uncertain Future

As Njoro Canners looks to the future, it appears that

effective continued operations will depend on the sustained
involvement of a few individual senior staff members who have
nurtured the project from the beginning. Several of these people
are expatriates. Efforts to train local staff for senior
management positions in the factory have thus far not been
successful. It also appears that the project will remain
politically vulnerable. The project's growth has led it to
become an important force in the regional economy where it
operates. Such an important presence has made the company

59
vulnerable to individuals seeking political gains either by
drawing on company resources or by criticizing the company. The
company has periodically been labeled an "exploiting middleman."
It operates within a larger political environment where farmers
are always "right" and companies (particularly foreign and Kenyan
Asian owned) are always "wrong" when any dispute arises.

The relative success of Njoro Canners has led many Kenyan

entrepreneurs to consider establishing competing French-bean
canneries. Projects have been proposed for Kakamega and for
sites in Rift Valley and Central Provinces. In one case a major
Belgian canning firm was considering a joint venture. Whether
any significant investment will be made is not certain. The
country's existing processing plants generally operate at well
under capacity. Improved coordination between producers and
processors is needed. Additional processing capacity is probably
not required.

Even if technical, organizational, competitive, and politi­

cal problems can be solved, the long-term prospect for the
project hangs in the uncertain shadow of particular technical
developments in Europe that could virtually negate Kenya's
present comparative advantage. There is some danger that Njoro
Canners will lose its cost advantage for supplying French beans
to the French market. Several European seed breeders have
developed a hybrid variety of green bean containing many of the
quality features of the French bean but the one-flush yield
feature of the bobby bean. Having one flush permits mechanical
harvesting. While this mixed variety has a slightly different
taste from the pure French bean and while mechanical harvesting
does lead to more damage and the presence of foreign matter, the
new variety can generate a canned product at 15-20 percent below
the cost of the Kenyan prodcct. This lower-cost mixed-variety
product could well draw away a considerable part of the luxury
extra-fine market demand. The latter would remain, but as a more
narrow market segment.

The time frame for such developments is uncertain. There is

presently inadequate quantities of the hybrid seed to meet
existing demand. Commercial production using the new variety was
unsuccessful in 1986, largely as a result of a drought in the
south of France. Njoro Canners may be "safe" until perhaps 1990.
Saupiquet and Njoro Canners management have decided to reduce the
risk associated with these technical developments. Njoro Canners
product line will be expanded, and Vihiga farmers will be
contracted in 1987 to grow both French beans and gherkins.

60
ConcludinQ Remarks

This case study of Njoro Canners/Hortiequip highlights the

following points about contract farming and research on this form
of organizational arrangement:

1) The contracting company must seek to develop an organi­

zational framework that improves farmer productivity and then
strive to progressively reduce the transaction costs arising from
this arrangement.

2) A system of smallholder contracting will generally

involve high transaction costs and "leakages" (whether of money,
inputs, or product), but the basic economics of crop production
may limit the company to this high cost option.

3) In smallholder contracting systems the effectiveness of

extension staff is of critical importance. Analysis must go
beyond company-farmer relations and examine company-staff and
staff-farmer relations.

4) Contract farming systems go through potentially signifi­

cant structural and/or policy transformations in response to or
in anticipation of internal project developments or external
events. Examining the rationale and impact of these adjustments
is crucial in understanding the "life cycle" of a contract­
farming project.

5) Under circumstances where contracts with neither farmers

nor staff are truly enforceable, the contracting company must
develop the capacity to "migrate" locationally.

6) Even where a company adopts an apolitical line, politi­

cal considerations necessarily intervene in smallholder projects.
Local political support proves essential for success, yet company
success tends to breed political opportunism.

61
 Notes

I Based on interview with Gilbert Bintein, General of Njoro

Canners on September 11 and 12, 1986.

2 Marketing in Europe, October 198b. Special article on the

vegetable canning industry in France.

3 Bintein.

4 Marketing in Europe, p. 58.

5 Ibid., op. cit. ; Bintein.

6 Saupiquet Annual Report 1985; Marketing in Europe, p. 60.

7 Calculated from confidential Njoro Canners data.

8 Based on information provided by Bintein.

9 Ibid.

10 Interview with Mr. Wadhwa of Corner Shop Ltd. on September 10,

1986.

11 Ibid.

12 This and the subsequent two paragraphs are based on the

interview with Mr. Wadhwa and interviews with former staff of
the Corner Shop and farmers in the Chango area of Vihiga who
participated in this scheme. Our rather negative findings
contradict the fairly rosy picture of the project presented in
a 1982 FAO document entitled "The Private Marketing Entre
preneur."

13 Bintein interview.

14 The social problems of a large female labor force were

emphasized by several Lake Naivasha farmers during interviews
held September 13-15, 1986.

15 Bintein interview. Mr. Wadhwa was later dropped from the

project as he was unable to finance the Hortiequip operation.

16 Kakamega District Development Plan 1984-1988.

17 Agriculture Development Plan for Vihiga 1968-1972.

18 Ministry of Agriculture, "The Marketing of Fruit and Vege

tables in Vihiga" 1969, p. 3.

62
19 Kakamega District Development plan, p. 5.

20 Development Alternatives Inc. (DAI) "A Strategy for the

DevelopmEn.t of Four Districts in Western Kenya" 1982.

21 Quoted in Martin.

22 Martin, p. 164.

23 ibid., p. 167.; Also based on our survey in Vihiga, September

19 6.

24 As reported by senior staff of Hortiequip (Vihiga) during

interviews of September 17 and 18, 1986.

25 Based on survey of Vihiga farmers.

26 DAI, p. 41; Kakamega District Development Plan.

27 Martin, p. 165.

28 MOA, Marketing of Fruit and Vegetables in Vihiga, p. 3.

29 Interview with production manager of Msambuani Industries on

Sept. 23, 1985.

30 One factor explaining the relative productivity of farmers in

the project is certainly control of clerk e±fectiveness. This
can be illustrated by examining results from two of the sub­
areas where we interviewed farmers--Chango and Mbale. Each of
these sub-areas has a collection center with more than one
control clerk operating out of each. During 1985 Chango
actually had four control clerks (each with more than 125
farmers) while Mbale had three control clerks (each with more
than 200 farmers). The characteristics of the farmers
attached to individual control clerks at these stations are
basically the same. Control clerks aren't allocated a
particular territory, so geography isn't a factor. All
farmers at one center plant and harvest at similar times.
Experience in the project should not differ according to which
control clerk a farmer is attached to. Thus differences in
average yields between farmers at the same locale but with
different control clerks can be largely explained by the
relative effectiveness of control clerks.

We lack individual farmer data and have only the mean

yields for each collection center. While the data shown below
do show variations in performance according to control clerk

63
at the same centers, in the absence of calculations of
standard deviations we cannot claim statistical significance.

1985 Yield Variations at Individual Collection Centers

Area/Code Combined Seasonal Yields Index

Chango A 188.9 kgs 100

B 188.6 99.8
C 175.3 92.8
D 161.8 85.6

Mbale A 181.9 100

B 158.2 86.9
C 151.7 82.9

While in Chango the aggregate performance difference between the

best and the worst control-clerk group was over 14 percent, in
Mbale it was over 17 percent.

31 Farmers interviewed in our survey complained of this problem.

32 Staff of Hortiequip (Vihiga).

33 Ibid. In only one of the 18 sub-areas where the project

operated did average farmer yields approach those obtained
elsewhere in Kenya for French beans. This was the Mbale area,
which had an average yield of 76.6 kgs per I kg of seed, but this
relatively good annual average stems largely from the excellent
results of a late year experiment whereby farmers were provided
with only 1/2 kilo of seed for a planting. These farmers
obtained an average of 80 kgs of beans or 160 kgs per 1 kilo of
seed. Prior to this experiment results in Mbale had not been
good. Over the entire year Mbale farmers wre provided with
inputs valued at Ksh 8755 and the areas's input balance for the
year was Ksh 4177 or 47.7 percent.

34 Hortiequip staff.

35 The project proposal predicted average yields of 100 kgs per

kilo of seed. None of the centers reached this averaqe during
the season.

36 Calculated from Hortiequip and Njoro Canners Records.

37 Interviews with Hortiequip (Kisii) staff, September 21-22,

1986.

64
 Published Sources

Development Alternatives Inc. "A Strategy for the Development of

Four Districts in Western Kenya." May 1982

Economist Publications.
Marketing in Europe.
"Canned Vegetables in France" October 1986.
"Prepared Foods in France" April 1981.
"Prepared Foods in France" April 1986.

Food and Agriculture Organization. The Private Marketing

Entrepreneur. Rome. 1982.

Hormann, D. and Thuo, J. "Smallholder Production and Marketing

of Vegetables for Export in Kenya." Unpublished paper. January
1979.

Martin, C.J. "The Agrarian Question and Migrant Labor: ThE Case
of Western Kenya," Journal of African Studies v. 11 no. 4. 1985.

Minstry (,_ Agriculture. 'The Marketing of Fruit and Vegetables

in Xenya: Case Study Number One Division of Kakamega District,
Western Provi, ." 1969.

Moock, Peter. ,he Vihiga S.R.D.P. Farm Level Survey: A Prelimi­

nary Report of Findings." I.D.S. Discussion paper # 111.

Staudt, Kathleen. "Inequities in the Delivery of Services to a

Female Farm Clientele: Some Implications for Policy." I.D.S.
Discussion Paper #247.

65
THE DEVELOPMENT OF THE KENYA/UNITED KINGDOM TRADE IN "ASIAN
VEGETABLES" WITH PARTICULAR REFERENCE TO THE PROBLEM OF
COORDINATING PRODUCTION AND EXPORT MARKETING IN KENYA

"Contract Farming in Africa"---Kenya Case Study # 2

67
 Since the late 1950s, Kenyan farmers and exporters have
airfreighted fresh fruits and vegetables to Western Europe and
the Middle East. Kenya's horticultural exports have included
several dozen tropical and temperate fruits and vegetables and at
any one time over forty different items may be exported. While
becoming increasingly important to the Kenyan economy, the value
and volume of Kenya's fruit Pnd vegetable exports has remained
relatively small compared to the supplies provided to the EEC by
nonmember Mediterranean countries such as Morocco and Israel.
Nevertheless, Kenya has held a predominant position for nearly
twenty years in one segment of the European market. This is the
market for "Asian vegetables," the group of vegetables that form
an important part of the traditional diet of several South Asian
communities and that are still widely consumed by various South
Asian and other immigrant communities in Western Europe,
particularly in the United Kingdom.

Kenya exports up to twenty different Asian vegetables.

These vegetables fall under various classifications. Some are
beans (i.e., valore, gwar, and chola), while others are peas
(tuwer and papdi), capsicums (various types of chillies),
marrows/gourds (dudhi, gisoda, tura, tindori, and mooli), leaves
(chillie leaves and patra), fruits (aubergine and gunda),
cucurbits (karela and tindola) , or fit into additional
categories (i.e., okra, tindo, saragwa, and gingra). Asian
vegetables such as chillies and okra have recently become popular
among sections of the indigenous West Curopean consumer
population, while other Asian vegetables are consumed almost
exclusively by immigrant (or second generation) communities.
Very few of the Asian vegetables have an "identity" of their own.
Instead, they are typically consumed and nearly always
distributed in combination with one another. For marketing
purposes these vegetables comprise a comprehensive basket.

This study emamines the Kenya/U.K. trade in Asian vegetables

beginning at the consumption stage in the U.K. and tracing back
the marketing chain to smallholder production in Kenya. For the
U.K.I examine the source and growth of demand for Asian
vegetables and the structure and characteristics of the Asian
vegetable marketing system at retail, wholesale, and import
levels. For the Kenyan dimension of the trade I begin by
examining the structure of the export trade and the political
environment in which it operates. I go on to discuss various
contributions of the Asian-vegetable trade to the Kenyan economy.

Special attention is then given to the problem of

coordinating production and export marketing in Kenya. I examine
the inefficiency and instability of the ties between farmers and
exporters, which are now beginning to undermine Kenya's
competitive position in the U.K. market. This unstable
production and marketing system serves as the backdrop to one

69
company's attempt in the early 1980s to introduce an Asian­
vegetable procurement system based on contract farming. I
exarmine the circumstances surrounding this contract farming
project, its performance, and its impact.

The Origin of the Trade

The historical origin of Asian-vegetable production in Kenya

is unclear but probably dates to the late 1860s or 1870s with
the demand and seed coming from Indian merchants based in
Zanzabar and along the Kenyan coast. The demand for these
vegetables increased significantly by the turn of the century
when thousands of people from the Indian subcontinent were
brought as indentured labor into the area to work on the Uganda
Railroad. Those workers not stricken by smallpox or malaria and
not mauled by lions needed to eat. By 1898 the rail line had
reached a place called Makindu (210 miles west of Mombassa),
where a camping station for the workers was established. By this
time 13,000 Asians were working on the rail line, requiring 21
tons of food per day. A major drought necessitated that
practically all food rations be brought up from the coast. It is
likely that the presence of these workers not only stimulated
Asian-vegetable production at the coast but also encouraged some
farmers along the Makindu River to grow these vegetables. The
total number of Indians brought over to work on the railroad was
32,000. ()

Only about one-quarter of the railway workers remained in

East Africa, but the local Asian population was augmented by
other immigrants from India. Most immigrants were from Gujarat
with smaller numbers coming from the Punjab. Many of the
immigrants were petty traders or children of peasants who were
released from the land. Many immigrants set up small
"dukawallahs" in the cities along the railway line selling basic
goods to Africans, Asians, and Europeans. (2)

With the expanding local Asian population, the market for

Asian vegetables was enlarged. Much of the production of these
vegetables was by Asian farmers in areas such as Kibwezi and Athi
River, who sent the produce by rail or truck to Nairobi and
Mombassa. By the 1950s, African farmers at Voi and the coast
also supplied the Mombassa market. Asian retailers would then
sell the produce to consumers. The local Asian-vegetable
marketing system predated the export trade by some forty years.

In the late 1950s and early 1960s there was a considerable

increase in migration of young Indian men to the U.K. A small
international trade in certain traditional Indian vegetablec was
initiated at this time. Vegetable traders in India sent small
consignments of produce to small-scale Indian companies in the
U.K., who then sold to the imm:grant community in London. Demand
continued to outpace supply. The U.K. Asian population expanded

70
rapidly in the 1960s as the initial male migrants were joined by
their families. Supply was constrained by insufficient air cargo
capacity between India and the U.K. (3)

During the 1960s Kenya emerged as an important source of

supply for the immigrant market. Actual Kenyan exports of
chillies and other Asian vegetables began in the late 1950s, but
this was only of tiny quantities. During the mid-1960s a few
Kenyan Asian-owned companies entered the export trade and
expanded Kenya's supplies of Asian vegetables. By 1969 this
export trade had reached over 750 tons/year.

The Kenya/U.K. trade in Asian vegetables has expanded more

than ten-fold over the past decade and a half, and this basket of
commodities has played an important role in the overall
development of Kenya's horticultural export trade. Over the past
fifteen years this group of vegetables has comprised over 30
percent of the volume of Kenya's fruit and vegetable exports.
For Kenya's most imporcant market, the United Kingdom, Asian
vegetpoles have comprised over 60 percent of the volume of
bilateral fruit and vegetable trade during the 1980s. Several
Kenyan firms that initially built their operations upon the
Asian-vegetable trade have since been able to diversify into
other products.

The Market for Asian Vegetables

The demand for Asian vegetables in Western Europe is heavily

concentrated in those cities that have sizeable communities of
South Asian ethnic origin. Given the strong historical ties
between the U.K. and the Indian subcontinent, the vast majority
of South Asian immigrants to Europe have settled in the U.K.
Smaller South Asian communities also exist in Amsterdam, Berlin,
and in each of the national capitals of Scandinavia. As nearly
95 percent of Kenya's Asian-vegetable exports have gone to the
U.K., we focus our discussion on that market.

The Asian Population in the U.K.

According to estimates made by the Office of Population

Censuses and Surveys (OPCS), there were over one million
residents in the U.K. in 1980 whose ethnic origin was in the
South Asian subcontinent. No British census has raised questions
on ethnicity, so estimates of the size and demographic structure
of the various ethnic groups in the U.K. population are based on
information about birthplace and parental birthplace taken in the
1971 census, up-dated and supplemented by estimates of births and
deaths, migration statistics, and information about ethnicity
raised in the 1979 and 1981 Labor Force Surveys. (4) The OPCS
provides the following estimates of the size and growth of the
local South Asian population during the 1970s:

71
 Population of Asian Communities in the United Kingdom
(Thousands of People)

Group Mid-1971 Mid-1976 Mid-1980

Indian born 307 390 460

E.African Asian 68 160
Pakistani/Bangladeshi 171 246 355

Total "Asian" Population 546 796 1007

(Source: OPCS Monitor PPI 81/6)

As can be seen, the local South Asian population nearly

doubled during the 1970s. Since data sets are not consistent it
is not possible precisely to subdivide this growth according to
net migration and net natural increase. Using data provided in
the issues of Internationt,,l Migration published by OPSC, however,
it is estimated that slightly less than half of this population
growth over this perioa was due to migration. Migration did play
a considerably more important role during the 1971-76 period than
in the latter half of the decade. This was particularly the case
for Asians who -migrated from East Africa.

Even with a decline in immigration levels, the population of

South Asian communities should continue to grow rapidly due to
their relatively high birthrates. While their numbers constitute
about 2 percent of the total U.K. population, over the 1977-83
period they comprised, on average, 4.5 percent of all live births
in the country. (5) Based upon prevailing population growth
rates, one Government report has estimated that the 1991
population of people of wholly Asian ethnic origin will be
between 1.25 and 1.50 million. (6)

The high birth rate of the Asian population combined with

the tendency for immigrants to have been concentrated among the
young has led to an age structure for the local Asian population
that is skewed toward younger ages. While 6.3 percent of the
overall U.K. population is between the ages of 0 and 4, a survey
among the Pakistani population in a section of Manchester found
20 percent of the population to be in thio age category. While
31.2 percent of the U.K. population is 45 years or older, the
sample of the Pakistani community found only 6.8 percent of the
group in this age category. (7)

The Asian "community" in the U.K. is actually a

"proliferation of distinct ethnic groups" with different
countries/regions of origin, different languages, and different
religions. At least three major religions, four major languages,
and four countries of origin can claim large groups among the

72
U.K. Asian population. The various groups differ in their rural
vs. urban origins, their settlement patterns in the U.K., their
occupational structures, and their dietary patterns. (8)

Dietary Habits

While the tastes of the younger generation are certainly

changing, a high proportion of immigrants of South Asian origin
continue to eat traditional foods rather than English foods.
This is partly out of sheer preference for certain foods and
partly to maintain their religious affiliations and social
customs. (9) A 1973 survey found that 79 percent of respondents
born in India or Pakistan ate only traditional foods at their
evening meals. The survey suggested that this eating p&ttern
would continue into the second generation. It found that most
Asian children were eating primarily traditional dishes with only
a small percentage preferring English food. (10) Demand remains
strong for certain traditional spices and vegetables and for a
growing array of Indian convenience foods that a few specialist
firms have supplied. A recent study estimates that the 1985 U.K.
market for Indian ingredients and processed foods was 40 million
pounds. (11)

Some vegetables are regarded as staple items in the South

Asian diet, while other vegetables are either delicacies or
spices/flavorings for which there are dried alternatives. The
staple vegetables include: carrots, okra, spinach, chillies,
peas, and potatoes. Other commonly eaten vegetables include
aubergines, karela, tomato, dudhi, and cabbage. Thus, one finds
some overlap between the traditional vegetable basket consumed by
the Asian population and that of the larger English population.
For some Asian vegetables there are locally grown substitutes.
For example, dudhi can be replaced by marrows and courgettes
while mooli can be replaced by radish or cabbage. There are no
acceptable substitutes for chillies, okra, or karela.

Vegetable consumption patterns differ for the different

subgroups. Gujarati Hindus are primarily vegetarians and thus
require greater quantity and more variety of vegetables. They
would be the prime consumers of items such as dudhi, gwar,
gisoda, papri, petra, tindola, and valore. Even when multiple
groups consume certain items their pacticular tastes moiy vary.
For instance Gujaratis use chillies as a pickle and thus want a
mild variety that has a nice dark green color. Punjabi Muslims
use chillies as a spice and thus require a pungent light green
variety. There are two main types of karela that are preferred
by different groups. Together with their different settlement
patterns (see below), these taste differences of the various
South Asian groups create a segmented market requiring a
specialist knowledge for effective distribution.

73
Concentrated Settlement Pattern

The local population of SoutY. Asian origin is concentrated

in a few major English cities. 1F.rly South Asian migrants
settled in areas experiencing Ir ior shortages either due to their
rapid rate of economic growth (-.e., Greater London) or due to
poor working conditions (i.e., in the Manchester or Leeds textile
industries). (12) South Asian communities are concentrated in
London, Birmingham, Bradford, Leicester, or Manchester. For
example, according to 1981 census data showing the regional
distribution of the local population according to country of
birth, of those born in India, Bangladesh, or Sri Lanka, over 53
percent reside in the Southeast, with over 38 percent living in
London alone. Different subgroups have had different settlement
patterns; for example, Pakistani Muslims are most heavily
concentrated in Birmingham, Liverpool, and Manchester, while
Gujaratis from East Africa ar- most heavi.y settled in Leicester
and parts of London.

In some cities the Asian population has come to form a

significant proportion of the overall population. For example,
between 1971 and 1980, the Asian population of Bradford rose 89
percent to reach an estimatEd 47,000. Its share of the city's
population rose from 8.4 percent to 17 percent. (13) For
Leicester, the 1981 census found that 19 percent of the local
population was born outside the U.K., of whom 80 percent were
born in India or East Africa.

The Asian-Vegetable Marketing System

General Features

Before examining the various levels in the Asian-vegetable

distribution chain it is appropriate to lay out some general
diitinguishing features of this trading network. Such features
include the following:(14)

1. Dominance of minorit, awned firms---

English fruit and vegetable importers and wholesalers have

played only a minor role in the servicing of requirements for the
country's ethnic minorities. Only recently have these firms
entered into the field of "exotics" primarily at the behest of
overseas marketing agencies. Conservatism, lack of understanding
o. a potential opportunity, as well as the preference for dealing
on a commission basis limited the participAtion of English firms
in the ethnic foods trade. The partiOular requirements of the
country's Asian and West Indian population have been met largely
by small-scale family companies with origins in these areas.

74
2. Fixed price system---

Unlike the general fruit and vegetable trade that until

recently has operated primarily on a commission/consignment
basis, the Asian-vc-getable trade has always operated with fixed
buying and selling prices. While import costs vary from item to
item, importers and wholesalers have tended to sell the wide
range of Asian vegetables at the same price level. This has
served to economize on transaction costs as price information
could be consolidated in one figure and the administration of
sales made essier. Traders look for an overall margin on their
basket of produce, and some items subsidize others. Prices may
remain steady for a considerable period. The most significant
influence on prices has come from factors out of the control of
participants--i.e., air freight rates and currency movements.

3. Quantity rather than price adjustment---

For Asian vegetables sent from Kenya freight, costs are

higher than f.o.b. prices and, for a low value item such as
aubergine, may be twice as high as the f.o.b. cost. Freight
costs account for nearly a third of the retail prices of these
vegetables. This, together with a system of minimum export
prices set by the Kenyan Government, determines that the trade
has minimal latitude for price reductions in the face of
surpluses. The limited spending power of most Asian communities
and the personal relationships between retailers and consumers
limits the scope for price increases in the face of shortages.
Demand patterns are consistent and relatively price inelastic.
The trade thus utilizes quantity adjustments rather than price
adjustments to match supply with demand. Given the relatively
small size of the trade, but the vast range of items exchanged,
shortages and surpluses of individual itejis are ever-present.

4. Fragmentation rather than concentration---

Unlike the general fruit and vegetable trade, which is

experiencing increased concentration at import, wholesale, and
retail levels and greater degrees of vertical integration across
stages, the Asian-vegetable (or ethnic) trade has experienced
greater fragmentation, particularly at import and retail levels.
At the import level a group of medium-scale, well-established
firms have faced increased competition from a large fringe of
small-ccale firms making deliveries direct from the airport to
retail shops. The Asian retail sector continues to proliferate,
reducing the clientele for each individual shop.

Retail Distribution

The retail sale of Asian veget-.les is predominantly from

the corner shop located in an inner-city area. Some shops carry

75
a wide range of spices and other foods and a more limited stock
of vegetables, while other shops are fresh produce specialists
and carry a bewildering array of fruit and vegetables, many of
which are unknown to the average Briton. Maiy cater to a
primarily ethnic-minority clientele. One survey of Asian
shopkeepers in three British cities found that, on average, only
30 percent of their customers were white. (15)

Asian retail establishments are a relatively recent

phenomenon in the U.K. Few Asian-owned shops existed prior to
the 1950s, and it was not until the early 1960s that there was
major growth in this type of investment within the Asian
communities. Desai (1963) reports that the first Gujarati grocer
in Birmingham started business in 1949 and that by 1961 there
were still only six Gujarati grocers. The growth of Asian
retailing was rapid in the 1960s and on into the 1970s with
individuals responding directly to the opportunities created by
the growth in the local Asian population to supply food,
clothing, and other items unknown to English shopkeepers.

The growth of the Asian retail sector occurred

simultaneously with a pattern of economic decay in some inner
city areas and with a trend toward increased concentration of
retail food sales. Some researchers have accounted for this in
terms of a comparative advantage of "ethnic entrepreneurs" in
servicing the needs of particular communities. The retailer,
dealing in a range of cultural items, goods, arid services, can
develop a certain niche that shields him from ,,tP.Ade
competition. (16)

Against this optimistic picture are a number o± studies that

have argued that the majority of Asian retailers are working
extremely long hours and generating relatively low returns. Not
only is the level of purchasing power within Asian communities
relatively low, but the expansion in the number of retail outlets
has spread the Asian consumer pound more thinly. There Ere now
too many businesses chasing the ethnic trade with insufficient
wealth in the communities to support the quuntity of retailers.
Their location in areas of high Asian population density limits
their access to the majority population. (17)

Indeed, for the past several years retailers have seen their
margins on Asian vegetables squeezed as costs of supplies have
risen with increased air freight costs, but heavy competition has
prevented them from passing on these cost increases to consumers.
Some retailers have encouraged local Asians to grow vegetables in
their backyards so as to provide them with a cheaper product and
thus some competitive edge. (18)

Many Asian greengrocers have on-goino orders from a

wholesaler who makes deliveries to the shop several times per
week. Generally, preference is given to suppliers who can

76
provide a full product range plus multiple-week credit. During
periods of peak demand or short supplies this retailer may seek
additional supplies from wholesale mdrket-based traders or small­
scale distributors with lower prices, but limited credit
arrangements.

During the 1980s the multiple chain supermarkets have shown

some interest in items sur, as okra and chillies to be Tncluded
in their overall range of "exotic" fruit and vegetables. The
volume of this flow is growing, but remains small.

The Wholesale Trade

Most of the wholesalers of Asian vegetables carry a full

line of fruit, vegetables, and spices for a largely ethnic
minority clientele. Based on interviews conducted during the
1984/85 winter, I have estimated that twelve firms, based in
13irmingham, London, or Bradford, account for three-fourths of
primary or secondary wholesale turnover for Asian vegetables.
Secondary wholesalers in cities such as Liverpool, Manchester,
Coventry, Leeds, and Glasgow will normally obtain their supplies
from Birmingham- or Bradford-based importer/wholesalers.

The history of many of these wholesaling firms has followed

a similar path. Most started their businesses in the late 1960s
or early 1970s importing small quantities of vegetables and
spices from contacts they had in India, Pakistan, Kenya, or
Cyprus. Initially, they used their own houses for storage and
made deliveries door-to-door to shopkeepers. The firms then
acquired warehouses and began selling near or in primary
wholesale markets. Contacts and business outside one's own
community and ethnic group were expanded as were the product
ranges of these firms. Even with the expanded clientele and
pr-duct range, Asian vegetables from Kenya have remained a key
compoi.nt of each firm's business, and the Asian retailer and
consumer the prime orientation. Due to the risks of importing
produce directly (see below), most of these firms have withdrawn
from this activity and now rely on a few importers for their
supplies.

Importing Asian Vegetables

The Asian-vegetable import trade began in the 1960s when

some Indian merchants began receiving prodtuce from India and then
selling it from the parking lot of Heathrow Airport. The
distribution of Kenya's Asian-vegetable supplies also focused
around the "parking lot merchants" until the early 1970s. At
that time several small companies began operating vegetable­
delivery services from vans. Distribution was still largely
concentrated in the London area, although significant Asian
communities were developing in the cities of the Midlands.
Eventually, several "van importers" established warehouses in the

77
vicinity of primary wholesale markets, reduced the extent of
their retail shop deliveries, and operated centralized
distribution systems. In 1973 a senior partner in Kenya's
largest export company emigrated to the U.K. and established an
import/distributing company. Un2.il the late 1970s this firm
would dominate the Asian-vegetab.e trade in the U.K. setting the
standards for quality, setting pric~e guidelines, and widening the
distribution network supplying the ethnic minority population.

In the 1980s the import trade for Asian vegetables has

featured a few long-established dominant firms and a large
competitive fringe of small-scale or specialist firms. Five
firms probably account for 65-75 percent of the U.K. imports of
Asian vegetables from Kenya, although no single firm has more
than a 20 percent share of the market. For some of the cities in
the Midlands, one or a few firms provide the bulk of imported
Asian vegetables. These and other firms also obtain Asian
vegetables and other "exotic" fruits, vegetables, and spices from
India, Pakistan, Cyprus, Zambia, Egypt, and South and Central
America. With one exception, each of the leading firms
speciaiizes in the importation of tropical fruits and vegetables
for distribution through "ethnic channels." Again with one
exception, each firm concentrates its activities in a single city
or small region. Each firm is Asian-owned and is typically a
family enterprise or two-family partnerhhip. Most of these
companies have an annual turnover of less than 5 million pounds.
These firms are based in either London, Birmingham, or Bradford.
Each fir.1i carries up to thirty individual products and deals with
perhaps 100-150 secondary wholesalers and retailers. (19)

In addition to these five firms there are a number of

smaller importers of Xenyan produce. Some of these firms are
primarily wholesaling operations that merely obtain part of their
requirements directly from overseas. Others are the modern-day
"van importers," who pick up small consignments of produce at
Heathrow Airport and make deliveries by van direct to retailers
in London and other cities. There may be up to fifty "van
importers" operating in the country. These companies generally
do not maintain stocks of produce, preferring immediate turnover.
Some of these firms are permanently in the trade while others are
simply "cowboy outfits" operating part-time or seasonally and
commonly stopping and starting up under a range of different
names.

The "van importers" have proven to be parti,;ularly important

in serving the London market. During the winter months they may
have a combined 30 percent share of this market. With
insignificant overheads and by nypassing the wholesale trade, the
van importers have been able not only to undercut in price some
of the larger importers, but also to provide the service of
delivery to retailers, but the "van importers" are not in a
position to offer the extended credit terms that larger

78
importer/wholesalers may offer. The competitive fringe provided
by the "Van importers' has reduced the market power of the larger
firms. It has also undermined previously stable Lrading
relationships between established importers and their customers,
however. Some of these importers have found it to be
unprofitable to continue in the Asian-vegetable trade and have
diversified their product range into more profitable items.

As is the case for a few of the leading firms, many "van

importers" are linked to family or friends in Kenya. While the
larger exporters tend to deal with the larger, well established
import/wholesaling firms, many of the small-scale, part-time
exporters have traded with the small importers willing to handle
a consignment of a ton or less. As the small-scale firms on each
side of the trade operate with limited overhead costs, they have
been able to undercut the riore established firms. While on the
surface this appears to he a sign of "healthy" competition and
the reduction of "inefficient" market power, e)amined more
closely this pattern appears to be reducing the incentives to
participate in the trade for those who are most able more fully
to develop the market.

Importer Dissatisfaction

U.K. Asian-vegetable importers have relied heavily on

supplies from Kenya. Some produce is procured from European or
other countries, but their production is either highly seasonal
(i.e., Cypriot okra), relatively more expensive (i.e., Dutch
aubergine; Indian, Mexican, and Brazilian okra), or not of the
varieties preferred by the local Asian consumers (i.e., Indian,
Pakistani, and Cypriot chillies). Given their highly specialized
product range, these importers are vulnerable to supply
disturbances on the Kenya side.

Importers generally have informal, "gentlemen's" agreements

with one or two Kenyan exporters to send a specified quantity of
each of 20-30 items, a specified number of times per week. These
will be on-going orders that might be subject to adjustment on a
weekly or bi-weekly basis via telex communication. The p yment
schedule for importers is related to their size of purchase,
although two weeks credit is the norm. Small importers may have
to prepay for their orders a week or consignment in advance.
Where transactions are between family companies, the importer may
have payment periods of up to three months. While some i±porters
have dealt with the same exporters for a number of years, most
importers report that they have shifted among several suppliers
in the past few years.

Most U.K. importers are dissatisfied with the ability of

their Kenyan suppliers to meet their requirements for quantity,
quality, and continuity of supplies. Many firms view Kenya as
the least reliable of the main source countries supplying the

79
overall U.K. fruit and vegetable market. Some firms see this
problem stemming primarily from the bottleneck in international
transport in Kenya during the peak winter supply months. In
recent years during the peak export months of December and
January, there has been inadequate air-cargo space for fresh
produce leaving Nairobi, particularly that bound for the U.K.
market. Significant quantities of produce have been wasted,
gone unharvested, or off-loaded from airplanes. (see below)

Other importers see their difficulties stemming from the

practices of exporters. They feel that certain exporters have
inadequate commitment to their customers and will chase short­
term profit-generating opportunities even at the expense of
"loyal" customers. In a large number of trading relationships
there exists a strong element of distrust. There are certainly
exceptions to this state of affairs, but most importers feel that
many exporters simply cannot be relied upon. The poor services
provided by some exporters has tended to generate external
"diseconomies" for the overall reputation of Kenya as a supply
source and has undermined the position of the more competant
firms.

Importers generally face produce quantity and quality risks

rather than price risks. Within an overall climate of distrust
importers perceive that the general rules governing the trade
include the following:

1. At particular times importers will not receive any

produce at all from particular suppliers because a) of a failure
on the part of the exporter to secure sufficient airfreight
space, and produce that is sent will go to preferred customers;
b) cargo is off-loaded at the airport in the last minute scramble
for cargo space; and/or c) the exporter has located another buyer
who is offering better terms and has thus redirected the
consignment. Under this condition importers will need to make
purchases from other importers to satisfy at least their most
important customers.

2. Importers may not receive produce of marketable quality

because a) the consignment has been transshipped and subsequently
deiayed; b) produce has dehydrated due to heat build-up within
the carton during hot periods; and/or c) the rupplier has failed
to grade and pack the produce properly. There is no standardized
Kenyan product; quality levels differ by grower and exporter and
quality varies week by week. Under this condition importers can
make claims against the exporter for the produce that is
unmarketable, and obtain a certificate from the local inspection
services to that effect. Most importers must be careful about
making "excessive" claims against exporters, however, or the next
consignment may "fail" to arrive at all. If quality
deterioration was due to transshipment, the importer still must
prove that the relevant airline was at fault.

80
 3. Importers will not generally receive what they have
ordered because a) within Kenya some items may be in short supply
while others have been harvested in surplus. The exporter will
seek to meet the quantity of the order by including larger
quantities of the surplus item within a consignment; and b) even
when communication is made to the exporter that particular items
are short or flooded, a lag time of a few days normally follows
before a noticeable response is made. Importers are particularly
worried about shortages, as these result in dissatisfied
customers. Some firms report over-ordering those items that are
"traditionally" under-consigned. Normally surplus produce can be
sold at cost.

Importers have no legal or other institutional remedy

against a supplier who willfully breaches an agreement. One
sanction, important in many trading relationships, is the threat
of lost future trade. This threat is powerful only for the
handful of larger importers. Another possible sanction is the
threat of "advertising" the wrongdoing, thereby undermining the
reputation of the exporter. This sanction seems to have cnly
limited value as most f2rms are painfully aware that many trading
relations may go sour for a variety of reasons and that the
breach of one agrecmcnt should not greatly damage the reputation
of a firm. Only repeated breaches of agreements should lead to a
firm getting a "bad name." Furthermore, some exporters have a
very short time horizon with their prime interest being the
generation of rapid seasonal profits and perhaps the export of
capital to overseas bank accounts.

Importers generally can spread these quantity and quality

risks by procuring produce from more than one Kenyan supplier as
well as from one or more suppliers in another country. Some
importers and wholesalers have encouraged British farmers to grow
chillies during the summer months. Still others have invested
their own resources in production schemes in such countries as
Mauritius and Egypt and even in the black "homelands" in South
Africa. Importers do not expect that alternative supply sources
will initially be able to match the quality of Kenyan produce.
Nor do they expect these sources to compete well initially with
the Kenyan supplies on the basis of price. Increased reliability
and continuity is the central objective in diversifying away from
Kenyan supplies.

The Asian-Vegetable System in Kenya

The Export Trade

Kenyan exports of fresh fruit and vegetables were introduced

during World War II with supplies going to Allied troops
stationed in East Africa and the Middle East. The export trade
to Western Europe began in the mid-1950s with the expansion of

81
commercial air transport. The European trade was initiated by
the European-managEd Horticultural Cooperative Union, which sent
supplies on consignment to firms operating out of London's Covent
Garden market. In the mid-1960s a few Kenyan Asian-o-ed firms
began exporting Asian vegetables and other items to the U.K.
These firms either had been local fruit and vegetable wholesalers
or had sizeable vegetable farms. The fruit and vegetable export
trade can be characterized by four major features: I) the
dominant role of the private sector; 2) the limited role in
export marketing of African-owned and managed firms; 3 its
fragmented structure; and 4) a major international transport
constraint. We touch briefly on each of these
characteristics. (20)

1) Private Sector Dominance

For nearly all agricultural crops and products marketed

domestically in Kenya or exported, the Kenyan Government has
played a substantial role either through price or territorial
controls or through direct involvement in physical marketing
activities. In contrast, the role of the Government in the
development of the fruit and vegetable export trade has been very
limited. In 1967 the Horticultural Crops Development Authority
(HCDA) was created. Linked to the Ministry of Agriculture, it is
a specialized parastatal empowered to regulate, control, or
involve itself directly in virtually all aspects of horticultural
production, processing, and marketing. While given extensive
legal powers, the HCDA has never received sufficient funding or
manpower to carry out most planks of its broad mandate. Its
prime activities have been a) periodic support for smallholder
horticultural production schemes, b) domestic marketing cf
onions, c) export licensing, and d) monitoring and regulating the
export trade.

The Authority entered marketing directly, not in pursuit of

an explicit policy objective, but primarily in order to raise
revenues to cover its operating costs. This occurred first in
the domestic marketing of onions and later in a small-scale
export operation. Still, the actual exports of the Authority
represent no more than I percent of total horticultural exports.

2) Limited Role of African-owned and Managed Firms

Since its initiation, the horticultural export trade has

been dominated by firms owned and managed either by Europeans or
Kenyan Asians. Kenyan Africans have had a minimal role in export
marketing. The HCDA has long maintained a liberal export
licensing policy in order to encourage potential exporters,
particularly African-owned firms. During the 1970s and 1980s
several African firms have entered the export trade. Some of
these were owned by civil servants and their wives. Most of the
African-owned firms have experienced considerable difficulties

82
and have withdrawn from the trade. These firms either had
difficulty obtaining adequate air cargo space, had insufficient
marketing experience and market contacts, or didn't get paid by
overseas buyers. The managers of these companies tenden to
divide their time between this business and several other
endeavors, further constraining their ability to ectablish a
stable position in the trade.

In line with a general Government policy for the economy,

there have been frequent calls for the "Kenyanization" of the
trade, sometimes made from fairly high levels in Government. As
all leading firms are already majority-owned by Kenyan (Asian or
European) citizens, the term can only be interpreted as a call
for "Africanization." The liberal licensing policy introduced in
the 1970s was not succeeding in reducing the dominance of firms
owned by non-Africans

In the late 1970s, export companies were put under pressure

to take on influential African personalities from public life to
"participate" in their operations. Failure to do so would have
resulted in the termination of one's export license. Generally,
ouch participation did not involve capital investment. Instead,
the "personalities" were paid service fees for providing come
measure of protection and support in overcoming bureaucratic
hurdles. Some of the "personalities" have been provided support
from their companies to develop their own farms.

Nine firms continue to account for 85-90 percent of the

volume of Kenya's fruit and vegetable exports. With the
exception of one European-managed company, each of the other
leading fruit and vegetable exporting companies is owned and
managed by Kenyan Asians. Only a few of these firms have
Africans in senior management positions, although their overall
staffs are largely African. African-owned firms have a combined
share of less than 7 percent of export volumes. Asian-owned
firms conduct nearly the entire trade in Asian vegetables.

Frustration of the official policy has led to recent

discussions about "transferring" the trade from established
exporters to rising Kenyan African entrepreneurs. This would
involve selective licensing, increased scrutiny over various
practices of existing exporters, and provision of preferential
treatment to a limited number of well-connected African-owned
firms. (21)

In recent years there have teen numerous official statements

deploring the pricing policies of fruit and vegetable exporters
as well as their alleged failure to repatriate the "rightful
share" of foreign exchange earnings to Kenya. At times, these
statements have taken on a strident line with claims made that
these exporters were "plundering of the economy." These publiz
attacks have generally questioned the integrity of the entire

83
industry and have not made distinctions between offending and
non-offending firms. At the same time some firns have been
accused of "exploiting" farmers. Most export firms see their
investments and future livelihood as being vulnerable to
politically-inspired interventions.

A few of the export companies have family living in the

U.K., which enables these firms to economize on the transaction
costs of export marketing. Some exporters deal directly with
affiliated family companies while others get assistance from
family members through the provision of market information and
perhaps through debt collection. Firms with family links are far
less vulnerable to various forms of importer opportunism. Many
Kenyan exporters have had consignments not paid for or had
importers make large claims on the basis of poor quality or
noncompliance with their orders in terms of product mix. Those
Kenyan firms that deal with family members overseas have not had
to "chase" their money or be subject to large claims. They have
also been less exposed to exchange-rate risks than o+her firms.
Their U.K. affiliates will generally absorb the deviations
between orders and actual deliveries and swallow their normal
margins whenever procurement costs have risen temporarily.
During periods of financial stress the overseas affiliate can
inject capital into the local operation by prepaying for orders.

3) Fragmentation of the Trade

In the 1960s the number of firms exporting fruit and

vegetables waE probably less than a half dozen. Since t!,e early
1970s the number of licensed exporters has mushroomed to over one
hundred. While not all licensed exporters do engage in trade,
and while only a limited number of firms contribute a large
proportian of overall export volumes, it can still be argued that
the Kenyan export trade is too fragmented either to maintain or
to expand Kenya's trade position.

The majority of firms holding export licenses have been

part-time exporters. They export only during short periods of
the year and/or are involved in this trade only as a
supplementary activity to selling tea, running a travel agency,
or holding a civil service job. Over the years many "cowboy
outfits" have sprung up in search of quick profits in this trade.
Their scale of operation warrants neither the investment in
marketinq infrastructure nor the investment n building up stable
relation-hips with growers and overseas buye.rs. Most firms have
neither the capacity nor the inclination tri plow back export
earnings into the horticultural sectox. Most firms have
insufficient turnover to obtain an economical return on
precooling and cold storage facilities or on the development of
their own extension staff.

84
 The fragmentation of the trade results not only in Kenyan
firms scrambling for farmer produce and air cargo space, but also
competing against one another for the same markets.
Fragmentation has also served to undermine the reputation of
Kenya as a supplier. The quality of produce and associated
services varies by exporter with small-scale, ad hoc exporters
not generally being able to satisfy importer requirements. This
undermines the overall image of the Kenyan trade and acts as a
"drag" on the business of the more competent firms. Regular,
long-term marketing relationships have been somewhat undermined
by the presence of an array of firms operating with minimal
overheads and able to offer produce in the short run aL a
discount. The fact that the HCDA tends to pass on to new
exporters the names and addresses of the overseas buyers of
existing exporters does not help preserve these stable trade
relations.

4) International Transport Constraint

Throughout most of the history of the horticultural export

trade, firms have had to contend with limitations on
i.nternational cargo facilities out of Kenya. Although it was
first muted as an idea in the early 19/.7., has been discussed
repeatedly since that time, and has been developed extensively by
other horticultural exporting countries, international sea
transport of Kenyan produce has never developed. The Kenyan
horticultural export trade has been based entirely on air
freight.

The -rvsonal inadequacy of air cargo facilities was felt as

early as t,.e 1950s, but th,. introduction of wide-bodied carriers
and a few charter lines in the 1970s was able to handle much of
the expanded production and trade. Still, access to air cargo
space proved problematic for smaller firms (lacking permanent
"reJationships" with airline cargo staff), especially during the
peak export months. Air cargo space has increased in the 1980s,
but not nearly as much as has the demand for it. Air cargo
limitations are felt throughout the main October-June export
period, but particularly during November to January. Most
produce going to the U.K. market must be transhipped via other
European countries.

The reasons for the air cargo shortage are many and the
problem can not be discussed in detail here, but a short list of
causal factors m.'qht includet
i) -,ne weak direct involvement of Ke,ya Airways in carrying
horticultural cargo and its obstruction of cargo plans pro osed
by alternativ, charter and IATA airlines;
2) the high customs duties on imports into Kenya that have
reduced the south-bound ciirgo traffic from Europe, and thus cargco
space for the return journey;

85
 3) the restrictions on charter licensing and permissible
types of cargo on charter flights as laid down by the Kenyan
Civil Aerodromes Board;
4) the high fuel charges to airlines relative to those
charged in other African countrie, with higher fuel taxes being
imposed against charter airlines;
5) the Kenyan Government's controlled freight rates for
horticultural produce, which are below IATA rates;
6) the growth of the Kenyan flower export trade. As freight
charges for flowers are higher thnn for fruit and vegetables, the
airlines prefer to take flowers; and
7) the growth of air cargo requirements out of South Africa
as a result of an expanded horticultural trade and increased
emigration due to the political situation. Most commercial
airlines stopping in Nairobi initiate their flights in
Johannesburg.

At any one time thirty or ior-t exporters may be seeking to

get cargo space from the commercial airlines. The airlines may
give several firms an indication of available space, but this is
subject to ctange as produce up-take from Nairobi will depend on
cargo up-take from previous stops 'particularly Johannesburg) and
passenger load. The competitive, last-minute scramble for space
is accompanied by varinus malpractices and a high level of
uncertainty for those firms that lack a strong pe.-cinal
relationEhio with th.. airline cargo staff.

Growth and Contribution of the As-an-Vegetable Trade

The Kenya/U.K. trade in Asian vegetables expanded

considerably from the late 1960s until 1983. (22) Since then
there has been a decline in the level of trade. The growth of
the Kenya/United Kingdom trade in Asian vegetables can be seen in
the following table:

86
Kenya/U.K. Trade inAsian Vegetables (Tons)

Year Auber- Okra Chillies Karela Mooll Dudhi Other SubTotal Total
gine Asian
Veg.
1968* 30 99 158 289 576
1978* 98 82 274 ------ 613 1067
i72* 746 151 471 -- -419 1787
1974 1068 152 688 181 250 98 115 3144
1976 1021 263 882 387 235 201 1184 4093
1977 126u 300 980 215 307 171 1126 4359
19/8 1382 361 1209 515 371 257 1223 5318
1979 1622 735 1508 661 365 295 ** **
1980 1618 812 1340 758 241 295 1544 668
1981 1666 978 1328 840 145 346 1554 6857
1982 1887 1121 1563 962 126 360 1664 7683
1983 2847 1627 1746 9680 181 477 1964 6942
1984 1767 1506 1625 914 30 571 2R 7 8440
1985 1701 1278 1940 979 4 523 1534 7959

(Source: HCDA Trade Statistics)

*Kenyan exports to all destinations. U.K. probably accounted for
over 90 percent of these totals.
** Data riotavailable.

As can be seen in the data, there has been fairly continuous

growth in the overall trade in Asian vegetables. (23) For, some
items, trade volumes have stagnated or declined over the past
five years. This is most notable for mooli. The market for this
product hms largely been taken over by cheaper Italian, Dutch,
and British supplies. Kenya's market share for aubergine has
kbeen substantially reduced as the bulk of increased U.K. imports
have been provided by the Netherlands and Spain. (24)

The trade downturn for 1984 and 1985 (and now 1986) reflects
changes on both the supply and demand sides. The major decline
in 1985 partly reflects the impact of the 1984 drought. The
declining Asian-vegetable exports are also a result of the
reduced emphasis that a few leading exporters have placed on
Asian vegetables as part of their overall export basket. These
exporters have placed increased attention on the procurement and
sale of higher-value items, particularly french beR,.s. On the
demand side Kenya is beginning to face increased -ompetition from
European and non-European sources for okra and chillies. While
Kenya still retains a competitive advantage due to its ability to
provide the full range of Asian vegetables, many alternative
sources are beginning to eat away at the virtual monopoly
position that Kenya once held in this market. Impoxter
dissatisfaction with the reliability and continuity of supplies
as well as the unevern quality of Kenyan produce is pushing this
source diversification at a faster pace.

87
Foreign Exchange Earnings

Using the HCDA's minimum export prices as a guide to actual

sales earnings for these crops, one finds that the foreign
exhange earnings for this group of vegetables have been the
following:

1981 Ksh 47.3 million

1982 54.5 million
1983 67.0 million
1984 76.3 million

The minimum export prices may understate the actual value of

sales by 10-20 percent. Even disregarding this fact, in 1984 the
export earnings for Asian vegetables were equivalent to 3,763,000
Kenyan pounds, which ranks this commodity group well above the
majority of the items listed as "principal" export commodities in
the Government's Statisitical Abstract. Export earnings from
Asian vegetables exceed those for all. individual categories of
manufacturod goods other than chemicals and cement.

Farmer Participation

While the aggregate qrowth and export earnings of the Asian­

vegetable trade are important, the rubsector's main impact has
been felt through exporter procu:.-ement of these vegetables.
Initially exporters obtained produce from their own farms or from
medium- to large-scale Asian or European farmers. In the late
1960s, in the wake of a series of Kenyan Government measures to
Africanize various aspects of the economy, an exodus of Asian
farmers from the Kibwezi area began. Africans who had worked on
the Asian-owned farms moved on to produce Asian vegetables either
on their own farms, on land leased temporarily to them through a
government irrigation scheme, or on larger African-owned
farms. (25) One Asian farmer whose family had lived in Makindu
for many years began in the mid-1960s supplying both his and
outgrower-farmer produce to Nairobi-based exporters. He
purchased from both small- and large-scale farmers. By 1972 he
withdrew from farming and established his own exporting company
called Makindu Growers and Packers. By providing technical
advice, market access, and (occasionally) production inputs, this
firm stimulated Asian-vegetable production for export from
Kibwezi farmers.

By the late 1970s small-scale farmers were becoming the most

important source of these vegetables. In the early 1980s, small­
scale farmers probably accounted for 75-80 percent of the Asian
vegetables that are exported. The trend in the Pid-1980s has
been a move back in the direction of procurement from larger­
scale production units. The major involvement of smallholders in
the Asian-vegetable sector contrasts, however, with the export

88
procurement systems for crops such as pineapple, passion fruit,
french beans, flowers, and strawberries, where a substantial
majority of produce derives from medium- and large-scale
farms. (26)

Also important is the location of Asian-vegetable

production. The bulk of Asian-vegetable supplies has come from
the semi-arid areas of Machokos District such as hatuu, Kibwezi,
and Mtito Andei, where over 3000 smallholder farmers are attached
to government-supported irrigation schemes. In recent years
Asian-vegetable production has also expanded to distant
Lotokitok, on the slopes of Mt. Kilimanjaro. Asian vegetables
have provided an impo-'tant source of income and employment for
these areas, becoming the most important (and widespread) cash
crop in certain locations.

Income Generation

Available data on farmer yields and sales of Asian

vegetables are extremely poor, and what data do exist show
tremendous yield variations among farmers. Prices paid Asian­
vegetable farmers also show considerable variation, yet even when
using relatively low yield price estimates, Asian vegetables
compare favorably with other cash crops and food crops in terms
of gross producer income. The data below for cash and food crops
are calculations by USAID/Kenya of the average gross income of
crops over the 1979-1983 period.

Average Gross Income Per Production Season (Ksh)

Cash/Food Crops Asian Vegetables

Sugar 18,559 Karela 24,000 (a)

Tea 11,227 Chillies 18,000 (b)
Coff'ce 9418 Okra 16,500 (c)
Pyrethrum Ex. 3736 Aubergine 12,000 (d)
Maize 1584
Oilseeds 1345

(a) Yields vary between 2-6 tons/acre and prices between Ksh
5-a/kg. Used here is a yield of 4 tons/acre at Ksh 6 per kilo.
(b) Yields range from 2-6 tons/acre and prices between Ksh
4-9/kg. Used here is a yield of 3 tons/acre at Ksh 6 per kilo.
Can get more than one crop per year.
(c) Yields range from 2-6 tons/acre and prices between Ksh
4.75-9/kg. Used here is a yield of 3 tons/acre at Ksh 5.5/kg.
Can get up to three crops per year.
(d) Yields range from 5-12 tons/acre and prices between I'sh
1.5-3/kg. Used here is a yield of 6 tons at Ksh 2 per kilo.

89
 This comparison is for illustrative purposes only. The data
for the cash and food crops is now slightly outdated. Even
though I have used relatively poor average yield estimates and
average price estimates toward the bottom of their range, some
smallholders may obtain less positive results. The estimates are
for gross rather than net income; however, one source has
calculated the net income for an acre of thin chillies, okra, and
karela to be Ksh 9000, 12,800, and 7000 respectively. (27) These
levels are higher than the estimated gross income for wa y other
cash and food crops.

Employment

Most Asian vegetables are labor intensive relative to other

crops grown in Kenya. They are grown throughout the year,
although the peak production of most items tak-s place over the
October-June period. The employment opportunities created by
expanding Asian-vegetable production have led many young people
in parts of Machokos District to remain in their home area rather
than migrate to Nairobi or other locations in search of work.
Compare below estimates of labor intensivity for different crops:

Man Days Needed Per One Hectare Crop

Hybrid Maize 152 Aubergine 277

Cotton 235 Okra 304
Coffee 294 Chillies 378
Karela 510

(Sources: Hormann and Thuo [1979]; own calculations)

The Asian-Vegetable Procurement System

While it may be said thet the Asian-vegetabl trade has

made a range of contributions to the Kenyan economy, this does
not imply that the production and marketing system for these
crops has functioned efficiency. To the contrary, the
coordination of production with marketing has been extremely
weak, and the overall system seems to operate in a state of
perpetual disequilibrium. Subsector participants, especially
farmers, operate under considerable uncertainty. In recent years
overall production has far outstripped demand, while on a
seasonal basis the supply of particular items has been
inadequate. Not only has produce wastage been high, but the
produce mix of exnorters has been thrown into an imbalance. This
has undermined the competitive position of Kenya in the U.K.
market.

In this section we discuss the general features of the

Asian-vegetable procurement system. We note the inefficiencies
and uncertainties that the system creates. Together with the

90
overseEIs market conditions and the wider political framework,
this systemic disequilibrium serves as the backdrop to one export
company's attempt o introduce Yormal contractual arrangements
into the procur.ement system. This case is discussed in the
subsequent section.

What we will discvss here is the main features of exporter

procurement of Asian vegetables from smallholder farmers. All
exporters obtain a share of their supplies from medium- to large­
scale growers. This share varies by company. Some corpanies
rely largely on a few larger growers with whom they no"e dealt
for many years. For these firms, smallholders may oniy be a
residual supply source. More commonly, exporters obtain the bulk
of their suppli- from smallholders and rely on larger farmers
primarily for i zms requiring greater investment (i.e., wires for
trellising) or 1. gher technical standards. Those Asian
vegetables that equire high humidity for growth are contracted
to larger farmerk at the coast. Chillies grow well at Lake
Naivasha and as exporters are already procuring french beans from
the large farmers there, this Asian vegetable is added to their
order.

The nature of exporter/largeholder relations differs

significantly froui that of exporter/smallholder relations. The
relationship is generally more personal, more intensive, arid
longer lasting. It sometimes is based on a higher level of trust
and loyalty. Bargaining power is not as skewed as in the
exporter-smallholder case. The relationship is also not as
politicized. Communication flows are better than in the
smallhoider case. For these reasons the exporter/largetolder
links have generally been satisfactory from the perspective o±
both exporters and farmers. Supplies from large farms are
inadequate to meet demand, however. Some large-farm areas are
not environmentally suitable for Asian vegetables. In other
areas where large farms exist and where some Asian vegetables can
be grown, farmers have preferred more familiar crops or crops
yielding iigher revenue per acre (i.e., french beans). Larger
farmers operating with pump irrigation systems have demanded
continued price increases to maintain their plantings. Seasonal
labor shortages have also constrained large-farmer production in
areas such as Kihwezi.

Smallholder producers of Asian vegetables have thus been

sought. Even at lower prices and with lower and varying yields,
smallholders in parts of Machokos District would find growing
Asian vegetables for export an attactive venture. The
procrement system for smallho der Asian-vegetable supplies,
however, has not functioned efficiently. Lr.c us examine this
system.

91
Demand for Smallholder Supplies

The demand for smallholder supplies of Asian vegetables is a

derived demp-.d. It is an aggregation of the requirements of a
large number of individual companies, which themselves derive
from:
a) the level and adjustment of on-going orders by U.K.
importers;
b) the quantity of air cargo space allocated and then
actually provided to the exporter;
c) the relative importuncc of Asian vegetables in an
exporter's produce mix; and
d) the expcrter's supplies of produce from larger farmers.

These are variables, not constants, and thus the quantity of

Asian vegetables required from smallholders shows continuous
variability. When combirned with variable production and poor
information flows, the seeds of disequilibrium are sown.

Typical Procurement Arrangements(28)

In a production area such as Matuu (Machokos District) the

exporter's contact with farmers is thrc.ugh his truck drivers and
a few local agents whom he may appoint to represent him. For
smallholder supplies most exporters work through agents, usually
local shopkeepers or farmers who own or rent a shed in a market
area. These aqents try to recruit farmers to grow for a
particular company. General procedures vary by company. Some
provide agents with cartons on a weekly basis and give orders for
a week, perhaps scattered over three or four days of pick-up. An
agent must distribute cartons and make sure farmers make
deliveries to his stall in time for the collections. Other
companies bring cartons only on the morning of collections and
specify their orders on that day. An agent may have some farmers
operating on accounts while other farmers deliver on a strictly
cash basis. Those with accounts generally receive a steady price
for individual products and may be paid monthly or fortnightly.
Those delivering on a cash basis will face widely fluctuating
prices. Exporters will inform their truck drivers of the daily
prices. The actual prices that "cash f~rmers" receive and the
extent of delay in payment depend upon their relal-irinship with
the agent.

The weak coordination of the trade can be illustratedi by

several features, Including:

I. Absence of Production Support--- Most exporters have had

no direct involvement in the smallholder production process.
They view themselves as trading companies neither capab).e of nor
responsible for Droviding smallholders wi4 h either productzion
inputs or technical advice. These are seen to be the
responsibility of other institutions. While the seeds/chemical

92
trades and various farmer associations are seen to be responsible
for the inputs side, the government extension service and the
HCDA are seen to be responsible for technical assistance to
farmers.

The location of some of the production sites, the specialist

nature of these crops, and the prior notion that these crops are
nr!+ important to Kenyan Africans have resulted in an absence of
production services to smallholder Asian-vegetable farmers. This
vacuum can most clearly be seen in the area of technical advice
to farmers. In the main areas of Asian-vegetable production the
numbers of extention staff have been few and their mobility
limited by inadequate transport means. Trained as generalists
and having only a few of the Asian vegetables described in their
Ministry of Agriculture Handbooks, these extenoion people have
not been in a strong position to make recommendations to farmers.
What they know about Asian vegetables tney have learned from
farmers. One extension worker views his activities as being
equivalent to "running in the fields."

2. Inappropriate Quantities or Produce Mix--- Each

participating exporter is continuously unable to obtain his full
vegetable requirements in the appropriate mix to meet overseas
orders. Each day he obtains a surplus of some items and
insufficient quantities of other items. Being short of certain
items is particularly prcblematic, as it %psets the entire
produce basket. ExporterL react to this situation of uncertain
product mix by a) over-crdcring supplies and then rejecting or
repacking produce, b) over-ordering supplies and keeping excess
items for shipment the next day, c) over-ordering supplies and
then selling excess items in their own retail outlet, or d)
exchanging items held in excess for short iterm held by oLner
exporters at the airport. Only a few exporters have their own
retail outlets and there is practically no demand for these items
by the local processing industry (i.e., processing firms import
chillie pzivder from the Far East) so option "c" is not commonly
pursued. Each of the other options are commo..

Option "a" shAfts quantity risks onto the farmers. Option

"a" can be carried cut in the field or in Nairobi. Exporte-s may
give their tr),ck drivers target qu~Litity figures for diffe6rit
vegetables. Once these targets are reached in the course cf
their collection rounds, the collectors may cease further
purchase of these items, perhaps on the basis of "poor quality."
Another tr'iditional practice has been to match sumplies with
orders at the last-minute documentation stage at the airport,
save some extra supplies, and then return additional surplus on
Rquality" grounds. Farmero report that sometimes they receive
back ca. tons that either axe not theirs or are half empty.

2. high Wastage and Speculative Production--- While a few

exporters do give an indication to farmers (or farmer groups) of

93
their expected requirement. over the course of an export season,
there is no coordinated planning procedure for Asian-vegetable
production. While Asi n-vegetable production is carried out all
year long in the main smallholder producing areas of Matuu and
Kibwezi, there are weather-induced production peaks in December-
February and April-June. During this first period Asian
vegetables must compete with higher value horticultural crops for
the available air cargo space. During the latter period there is
generally a surplus of many items. Most farmers growing Asian
vegetables either must leave a sizeable proportion of their crop
unharvested or face considerable wastage due to the lack of a
sales outlet. Most farmers obtain seeds and then plant
speculatively, hoping that a buyer will be found at harvest time.
Wastage of produce may be 30-50 percent at times. Even when
farmers do have ongoing relations with exporters, the latter
sometimes give short-term notice to stop harvesting particular
items. (29) Farmers located in areas with poor access roads may
have even higher levels of wastage as some exporters simply do
not send their trucks to these areas during periods of heavy
rain.

3. Producer Price Variation--- Producer prices exhibit

wide variability for the same crops in the same places. These
price differences are not generally linked to quality
differences. Rather, they are linked to short-term supply and
demand conditions, the relative desperation of competing
exporters, and price manipulations of the local agents serving
the exporters. Most -xporters pay different farmers different
prices. Sometimes farmers who have accounts with exporters are
paid higher prices, while other times farmers selling on a cash
basis receive a premium. Even when a company has established a
consistent policy, its implemention by staff or local agents may
involve considerable discretion. Company staff collecting
produce and paying cash are sometimes in a position to pay
farmers below the company's stated price. Local agents who may
also be farmers are in a position to underpay less-informed
farmers;

4. Quality Variation--- Produce quality exhibits wide

variation at farm and export level. The industry lacks a
consistent set of quality guidelines for many of the Asian
vegetables. Different exporters sc't different quality standards,
and produce rejected by one firm may be accepted by another. In
addition, quality standards are adjusted by exporters in the
context cf supply and demand conditions. We noted above the
upward shift in "quality standards" when supply exceeds demand.
Qual ty atandards are adjusted downward over the July-September
period when some? 7:rops are in short supply. Quality control is
thus a vehicle .'or quantity control. Not only expuvters behave
opportunistically in relation to produce quality. A common
practice of farmers is to put good quality produce at the top of
a carton and bad produce on the botLom, hoping that the carton

94
will pass through the exporters and government inspectors
undetected. Previously it was the exporter who paid for this
practice through the quality claims made by overseas buyers.
More recently, some exporters have each contributing farmer write
a designated code number on the side of the carton so that the
culprit can be detected and deductions made on future purchas:s.

5. Information Problems--- Small-scale, Asian-vegetable

farmers are poorly informed about the changes in supply, demand,
or the air cargo space situation. With such a large number of
exporters and the uneven buying behavior of some, farmers have
difficulty gauging demand. Communications are very poor between
Nairobi and several growing areas, and information is generally
passed to farmers by company collection-truck personnel. Delays
in communications may result in farmer losses as produce is
harvested without exporter intention to purchase. Exporters tend
to pass on only short-term information regarding the quantity of
requirements. When local agents are responsible for providing
information to farmers, there is scope for distortion. Local
government staff do not understand the general patterns and
complexity of the trade, and are thus not in a position to advise
farmers or a production and sales strategy.

In a contentious trading environment information becomes a

perishable commodity. Information is a key element in reducing
risks. As long as farmers c=an be held in the dark, the risks of
cargo off-loads and supply/demand imbalances can be shifted to
them. Information flows take a "negative" form. Exporteia will
inform farmers when the overseas market is depressed or when the
quality of produce is below some standard set by the exporter.
Positive feedback on good produce or good sales re-sults is rare.

6. Weak Intermediation--- The weak bargaini ig position of

farmers, the poor information flows, and the a?-'.ence of effective
production planning would all appear to call for the involvement
of farmer cooperatives or associations in the Asian-vegetable
system. A large number of such groups have either emerged
ostensibly to help vegetable farmers, or have diversified beyond
interests in coffee or cotton to include vegetable farmers.
Generally, these coooperatives have made only a minimal
contribution to the Asian-vegetable sector. Some of theae groups
are "paper cooperatives" consisting of a list of namec and office
holders. Other groups have "bodies with no legs" lacking support
and legitimacy in the eyes of farmers and being used for
political purposes by exporters rather than carrying out actual
marketing functions. Cooperative officials have been adept at
corresponding with exporters and government officials, laying out
terms of trading agreements or asserting the rights of farmers,
but vegetable cooperatives have been singularly unsuccessful in
coordinating the oroduction and marketing of the farmers on their
lists.

95
 While not averse to the idea of coope-ation, many farmers
have come to associate formal cooperative organizations with the
deduction of cesses from farmers in order to pay for the offices,
telephones, and trips to Nairobi for a few "big men." Where
horticultural cooperatives have operated, internal power
struggles have frequently led to the breakaway of splinter groups
with both exporters and government officials not being clear
about whom to deal with.

7. Widespread Mistrust--- Exporters perceive most farmers

as opportunists selling to whoever provides them with the best
terms at arjy one time. Farmers view exporters as unscrupulous
and unreliable. Commitments are made to tie down the other party
and reduce one's own risks. Under a range of circumstances the
commitment will be readily broken. Cooperative officials
mistrust exporters and farmers while the latter two mistrust the
cooperative officials. Farmers view the HCDA as supporters of
exporters while the exporters view the Authority's intentions
with suspicion and its direct participation in the trade with
alarm.

A Contractual Scheme for Asian Vegetables

Within the context of this rather chaotic trading network an

effort was made between 1982 and 1985 to organize
exporter/smallbolder relations on a contractual basis. The
scheme involved approximately 500 smallholders in the Matuu area
linked by contract and farmer groups to the company, Kenya
Horticultural Exporters Ltd.

Background of Matuu-Yatta (Machokos District)

During the period 1954-59 the 37 mile long Yatta Furrow was
constructed by a work force of Mau Mau detainees. The furrow was
fed by the Th!.ka River and was initially geared toward supplying
water for domestic use and for cattle. Not until the mid-1960s,
with the initiation of settlement schemes, was water from the
furrou used for irrigation purposes. Throughout the late 1960s
and early 1970s small groups of people were settled on nne- to
three-acre plots near the furrow with feeder channels providing
irrigation water. The first plantings on these plots was in the
spring of 1967. (30)

From the beginninig the Matau farmers plante:'d vegetables on

the irrigated parts of their land, and maize and cowpeas on rain­
fed sections. Ivailability of water permitted the farmerc to
produce tomatoes, cabbages, and chillies at times wrtjn supplies
were short from the rain-fed areas in Central Province and in
olher parts of Machokos District. During these times Nairobi
traders would travel over nontarmac roads to reash the scheme.
At other times of the year Matuu farmers were heEi.ly constrained
by transport, as bus links to Nairobi or Thika were weak and

96
preference was given to passengers over produce. Insufficient
coordination of farmers restricted the hiring of lorries to
transport produce to Nairobi. (31)

In the mid-1970s the Horticultural Crops Development

Authority attempted to assist Matuu vegetable growers by
establishing a few grading, packing, and collection centers and
linking local farmers to the nation-wide Horticultural
Cooperative Union, to food processors, and to exporters. Various
companies made inquiries through the Ministry of Agriculture as
to whether the Natuu farmers could increase their production for
export. In 1977 Schluter and Co. requested birdeye chillies for
local processing. A year later M/S Kenez came forth with a
request for 30 tons of Asian vegetables per week to expc-t. In
1979 Al-Khaldiya Trading company inquired about supplies of fruit
to export to Saudi Arabia.

As with the efforts of the HCDA, these firms needed to

establish a link with a local organization. The only existing
farmer's organization was the Masinga Farmers Cooperative Union,
which was handling cotton. The HCDA stations were turned over to
the Union to administer, and exporter requests were passed on to
the cooperative. Few farmers felt that the Union represented
their interests, however, after it had generally mismanaged their
cotton crop and delayed payments for their vegetables. The HCDA
packing stations were closed and the produce inquiries were not
followed up. (32)

Still, by the late 1970s a few expnrters of Asian vegetables

had become aware that good quality vegetables could be obtained
from Matuu. They thus employed somc' local farmers to act as
their agents, buying from other farmers and then meeting the
exporter's trucks in Thika. I+ was not until the 1980/81 season
when the Thika-:(itui road was tarmacadamed that exporter trucks
actually went to the fatuu area. Only two exporters were
purchasing on a sustained basis in Matuu. Neither firm was
directly involved in supporting production. A few other
exporters made purchases ui- an ad hoc basis.

Farmers growing Asian vegetables for export were not

satisfied with the prevailing marketing arrangements.
Fluctuating prices, uncertain purchases, unreliable payments, and
quality adjustments were seen as common, and Yarmers had no
bargaining power vis-a-vis exporters. A group of farmers
contacted the director of the Horticultural Crops Branch of the
Ministry of Agriculture asking for his ar sistance. This director
himself had a fdrm in Matuu. He encouraged the farmers to form
an association or "self-help" group and put them in touch with an
exporter who might consider a more formal marketing link with the
Matuu farmers. This firm was Kenyr Horticultural Exporters Ltd.

97
Kenya Horticultural Exporters Ltd. (KHE)(33)

KHE is a partnership of two families, both with origins in

Gujarat, India. The families entEred into business together in
the mid-1950s to form a fresh-produce retail outlet. The firm
imported fresh fruit and expanded into local wholesaling,
especially for potatoes, onions, and garlic. In the mid-1960s
with the involverent of several European farmers at Naivasha,
they initiated an export trade. At that time the only other
important exporter was the Horticultural Cooperative Union,
although there were a few small-scale competitors.

The company's early exports consisted primarily of french

beans, pineapples, and strawberries obtained largely from
European farmers and sent on consignment to a broker in London's
Covent Garden Market. in the late l960s the company began
exporting Asian vegetables to two Indian firms based in London.
Asian vegetables were obtained from a few European and Asian
farmers.

During the 1970s KHE emerged as the leading exporter,

expanding its volume of trade and significantly diversifying its
product mix anu Aarket outlets. It was tv.e first company to
enter the West --,-man market and played an important role in the
opening of the market for Kenyan french beans in France and
Belgium. The company handled a quarter to a third of Kenya's
fruit and vegetable exports over the decade. Asian vegetable
exports to the U.K. remainad important, accounting for 30-40
percent of the company's export volume. In 1973 one cf the
company's founders emigrated to the U.K. and shortly thereafter
esta)?lished his own fruit and vegetable import and dismributing
compaly. This U.K. affiliate played a major role in expanding
the distribution of Asian vegetables outside of the Greater
Lo.don area.

The company continued to obtain its supplies from medium- to

large-scale farms in areas such as Naivasha, Thika, Embakasi,
Kibwezi, and the coast. It was developing a reputation for
2eliability in its dealings with farmers. For this reliability
farmers needed to pay a risk premium--KHE's producer prices were
generally 10-2C percent below those of its competitors. Having
developeo excellent relations with several airlines, having
strong overseas marketing links, and purchasing in sizeable
volumes, KHE was able to exercise considerable bargaining power
in local price negotiations. "Loyal" farmers couid obtain inputs
and credit from the company. If unforeseen market downturns
occurred these "loyalists" would be compensated for part of their
production costs. Thp company was the first to provide written
contracts to farmers growing vegetables for export. This was
undertaken with several farmers growing french beans and sweet
pepper.

98
 KHE has continued to expand its trade in the 1980s. It is
one of only a few Kenyan firms that have maintained a reputation
in Europe for quality pruduce and reliable service. At any one
time the company is exporting to up to a dozen countries and can
send 50 or more different items. While Asian vegetables and
french beans have continued to comprise a major part of the
company's export volume, the company has been Kenya's leading
exporter of nvocado, mango, passion fruit, and more exotic
produce such as apple ba-ianas. In recent years KHE's exports of
fruit and vegetables have reached the following levels:

1982 4315 tons

1983 5170
1964 5881
1985 5423

Over this period, the company has accounted for between 21

and 25 percent of the total volume of Kenyan fruit and vegetable
exports.

When KHE was approached in 1982 by the Ministry of

Agriculture official on behalf of the fatuu farmers, the company
was in a confident mood. By that time it had succeeded in
developing strong marketing links to a number of countrnes. Its
U.K. affiliate was diversifying its product range and was
becoming actively involved in marketing channels supplying
multiple chain supermarkets. KHE was in the process of hiring an
experienced horticulturalist who had managed the farm of one of
the company's main suppliers. It had just moved into a new Ksh
24 million complex incorporating offices and packing, grading,
and cold storage facilities. The company's operations were
previously scattered among three Nairobi sites. The cold storage
facilities would not only help deliver a higher quality product
with a longer shelf life, but would enable the company far
greater fleximility in its procurement arrangements. The cold
storage facility would enable the firm to carry out more
effective grading and quality control and to accommodate
surpluses of produce.

Thus, KHE in 1982 was in a confident mood looking to expand.

In terma of Asian vegetables the company had been experiencing
procurement problems as its policy of low k-ut steady prices was
making the firm uncompetitive with other exporters whenever
supplies of particular items were short. The other exporters
merely increased thei- prices ano made cash purchases. In
addition, the company was finding that some of its Iraditional
suppliers were not able to grow okra and chillies in sufficient
quantities and at hich quality. The =ompany's Asian-vegetable
export mix was thus out of balance and was constraining the
marketing effort of its U.K. affiliate.

99
 The Matuu situation appeared to provide the company with a
tremendous opportunity. The farmers there were looking for a
reliable buyer. Several Asian vegetables as well as other items
could possibly grow well there under irrigation. The company had
never formally contracted smallholder farmers before, but a
contractual framework was viewed as the best way t( signal the
company's long-term intentions both to the farmers of the area
and to government officials aware of the marketing problems faced
by the Matuu farmers. The company hoped that if indeed Matuu
became a major new source of export produce, then its contractual
links would enable it to have prime access to the additional
supplies.

The Scheme

Over three seasons--1982/83, 1983/84, and 1984/85--KHE

operated a contracting scheme for Asian vegetables and selected
ether items in the Matuu area. At the he ght of the scheme more
than 300 farmers were selling produce to KHE, and this enabled
the company considerably to expand its exports of Asian
vegetables. In the beginning of 1985 the project virtually
collapsed in the face of the drought-induced shortage of produce
and severe competition from other exporters for the farmers'
output. Since then the company's presence in the area has
diminished greatly, and during the 1986/87 export season no more
than 30 Matuu farmers sold to the company. Still, Asian­
vegetable production has continued to expand in Matuu. It is
KHE's competitors who are picking the fruits of this expansion.

1982-83

In June of 1982 a contract was worked out between KHE and a

committee representing the Matuu farmers. The program laid out
was extremely ambitious, reflecting the newly strengthened
confidence of the company. Matuu farmers would grow for KHE not
only several Asian vegetables they were familiar with, but also
substantial quantities of french beans and smaller quantities of
melons and even gooseberries. The company intended to enter with
a "blanket," spreading seeds, chemicals, and advice, and
generating e major new supply source of export produce. There
would be no trial period. Inputs would be distributed and KHE
purchases would begin in October. The program specified KHE's
weekly requirements over a period from October 1 to May 31 as
well as guaranteed prices that would hold over the entire period.
Production outside of this period would be at the farmers' risk
and would be purchased at negotiated prices.

KHE would not deal directly with each of the individua3

smallholders. The company had not previously operated in the
Matuu area and had no past contact with any of the large- or
small-scale farmers in the area. As it wanted to develop a
project on a fairly wide scale it required local intermediaries.

100
The Matuu Horticultural Marketing and Suppliers Committee,
comprising some of the area's larger, more influential farmers,
was seen as an appropriate intermediary. While initially the
Committee was supposed to play the role of communicator,
negotiator, and edvisor for the farmers, the intention was that
the Committee would seek small farmer members and register as a
formal cooperative.

Farmers preferred that the Committee remain purely a

communicative and advisory body with no decision-making
authority. They resisted the Committee's efforts to raise
contributions from them to set up an office and cover the petty
expenses of the Committee. The farmers preferred that KHE deal
either directly with them as individuals or through a number of
collection stations. Having individual accounts with several
hundred smallholder farmers was viewed by KHE as both expensive
and administratively infeasible. KHE's horticulturalist and an
agricultural officer in the area established eight collection
centers in the area. Twenty to twenty-five farmers were assigned
to each center, and they elected a center manager. KHE would
hold separate accounts for each collection center and provide
inputs and payments through their managers.

The KHE horticulturalist instructed each center on what

crops and what acreage to plant and provided the inputs to the
centers. At each center he initiated a small nursury to
faciliLate the transplanting of seedlings. He provided some
instruction to center managers and individual farmers on
production techniques and grading. Other company staff worked
part-time on the project, especially in monitoring farmer grading
and packing. The company had insufficient manpower, however, to
provide more than a minimalist extension service.

In drawing up the contract, each side acknowledged the

prevalence in the trade of sudden quantity adjustments on the
part of exporters. Thus, a clause was written into the agreement
that "the KHE will undertake to collect all exportable produce at
the given collection time. In the event of unavoidable
circumstances, the KHE will negotiate with the committee and put
in writing a suitable value of compensation for any uncollected
produce." This clause would theoretically lower the impact of
the major marketing risk facing Matuu farmers--i.e., lack of a
market outlet for their crops.

Distribution of inputs began in June 1982, initially on a

small scale. For several months the company provided a total of
about 20 kilos of seed/month. Nurseries were started at each of
the collection centers and on some of the larger farms. Among
the Asian vegetables, the company wanted to have Matuu farmers
racentrate on only a few items that were upsetting the export
boket because of their short supply. Particular attention was
given to okra, thin chillies, and fresno chillies. Matuu farmers

101
were also keen on growing aubergine as they knew it grew well in
the area and was far less labor-intensive than some of the other
crops. By October, input distribution was at full steam with
okra seed alone being supplied at the rate of 60 kgs. per month,
enough for 20 acres of planting. Most farmers were planting 1/4
to I acre of Asian vegetables. Several larger farmers, who had
individual accounts with KHE, planted up to five acres of Asian
vegetables.

Matuu experienced adequate rainfall over the 1982-83 season

to produce a good crop. Over the October 1982-September 1983
period, KHE purchased 575 tons of Asian vegetables from the Matuu
area. This represented more than 30 percent of the company's
exports of thiE group of vegetables for that year. KHE's
purchases in the area had a value of Ksh 2.58 million. Four
items accounted for 84 percent of KHE's Asian-vegetable purchases
in Matuu. These items were okra (195 tons), aubergine (133
tons), thin chillies (99 tons), and fresno chillies (59 tons).
The Matuu farmers had prior experience with thin chillies, so the
good results for this crop were not rprising. Fresno chillies
were introduced by KHE and brought id harvests from November to
May. The results for okra were di,. jointing, although supplies
from MLatuu did help KHE improve okra's position in its overall
export basket. The 60 kilos of okra seed per month that KHE
provided from October to June should have generated 40 tons of
produce per month, even with a poor yield of 2 tons an acre.
Actual okra purchases were the following (in tons):

Oct 1.6 Feb 18.0 Jun 50.9

Nov 6.6 Mar 18.2 Jul 11.8
Dec 11.4 Apr 26.6 Aug, 11.8
Jan 17.9 May 23.3 Sept 3.3

Only in one month, June, did purchases come anywhere close

to expected levels. Okra supplies in June were actually in
excess of KHE's needs, and it brought that product into surplus
at the time when Cypriot okra was coming onto the U.K. market.
The subsequent collapse of supplies over the July-September
period was weather induced with chilly evenings ..estricting okra
growth. In the course of the eeason, competing exporters had
made cash purchases of some of the produce grown under the KHE
contract. Okra was one product where such "leakageO was
imrortant. When these other exporters stopped purchasing okra in
June, the entire crop was left for KHE. The inadequate supplies
at other months cannot be accounted for by leakages alone. Many
okra fields were hit by disease, and yields were very low.

Aubergine also proved to be a problematic crop for the

season because of extremely uneven deliveries. Farmers utilized
the KHE contract as a sort of safety net, planting speculatively
outside of the contract, looking for alternative buyers at higher
cash prices, but then falling back on the KHE commitment when

102
market circumstances necessitated. KHE specified in the contract
that its requirements were 12 tons/month. Actual KHE aubergine
purchases were the following (in tons):

Oct 1.3 Feb 19.8 Jun 10.4

Nov 5.2 Mar 21.9 Jul 6.6
Dec 9.4 Apr 25.1 Aug 4.7
Jan. 9.7 May 9.8 Sept 8.7

Tne figures show that during the main October-May season the
company's requirements were not met in five of the eight months,
but that in the three other months deliveries were approximately
double the company's expected requirements. A surplus of
aubergine had emerged by mid-February and the farmers needed the
KHE outlet. The company was not sure whether excess supnlies
were due to better than expected yields or entirely to
overplanting, and so continued to buy the produce on offer. By
late April the company received a telex from its U.K. buyer
noting that the aubergine market was depressed, that KHE was
sending too large a volume, and that there were st;.re quality
problems. The company immediately stopped its purchases of
aubergine from the Matuu farmers. It informed the Head of the
Horticultural Branch of the Ministry of Agriculture that this
step was being taken because of the quality problem. The Matuu
Committee argued that KHE graders were inspecting the produce and
passing it for loading into the collection trucks as before.
While acknowledging that heavy rains had affected some of the
crop, the Committee argued that some of the crop was still good
and that KHE needed to abide by the clause to take "all
exportable produce" or else provide due compensation. The
diapute ended several weeks later with KHE undertaking limited
purchases. No compensation was provided to farmers as the
company showed that it was making purchases in excess of the
contract.

An effort to have the Matuu farmers grow bobby beans during

the 1982-83 season proved to be a disaster. The effort was
concentrated on some of the larger farms in the area, rather than
the settlement farmers. The beans encountered severe disease
problems. Nearly two tons of seed were lost.

1983-84

The 1983-84 season was highly successful for the project.

New collection stations were started and additional farmers
sought individual accounts vith KHE. At its peak perhaps 500
farmers were linked into the KHE system. KHE increased the level
of input supply and expanded the range of Asian vegetables that
it purchased from Matuu. Several nurseries wer& operating
effectively and helped provide higher quality aubergine and thin
chillies. Over the period from October 1983 to September 1984,

103
KHE purchased nearly 839 tons of vegetables from Matuu at a value
of nearly Ksh 4 million. These purchases accounted for about 45
percent of KHE's Asian-vegetable exports that year.

It is possible that a similar volume of purchases was made

in the area by competing exporters buying not only from the
farmers ostensibly growing under the KHE contract, but additional
farmers who were encouraged by the income obtained by the
contract farmers. While the other exporters were riot providing
inputs, the farmers were obtaining seeds outside of the KHE
conty-act from shops in Nairobi. The contractual scheme was thus
generating a general production expansion in the area.

KHE's exports of Asian vegetables expanded over the year as

its basket was more closely coordinated with the requirements of
its U.K. affiliate. Additional supplies of good quality okr3 and
chillies were sent to a buyer in France. ThE bulk of KHE's
requirements for several relatively minor items was obtained from
Matuu.

Still, the year was not without problems. While less

dramatically than during the first year, rupplies continued to be
uneven and deliveries rarely reflected the requirements set out
in the 1983-84 contract. Aubergine supplies continued acting
lie a rollar coaster, sometimes below orders and sometimes
considerably above. The company's monthly order (for October to
May) for Asian vegetables was about 93 tons. Actual purchases
averaged 85 tons; but two months featured purchases of less than
70 tons, and two months had purchases of over 108 tons. The
company provided large quantities of chola seed hoping to
increase production of this item. Chola is a type of pigeon pea
that the local farmers like to eat. The company was not getting
the deliveries of the crop that it had expected and discovered
that farmers were eating che leaves of the plant or selling the
crop locally.

Some pronlems were encountered with collection center

managers not paying farmers. As there were no banks in the Matuu
area, KHE would write a check in the name of the manager who
would then be responsible for distributing the money to
individual farmers as per the receipts they were given at produce
delivery. Several center managers were dishonest, and farmers
began losing confidence in the collection center system. Some
centers closed with a few farmers obtaining individual accounts
with KHE while other farmers decided to sell to other exporters.

1984-85

The KHE-Matzau contracting scheme complgtely unravelled

during the 1984-85 season. The short rains uf March-April 1984
were lower than normal and the long rains of September-October
1984 completely failed. Drought conditions had set in in many

104
parts of the country, adversely affecting agricultural
production. Matuu farmers were still able to draw on the
irrigation water of the Yatta furrow. The production of Asian
vegetables continued to expand up until about February of 1985.
KHE's purchases were at levels similar to those of the previous
year, but farmers were restricted from irrigating during the day
due to the shortage of water and the threat of water supply to
Kitui town.

Reduced Asian-vegetable output in Matuu and shortages of

supplies from other areas resulted in a chaotic scramble for
supplies over the March to June 1985 period. Many exporters were
attempting to obtain produce in Matuu and were offering prices
well above those offered on the KHE contract. Compare below the
prices offered by KHE with the prices reached in the cash spot
market:

Prices per Carton (Ksh)

KHE Spot Market

Okra (6 kg) 25.5 70

Chillies (5 kg) 20.0 50-55
Aubergine (6 kg) 13.5 25-30
Karela (6 kg) 30.0 70-80

Thus spot market prices reached levels mare than double those
offered by KHE. KHE did not react to the situation fast enough.
It initially maintained a policy of not entering into a cash
price war, hoping that it had generated through its efforts
sufficient loyalty from its farmers in Matuu. This view proved
to be naive. Farmers were being swamped with attention by other
exporters. An attitude spread that there was tremendous demand
for Asian vegetables, that exporters could earn profits even if
paying double the KHE price, and that KHE had actually been
cheating them for a long period. KHE contacted the hatuu
Committee and asked for their assistance in preventing farmers
from selling outside the contract. The Committee responded that
the problems were the company's fault since it had been
"exploiting" farmers. KHE finally did react to the situation and
sent out circulars announcing increased short-term prices. OChher
export2rs merely adjusted their prices upwards to compensate.
KHE suffered a costly loss by not recovering a large number of
cartons that it had distributed during the season.

1985-86

KHE was not ready to abandon its efforts in Matuu. In May

of 1985 it made proposals to the Matuu Committee for the
following production season. It was agreed that all farmers
growing for XHE must formally register with the Committee and
would not "be allowed to sell any of his/her produce to any other

105
buyer" or "be liable to paying damages to both the Group and
KHE." They also agreed that "all farmers for KHE will only plant
according to the programme as provided... (and) no member of the
Group will be allowed to plant outside that programme."

The agreement was actually a last ditch illusion to save the

project. Neither members of the Committee nor most farmers
perceived that they had an interest in abiding by the terms.
Other exporters were now more active in the area, setting up
collection stations of their own. The KHE contract would truly
be a safety net to fall into when higher price offers were not
available. The KHE contract for the 1985-86 season called for 80
tons of vegetables per month. During October and November actual
purchases averaged 11.3 tons. The project had indeed collapsed
to competition. Only a small number of farmers continued
supplying KHE on a continuous basis, and KHE supplies of inputs
and technical advice virtually stopped.

Project Impact

Matuu is presently the leading source of Asian vegetables

for export with purchases of nearly 100 tons/week being made at
the height of the export season. Up to 2000 local farmers may be
involved in this activity, with up to a dozen exporters
purcnasing on a consistent or periodic basis. The most important
impact of the KHE project and its wider stimulation of Asian­
vegetable production has been its injection of increased income
and employment opportunities into a relatively deprived area.

In each of the past few years the Mtuu farmers have

probably supplied in the area of 4000 tons of Asian vegetableF
per year. Such a level of sales has a farm-gate value of between
Ksh 20 and 25 million. Over the course of its three year
project, KHE alone made purchases valued at over Ksh 10 million.
Over this same 1982-85 period the Njoro Canners project in Vihiga
made payments to farmers totalling about Ksh 11.7 million, but
while the payments of Njoro Canners were spread across some
15,000 farmers, KHE's payments went to little more than 500
farmers. Sizable income increases have enabled many farmers to
start small businesses, build permanent structures on their
farms, and pay school fees. The impact of Asian vegetables can
be most clearly seen in the development of Matuu town. In 1S-79
the town was a small site with only two shops. The town has
grown ot a phenomonal rate and now includes numerous streets
filled with shcps and various service businesses and cottage
industries.

Asian vegetable production has also greatly affected the

value of land in the Matuu area. In one settlement scheme area
the cost of leasing land has risen from 400 sh/acre in 1983 to
2500 sh/acre in 1986. As for purchasing land, the cost in one
area has risen from 1000-2000 sh/acre in 1977 to 6000 sh/acre in

106
1986. In another area land values have risen from 3000 sh/acre
in 1982 to 10,000 sh/acre in 1986. The costs of part-time
agricultural labor have also been affected. Wages for
agricultural labor have risen from 5 sh/day in the early 1980s to
10-12 sh/day in 1986.

The project also had an impact on KHE. Through its

operations in Matuu, KHE was able to build up its level of Asian­
vegetable exports over the 1982-85 period. Its U.K. affiliate
was able to strengthen its competitive position in this product
area. KHE was also able to send high quality okra and chillies
to France. The impact of the project on KHE's Asian-vegetable
trade can be seen in the following figures:

KHE Asian-Vegetable Exports

1980/81 1220 tons

1981/82 1350
1982/83 1830
1983/84 1850
1984/85 1750
1985/86 1085

(Source: Own approximations using disaggregated KHE export data

according to customer)

One can see from the figures that the collapse of the
project in the beginning of 1985 adversely affected the company's
overall exports of Asian vegetables. While Asian vegetables
comprised over 40 percent of the company's export volume during
1982/83, they compix.sed less than a 20 percent share during
1985/86. Despite the initial success of the project for KHE,
most of the lessons that the company has learned from its
experience have been negative. In the aftermath of the project,
the company has sought to reduce the risks and transaction costs
involved in Asian-vegetable procurement by concentrating on large
farmer supplies. (see below)

Alternative Non-Market Solutions?

With the collapse of the KHE project in Matuu, the

procurement system for Asian vegetables has largely returned to
its status quo ante disequilibrium situation. Smallholder Asian­
vegetable farmers are faced with a situation of a) weak
bargaining power vis-a-vis exporters, b) uncertainty over prices
and the proportion of their harvest that will be purchased, c)
poor access to information on demand and transport, d) difficult
access to production inputs, and e) poor access to useful
technical advice.

107
 Difficulties in obtaining reliable and high quality supplies
of Asian vegetables from smallholders is leading some firms to
consider alternative sources. KHE has decided to concentrate its
Asian-vegetable procurement on larger farmers. During the 1986­
87 season less than fifty farmers throughout the country supply
KHE with Asian vegetables on a regular basis. Four farms supply
60 percent of the company's requirements of thin chillies.
Supplies of fresno chillies come from only six farmers. For
aubergine, three farms now supply the bulk of the company's
supplies with one farm alone providing 50 percent of
requirements. Karela supplies are coming largely from one farmer
who manages a series of farms at the coast. If the company can
interest a few large Kibwezi farmers in growing exclusively for
it, then it may withdraw from Matuu altogether.

Simultaneous with KHE's attempt to recruit a few large

farmers, the company has Legun a process of backward integration
via the development of a few farms owned by senior partners in
the company. Investments in drip irrigation systems are being
made on two farms. %lready this year nearly a quarter of the
company's thin chillies requirements will be produced on one of
these company farms.

Kenya's second largest fruit and vegetablo exporter, Makindu

Growers and Packers, has also begun to explore non-market
solutions to the problems of the Asian-vegetable trade. This
firm was mentioned previously. It had actually begun in farming,
moving later into strictly export marketing. The firm has relied
upon a mix of small- and medium-scale farmers in Matuu, Kibwezi,
and Lotokitok for its supplies and has sold to a large number of
different importers in the U.K. In 1985 one of the company's
senior partners emigrated to the U.K. where he set up an import
company. That firm handles distribution of Makindu's products in
Londun. The uncertainties of Asian-vegetable procurement as well
as an interest in diversifying into other product lines has led
Makindu to begin development of its own farm also.

These patterns of increased vertical integration by two

firms, which perhaps have the best reputation in the Asian­
vegetable trade, are probably beneficial to the maintenance of
Kenya's competitive position in this trade. The present
fragmentat on of the trade is undermining its long-term
viability, but backward integration by exporters into production
reduces the slope for smallholdex participation in the sector.
The rationalization of smallholder Asian-veqetable production
does appear necessary. Such a rationalization process should
require not only a reduction in the planting oL some items, but
an improvement in the yields and quality of the planted crop.
The fragmentation of the sector virtually assures that output
reduction will be achieved only through gradual smallholder
disillusionment with ai uncertain and unstable marketing system.
Neither the private sector nor the official agricultural

108
establishment is willing or presently nble to bring about the
necessary yield and quality improvements.

The instability and inefficiency of the smallholder Asian­

vegetable component has recently attracted government interest
with a wide range of possible interventions muted. A 1984
Ministry of Agriculture study on the problems at Matuu made the
inevitable recommendations that HCDA be streng'hened and that
more extension officers be assigned to the area and provided with
more technical information about Asian vegetables. Also
recommended was that the Matuu Committee should register as an
official cooperative, that all farmers should register with zhat
cooperative, and that all exporters should sign binding
agreements with the cooperative. (34) Neither farmers nor
exporters Ysve shown much enthusiasm for this arrangement and the
idea remaii.j floating

During 1985 and 1986 both the Ministry ol Agriculture and

the HCDA have made various problem-solving suggestions and
proposals for government interventions. Each proposal has sought
to introduce controls over one or morE dimensions in the trade.
For example, one report issued by the Ministry called for the
introduction of production quotas for farmers. How such a quota
system would be devised, let alone enforced, was not
discussed. (35)

HCDA has toyed with a package of policies for implementation

in the Asian vegetable sector. Most of its proposals, however,
have been targeted on the symptoms of the sector's
inefficiencies, rather than the actual causes of these
inefficiencies. Little discussion has related to reducing the
fragmentation in the export trade, countering the unrzontrolled
growth and variable quality of production, or improving the
provision of technical advice and inputs. The air freight
constraint continues. Most proposals have been control-oriented.
These policies: have been brought up at various meetings between
the HCDA, exporters, and farmers, and have generally sailed
through as resolutions even though only a minority of
participants view them as enforceable (or even desirable).

One issue generally discussed at these meetings is the

unscrupulous behavior of "middlemen" acting on behalf of the
exporters. Typically, a resolution will be passed stating that
there will no longer be middlemen between farmers and exporters.
In practice this is impossible as exporters cannot deal directly
with each individual smallholder (who may deliver one or a few
cartons of produce per day) and even where exporters have set up
collection stations, the managers of these stations inevitably
take on the characteristics of the dreaded middleman who is able
to take advantage of less informed farmers. Most of the
"middlemen" are local farmers, not some elusive character lurking
in the shadows of night. Without such middlemen, most existing

109
exporters would be hard put to obtain produce from smallholders
on any consistent pattern whatsoever. What weak information
flows that do exist between farmers and exporters are largely via
the presence of the "middlemen."

A second resolution frequently passed is that each farmer

will register with one and only one exporter and each party
should sign a written agreement stating terms of exchange. A
copy of this agreement should be sent to HCDA. Thus, in the
absence of trust, contracts are seen to be an appropriate means
of improving production-marketing coordination. Neither most
exporters nor many farmers wish to enter into enforceable
contracts. Voluntarily drafting such contracts would typically
be done in an attempt to "lock-in" the opposite party to future
transactions. As both parties are aware that each is likely to
default at one time or another, the contract merely represents an
illusJon of commitment. If exporters were told that they must
commit themselves in writing to purchasing specified quantities
of produce, then they would simply specify quantities well below
their actual requirements and then obtain the balance
"unofficially" from noncontracted farmers. A contract-farming
system cannot be imposed by government in circumstances where
there is a surplus of (uncontrolled) production and where there
is a multiplicity of buyers.

The variability of prices among exporters and over time is

another issue raised in neetings between the HCDA and farmers.
The proposed "solution" is a controlled producer-price system
with prices worked out between the HCDA and exporters and then
communicated to farmers. Official producer prices would probably
be followed initially, buL the structures of production and
export marketing would soon result in the reintroduction of
variations. Otherwise, farmers with top quality produce and
consistent supplies would obtain the same prices as farmers
producing mixed-quality produce on a sporadic basis. Official
producer prices would probably not be flexible enough to ennble
adjustments to short-term supply and demand changes. The
reduction of price uncertainties would probably lead to
accentuated seasonal gluts and farmers would welcome access to
buyers at below the official price.

Concluding Remarks

This report examined several features of a complex

production and marketing system. The analysis began by examining
the demand and distribution of Asian vegetables in the U.K. and
traced back the marketing channels through to the production
stage in Kenya. Particular emphasis was given to the structure
and constraints of the export trade and the poor level of
coordination between production and export marketing. The report
went on to analyze a contract farming scheme implemented by one
of Kenya's leading horticultural exporters in the early 1980s.

110
While the project did contribute to a major expansion in Asian­
vegetable production among smallholders, market forces made
contractual enforcement impossible and the contracting company
progressively lost control over the crop.

Competitive forces abroad and the changing business

strategies of several exporters appear to dictate a
rationalizatlon of smallholder Asian-vegetable production.
Farmers are not in a position to guide this process collectively
and are thu& vulnerable to boLh the vagaries of the market and
the uncertain effects of piece-meal government interventions.
The government has concentrated its attention on monitoring
exporter behavior and has not laid down the institutional
machinery to support farmers. The export trade is fragmented and
largely unprofessional. Greater coordination between production
and marketing appears elusive, and the Kenyan export trade in
Asian vegetables will decline.

Notes

I. Miller (1971) pp. 3 96-98.; Interview with Mr. Omii Bij of

Makindu Growers and Packers, October 10, 1986.

2. Tandon and Raphael 4

(1984), p. .; Robinson (1986), p.40.

3. This information was provided by several London-based Asian­

vegetable importers interviewed December 1984 to February 1985
and November 1986.

4. OPCS, 1982 Population Trends.

5. OPCS, Birth Statistics 1983.

6. The Immigrant Statistics Unit (1979) as reported in Robinson,

p. 36.

7. Central Statistical Office (1985); Anwar (1979).

8. Aldrich et al. (1984); Robinson (1986).

9. N.O.P. Market Research Ltd. (1974); Hunt (1975); Key Note

(1986).

10. As reported in Wilson (1977).

II. Key Note (1986), p.9.

12. Jones (1978), as mentioned in Robinson (1986), p.29.

13. Jones (1983).

III
14. For a more detailed examination of these features see Jaffee
(1986a).

15. Aldrich et al. (1984), p.199.

16. Ward (1983).

17. This is the general argument put forth by Aldrich et al.

(1981; 1984).

18. Loughborough (1984); personal communications.

19. Based on interviews with importers.

20. See Jaffee (1986b) Yor a more detailed discussion.

21. The difficulties of accomplishing this "transfer" and

suggestions of potentially more efficient methods of increasing
African participation in export marketing are discussed in Jaffee
(1986b).

22. A British team advising the Kenyan Government in the late

1960s predicted that the trade in Asian vegetables would level
off at around the volume reached in 1969 and would subsequently
decline. The prediction was based on the assumption that there
would be no major population increase within the U.K. Asian
community and that consumption patterns among this community
would shift away from traditional foods.

23. Several short periods of rapid growth or decline can be

linked to institutional changes. Particularly significant trade
growth took place over two nubperiods: 1972-74 and 1981-83. It
was during the first subperiod when a partner in Kenya's biggest
export company emigrated to the U.K. and started an
import/distributing company. During the second growth subperiod,
a U.K. firm dealing primarily in Kenyan produce embarked on a
major expansion program via investments in storage and transport
and making deliveries to several cities. The 1984 downturn in
the trade may be partly accounted for by the bankruptcy of this
latter firm, the resulting increased fragmentation of the trade
in the U.K., and the financial losses borne by Kenyan exporters
dealing with this firm.

24. It is likely that Kenyan export data for aubergine are

inaccurate. In recent years there has been a considerable
decline in U.K. importer interest in the Kenyan aubergine with
the greater availability of European aubergine supplies. Such a
decline in demand is not reflected in the trade data. One
explanation may be that some exporters are falsely declaring
other produce as "aubergine" since aubergine have a lower f.o.b.
value than other vegetables, and making such declarations would

112
reduce the foreign currency values that would have to be
repatriated to the Kenyan Central Bank.

25. Thuo and Horrman (1979), p. 8 .

26. See the sections on the structure of the horticultural trade

in Jaffee (1986b).

27. Crop budgets made by Kenya Horticultural Exporters Ltd.,

1986.

28. Based on interviews with exporters, government extension

workers, cooperative officials, local agents, and farmers in
September 1985 and October 1986.

29. One notice seen in October 1986 read "Kindly stop the
harvesting of aubergine. Sorry for any inconvenience."

30. Ministry of Agriculture (1974), p.14.

31. Ibid., op cit.

32. Wekundah (1985), p.2; Farmer interviews.

33. Based primarily on interviews held with Atul Dhanani and

other senior staff of KHE.

34. Wekundah (1984).

35. Machokos District 1985 Horticultural Annual Report.

Bibliography

Aldrich, Howard, John Cater, Trevor Jones, and David McEvoy

(1981) "Business Development and Self Segregation: Asian
Enterprise in Three British Cities," in Ethnic Segregation in
Cities. C. Peach et al. (eds) London: Croom Helm.

Aldrich, Howard, Trevor Jones, and David McEvoy (1984) "Ethni.

Advantage and Minority Business Development," in Ethnic
Communities in Business. R. Ward and R. Jenkins (eds) London:
Cambridge University Press.

Anwar, Muhammad (1979) The Myth of Return: Pakistanis in Britain.

London: Heinemann.

Central Statistical Of-ice (1985) Monthly Digest of Statistics.

Numb. 471. March.

Desai, Rashmi (1963) Indian Immigrants in Britain. London: Oxford

University Press.

113
Hormann D.M. and Thuo, J.M. (1979) "Smallholder Production and
Marketing of Vegetables for Export" Unpublished paper.

Hunt, Sandra (1975) "The Food Habits of Asian Immigrants"

Published by Van Den Berghs and Jurgens, London.

Jaffee, Stephen (1986a) "The Demand and Distribution Channels for

Kenyan Fruit and Vegetables in the United Kingdom: An
Institutional and Economic Analysis." Report to USAID/Kenya.

(1986b) "The Kenyan Horticulture Export Sector:

An Economic and Institutional Analysis of Alternative Marketing
Channels," Report to USAID/Kenya.

Jones, Trevor (1983) "Residential Segregaticn and Ethnic

Autonomy," New Community. Aut-Winter.

Key Note Publications, (1986) "Ethnic Foods."

Miller, Charles (1971) The Lunatic Express. New York: Ballantine

Books.

Ministry of Agriculture. (1974) "The Marketing System of Fruit

and Vegetables in Kenya. Case Study #4: Machokos District of
Eastern Province."

Ministry of Finance and Development. Machokos District

Development Plan 1984-1988.

National Opinion Poll Market Research Ltd. (1974) "A Report for
the Tropical Products Institute."

Office of Population Censuses and Surveys. (1985) Birth

Statistics 1983

(1982) Population Trends (28)

(1981) OPCS Monitor Ref. PPI

(Various Issues) International Migration.

Robinson, Vaughan. (1986 Transients, Settlers, and Refugees:

Asians in Britain. Oxford: Clarendon Press.

Tandon, Yash and Arnold Raphael. (1984) The New Position of East
African Asians: Problems of a Displaced Minority. London:
Minority Rights Group Ltd.

Ward, Robin (1983) "Ethnic Communities and Ethnic Business," New

Community. Aut-Winter.

114
Wekundah, J.M. (1984) "Analysis of Horticulture Production and
Marketing in Yatta, Machokos District" Ministry of Agriculture
and Livestock Development.

Wilson, R. (1977) "The Market for Selected Tropical Foods Amongst

Immigrants in the United Kingdom," Tropical Science. p. 88-107.

Winter, J.D., Jacqueline Stother and Daidy Kay. (1969) "A Report
to the Kenyan Government on the Market Prospects for Selected
Horticultural Commodities in Western Europe," Tropical Products
Institute. August.

115
CONTRACT FARMINGP MARKET CONDITtONS, AND THE VEGETABLE
DEHYDRATION INDUSTRY IN KENYA, 1964-1982

"Contract Farming in Africa"---Case Study #3

117
Introduction

The vegetable dehydration industry is the first case in

Kenya where small-scale African farmers became a party to written
production contracts with an agricultural processing firm, but
this smallholder contracting component was perhaps the only
successful dimension of a project that spanned two decades and
featured numerous changes in ownership, management, and operating
strategy. Severe problems were faced in largeholder and nucleus­
estate raw material procurement, in processing and in mLrketing.
While initiated in 1964 largely for the social benefit of
improving the welfare of newly settled African smallholders in
the former "White Highlands," the project never succeeded in
making a transition into an economically viable venture.
Government subsidies, foreign investment, and multinational
marketing were all marshalled to put the project on a sound
commercial basis but low levels of operational efficiency and
adverse changes in :,iarket conditions (both loca3 and
international) led the project into a financial abyss.

Little has been written on the project and few publically

available documents provide any information about the
participants, organization, or performance of the project. (I)
The information presented in this study has been drawn primarily
from a selected number of government documents and from sections
of company records. I have also relied on information provided
by the former agricultural manager of the project. Information
about the world market for dehydrated vegetables was obtained
from secondary sources.

This study provides only an initial overview of the

project's development, market environment, internal structure,
and performance. Many important dimensions of the project
warrant further study. A fuller understanding of the
microeconomics of the project as well as the institutional
linkages among participants would require a more thorough review
of company records and a wider range of interviews with project
participants than was possible in the course of this research.

We begin by discussing general features of dehydrated

vegetables and their international market. We then provide an
overview of the foundatiox. and the early performance of the
Kenyan dehydration industry. This covers the period from 1964 to
1972 when the industry was oriented primarily toward providing a
market outlet for small-scc~le farmers, but was not economically
sustainable because of its limited operating scale as well as its
management and marketing problems. In the early 1970s a plan was
developed to expand the industry and to link it to an
international expert in the field. We examine the new project
concept and its main participants. We then move on to the core

119
of the study: an examination of the organization and performance
of the industry over the 1975 to 1982 period. We discuss the
international market environment in which the new project was set
and explore the marketing, processing, and raw material
procurement problems that were faced by the new project. In
problem, we contrast the
discussing the raw material supply
schem.:e
relatively successful smallholder-farmer contract farming
from large-scale farms
with the problematic sourcing of supplies
and company estates. We then provide a few concluding comments
and raise a series of questions for further research.

Dehydrated Vegetables

Dehydrated vegetables have been produced in small quantities

since the 19th century. The product was used by British naval
expeditions in the mid-19th century and by both soldiers arid
civilian populations during subsequenL wars. Advances in
processing technologies after World War II brought significant
improvements in the quality of dehydrated vegetables. (2) The
demand for convenience foods began to grow in the 1950s and
accelerated in the following two decades. The dehydrated
vegetable industry 4ould benefit frov this growing demano for
convenience foods.

Dehydrated vegetables are less bulky and lighter in weight

than fresh or other processed vegetables. They are cheaper to
pack than canned vegetables and do not require refrigeration as
do frozen vegetables. Dehydrated vegetables have a long shell
life, extending several years for some items. (3)

The major use for dehydrated vegetables is in the

manufacture of dried (or packet) soups. The demand for
dehydrated vegetables is thus a derived demand, based on
production and consumption of (primarily packaged) soups.
Secondary uses of dehydrated vegetables are in baby food, canned
soups and stews, and a variety of ready-made meals. Seventy-five
percent of West European imports of dehydrated vegetables are
supplied *,o soup manufacturers. A furth-r 20 percent is supplied
to the catering sector and to institutions (i.e., hospitals and
schools). The remainder is used by general food manufacturers or
sold directly to consumers as dehydrated vegetables. (4)

In most Western European countries the domestic production

of dehydrated vegetables reached a peak sometime in the 1960s or
early 1970s and declined thereafter as a consequence of rising
raw matrial and labor costs. In some countries a small number
of large and diversified firms have continued to produce smaller
quantities of high-value, high-quality dehydrated vegetable
items. Consumer and manufacturer demand has been met largely by
increased imports from Eastern Europe, the Mediterranean, and

120
Asia. A large number of countries, both industrialized and
developing, now supply the West European market, and competition
is heavy both in terms of quality and price. Price flucuations
are common owing to changes in supply and/or demand
conditions. (5) Transport costs play a relatively insignificant
role in the relative competitiveness of different countries.
Dehydrated vegetables are sent by sea freight, and transport
costs tend to be 10-15 percent of import costs.

In most supplying countries vegetables for dehydration are

grown almost exclusively on contract for processors with
contracts stipulating acreage, planting periods, varieties, stage
of maturity at harvest, dElivery dates, grading, and prices. It
is generally considered that required continuity as well as
varietal specificity of raw material cannot be assured by buying
on the fresh market. k6)

The world dried-soup industry is dominated by three firms--

Unilever, Nestle, and Knorr (CPC Intl). These firms hold a
preponderant market share in nearly all Western European
countries. As the main users of dehy-rated vegetables these
firms have strongly influenced the standara trading practices in
the industry. These firms have set high quality standards for
their suppliers in terms of cut, color, moisture content,
bacteria level, flavor, and rehydration time. Historically,
price has been a secondary factor after quality in supplying raw
materials to this market sector. The soup manufacturers have
generally preferred not to purchase directly from overseas
producers, but instead buy from well recognized importers who
have the capacity to test, reprocess, regrade, and repack
supplies. Developing-country exporters thus tend to deal with
brokers or importer/packers rather then directly with soup
manufacturers.

In contrast to the soup sector, for bLyers serving the

catering/institutional sector, price is a major consideration and
quality standards are set lower. Standards set by baby food
manufacturers are the highest, but supplies fetch a considerable
price premium. This sector is small in volume relative to the
former two. (7)

VeQetable Dehydration in Kenya

1964-1972 Subsidized Trial and Error

Foundation

In 1964, less then a year after an initial investment

proposal was submitted, Pan African Produce and Development
Company started dehydrating vegetables at a smali Naivasha

121
factory. Th- factory had a capacity of producing 450 tons of
finished product annually. The company's main sponsor and
shareholder was Biddle and Sawyer Company, a London-based firm
that had been prominent in the marketing of Kenyan pyrethrum.
With Kenyan Government approval, also investing in the project
was the Development Finance Company of Kenya (DFC). Minor shares
were held by a few other private parties.

The main reason for the Government's interest in the project

was the creation of an outlet for the vegetables produced by the
smail-scale farmers who were settling near Lake Naivasha and in
Nyandarua District (i.e., the Kinangop Plateau) under the One
Million Acre settlement scheme. At Kenyan independence large
European farms in the highlands were purchased by the Government
with British financing. Settlement schemes were developed to
allocate land to smallholders and landless Africans. Seventeen
settlement schemes, each with a size varying from 10,000 to
18,000 acres, were established. African settler families were
provided with plots of 20 to 60 acres, although generally only 5­
10 acres of each plot were arable. With the backing of the
Ministry of Lands and Settlement, each settlement scheme was to
develop its own cooperative with its own administration,
technical equipment, workshop, and agricultural advisor. Farmer
membership in these cooperatives would be mandatory. (8)

As originally conceived, the project would combine private

and public interests in a production scheme that theoretically
would not only generate export earnings and improve the welfare
of newly settled smallholder farmers, but would also assist in
developing the country's cooperative movement. In theory the DFC
shareholding was being held in trust for the grower cooperatives,
which after accumulating a sufficient surplus, would purchase
these shares on behalf of their farmers.

Outgrower Contracts

As noted above, Kenyan Government support for the project

rested largely on the expected benefits that would accrue to the
settlement farmers. Most of the lond held by these farmers was
kept under permanent pasture to support their livestock. Milk,
sold through the settlement cooperatives, would become these
farmer's main source of income. Arable land was used to produce
maize, potatoes, and other vegetables. Carrots grew particularly
well in the Kinangop area. The farmers faced problems marketing
their produce as the road network in the Kinangop was poor and
Nairobi traders had easier access to vegetable-growing areas
nearer to the capital.

The dehydration company decided to base its raw-material

procurement system on production contracts with farmers. Since

122
it was felt that contracting directly with newly settled farmers
would be administratively difficult and financially risky, the
company decided to enter into written contracts with the
settlement cooperatives. The cooperatives would act as
"channeling funnels" for inputs and technical assistance and as
units for production planning. It was further felt that the
cooperatives would be well placed to assist in contract
enforcement and debt collection, since they would also be
marketing the farmers' milk and, if necessary, deductions could
be taken from payments for this commodity. (9) We examine these
production contracts in a later section.

The company's procurement of raw material incorporated two

other groups of farmers. On group consisted of people working
in the afforestation schemes of the Ministry of Natural
Resources. Workers employed on these sc:hemes to clear bush were
permitted to utilize space between tree rows for agricultural
purposes. Each worker had access up to 7.5 acres. Four forest
stations would serve as the intermediary between these farmers
and the company. (10) Both the settlement farmers and the forest
station workers were initally contracted to grow primarily
carrots for the factory.

Large-scale European farmers operating around Lake Naivasha

comprised the third group to benefit from the project Some of
these farmers had begun growing vegetables in the 1940s,
initially growing potatoes and onions for local sale and later
starting to grow capsicums and French beans under irrigation for
fresh export to Europe. (11) Growing vegetables for the factory
was a useful supplement to these other activities and helped
defray the high initial investment that these farmers were then
making in infrastructure and irrigation systems. Farmers were
particularly interested in growing for the factory during the
export off-season. With these larger farmers written contracts
were rare. The provision of seed by the company, and the
farmer's commitment to provide his output to the factory, were
based on trust. These farmers would concentrate on specialist
crops such as French beans and capsicums. Producer prices would
be collectively negotiated based on agreed estimates of
production costs. (12)

Erratic Performance

The project was supported lar( ?ly on social and political

grounds, rather than on commercial grounds. The private
investors viewed the investment as a pilot project to examine the
technical and market prospects for a larger venture. They had no
technical expertise in the field and the managers appointed to
run the factory had no experience with dehydrating vegetables.
Machinery and equipment were purchased from several sources, some

123
local and some foreign. Some machinery was badly designed or not
in full working order. As an assessment of raw-material
procurement potential had not been made, it is not surprising
that some of the equipment purchesed was for use in processing
vegetables that could not be procured economically in the
Kinangop area. (13)

The project sputtered along for four years making continuous

losses. A management overhaul in 1966 had only a minor effect on
performance. It was becoming obvious that the factory s very
small capacity made the entire operation uneconomical. Overheads
were swallowing sales earnings. The factory was only operating
at less than 50 percent of its small capacity in several years.
The company was exporting small quantities of low quality carrot
powder to the United Kingdom. Exports marketing was ad hoc,
involving little preplanning or long-term contracting. European
manufacterers would not enter into longer-term trade arrangements
because of the uncertainty of supply and quality associated with
the Kenyan product. The annual export levels were the following:
1965--102 tons; 1966--117 tons, and 1967--217 tons. These export
volumes, combined with the low prices that the Kenyan product
could fetch, led to continued financial losses. In March of 1968
the company went into receivership. Later that year the factory
was purchased by the Kenyan Government and renamed Pan African
Foods (1968). The government wished to prevent the closure of
the factory with its subsequent adverse effects on the contracted
farmers.

Following the government's purchase of the factory several

adjustments were made that improved some aspects of the company's
performance. Additional machinery was added to the factory to
bring its capacity up to 600 tons of finished product per year.
It was also decided that the factory's raw material intake
required greater diversification. Smallholder production had
concentrated on carrots, and this item formed most of the
factory's supplies. This contributed to financial problems as
carrots generate a lower profit margin and lower unit sales
earnings than vegetables such as green beans, capsicums, and
onions. To increase commercial viability the company would have
to put greater emphasis on procuring the higher-value vegetables.
This raw material diversificaton would require greater reliance
on the Naivasha farmers. The Kinangop area features low
temperatures at night plus clouds and high humidity in the early
morning. Thus, crops such as onions or beans which have high
photoperiod sensitivity do not grow well there. (14) Smallholders
would be encouraged to grow more leeks and cabbages.

Some success was made in diversifying raw material supplies.

While in 1970 eighty-two percent of the weight of raw materials
processed consisted of carrots, by 1972 the share of carrots was

124
down to sixty-seven percent. The company had succeeded in
increasing smallholder supplies of leeks and cabbages and large­
farm supplies of capsicum, beetroot, and French beans. (15)
Total raw material supplies by "he Naivasha growers were,
however, showing signs of instability by the early 1970s. Large­
farmer supplies to the factory fel. from 4306 tons in 1970 to
2971 tons in 1971 and down to 1960 tons in 1972. Many large
farmers were becoming more actively involved in the fresh export
trade, adding crops such as courgette to the initial basket of
French beanp and capsicum. Exports of fresh capsicums to Europe
increased four-fold between 1969 and 1972. Another outlet, that
of the Nairobi greengrocer serving a higher income clientele,
also grew in size and paid prices above those of the factory. (16)

Performance in the smallholder component was more favorable.

Cooperative vegetable supplies to the factory more than doubled
from 2304 tons in 1970 to 5234 tons in 1972. This occurred
despite the fact that by 1970 the Ministry of Lands and
Settlement had lost interest in the project bnd no longer wanted
the project justified on the basis of the social benefits
accruing to newly settled farmers. Initially the settlement
schemes had been underfinanced and lacked effective institutional
structures to channel the needed finance, equiprnent, and
technical assistance to the farmers. The Pan African Foods
project thus required the support of the local Ministry of Lands
and Settlement officials to get the cooperatives sufficiently
organized to perform project-related functions. Cooperative
staff had been both meagre and unqualified. However, by the time
the Ministry withdrew its support the cooperatives had built up
their own staffs. While some cooperative management problems did
arise it does not appear that these problems were nearly as
debilitating as those facing horticultural cooperatives elsewhere
in Kerya. Sometimes payments to farmers were delayed until
cooperative bills were paid and sometimes limited quantities of
inpuLs did "disappear." Still, overall cooperative performance
was adequate.

Trading performance over the 1968-1972 period was varied,

although better than during the earlier years of the project. (17)
Exports varied from year to year with the project being adversely
affected by drought in the Kinangop during both 1969 and 1971 and
by heavy rains during 1970 which resulted in extremely high
moisture content in carrots. Export levels were the following:
1968--595 tons; 1969--450 tons; 1971--297 tons, and 1972--572
tons. In 1968 the company diversified its sales into the West
German market, and by the early 1970s this was the company's
largest market.

While the quality of the factory's product did improve over

earlier years, Kenyan sales were still at lower prices than other

125
major suppliers. Kenyan supplies were largely being sold as
second quality to the catering/institutional sector as bacteria
count was higher than the limits set by the soup
manufacturers. (18) Obtaining long-term contracts thus remained
difficult. Quality control problems reduced the prices the
company could obtain. Some indication of the magnltLce of these
quality-related price discounts can be seen in the following
figures:

West German Import Prices for Carrots

(S per ton)

Year Average (All countries) Kenya

1969 817 701
1970 821 795
1971 757 730

Source: ITC 1972 (19)

From 1968 to 1972 the company operated in the red. In most

years the factory was provided with an annual government subsidy
of 20,000 pounds ($56,000) in order to cover its expenses.

1973-1974: Enter the Experts

A 1970 government working party examining the condition of

the horticultural sector argued that since its estabiishment the
vegetable dehydration project had been operating on an ad hoc
basis, never deve±oping a souni, long-term plan to develop the
industry and never adequately utilizing experts in this product
field. The group recommended that the governmerit enter into a
joilit-venture project with a major European or American firm that
wouLd provide finance, technical know-how, and established
distribution outlets. After several aborted contacts, the
goverument finally agreed to a proposal made in 1973. (20)

The new project would entail majority government control

through the shareholding of the Industrial and Commercial
Development Corporation and a minority shareholding by Sifida
Investment Company (Swiss), Bruckner Werke (W. Germany), Barclays
Overseas Development Corporation (U.K.), and several other
shareholders. The new project would involve $3.5 million of new
investment in the form of equity and debt. A new factory would
be built near the old factory site. It would have a capacity to
produce 3000 tons of dehydrated vegetables annually.

The central participant in the project would be Bruckner

Werke. Bruckner has been the largest producer of dehydrated

126
vegetables and potatoes in West Germany and has a major share of
that country's imports and exports of dehydrated vegetables.
Bruckner would be responsible for obtaining and installing the
machinery for the new factory. Also, in coordination with
company management, Bruckner would determine an annual program
for raw-material supply to the factory and a processing plan
which would result in a product mix and volume of supplies
sufficient to meet sales contracts. Bruckner would provide
technical assistance related to raw material production as well
as processing and packing methods. Finally, Bruckner would have
exclusive overseas marketing rights to the Kenyan company's
output. Any local or foreign sales that the company wished to
make on its own would require the approval of both Bruckner and
SIFIDA.

At full operation four years into the project, the company

expected to be producing 2560 tons of dehydrated vegetables using
nearly 33,000 tons of raw material. According to the production
plan, output and raw material sourcing would be as follows:

Product Planned Output Procurement

(Dehydrated Large Farms Smallholders
Product)

Carrot 975 tons 25% 75%

Onions 570 100
Leeks 400 50 50
Peppers 250 100
Beans 200 100
Cabbage 125 75 25
Beetroot 45 50 50
Tomatoes 45 100

Source: SIFIDA

Using company estimates for yields and required acreages,

one finds that the investment plan called for raw material
supplies from large farms of 20,115 tons (62 percent) while
supplies from smallholders would be 12,670 tons (38 percent).
This would represent a doubling of smallholder deliveries and a
ten-fold increase in large farm deliveries over the actual 1972
levels. Considering differential values for the various crops,
approximately 3/4 of farm-level income would accrue to the large
farmers under this plan.

While acknowledging that irrigation costs require large

farmers to plant crops bringing maximum revenues and while noting
the increased interest in producing veqetables for export, the
foreign investors were confident that raw material requirements
could be met: "No serious difficulties are foreseen to increase

127
the present production of fresh vegetables (8000 tons p.a.) to
the quantity needed for the new factory (33000 tons in 1977)
." (SIFIDA, p.2) There was thus considerable optimism about
the potential to increase raw material in-take to meet the new
factory's large capacity.

There was also considerable optimism felt about marketing

prospects. Past trends led the compAny to believe that West
European demand would continue to rihe at a steady 5 percent per
year. For the three largest markets--West Germany, the U.K., and
the Netherlands--combined imports of dehydrated vegetables more
than doubled from 1965 to 1970 from 16,102 tons to 35,566 tons.
Growth in imports had been steady year-by-year as domestic
production of dehydrated vegetables declined in several
countries. (21) For example, West German production of dehydrated
vegetables actually peaked in 1963, declining ther,. fter. The
market for dried soups continued to grow at a fast pace. Because
of the low capacity of its factory, Pan African Foods had not
been able to take advantage of the expanding European market
during the 1960s and early 1970s.

The invectment proposal appeared to provide solutions to the

project's existing problems and considerable confidence in
expanding the industry. Commercial viability would be guaranteed
by the expansion of capacity, by the increased emphasis on higher
value products, and by the participation of a firm with technical
expertise and excellent marketing skills and contacts. While the
relative importance of large farms for raw material supplies
would be increased, the company's plan included an expectation of
expanding smallholder deliveries, thus increasing income flows
into the settlement schemes.

The joint venture investment was approved by the Government.

It represented for several parties a risk-reducing effort. For
Bruckner Werke the project represented an opportunity to
diversify its sources of dehydrated vegetables and thus reduce
the risk of shortfalls from its other suppliers. The company's
minor equity holding did not represent a substantial investment
and even this was off-zet by earnings associated with the
procurement and insta)iment of the new plant and equipment. The
new initiative also ei.abled various government officials to
reduce their political and institutional risk, zs now the project
had incorporated "international experts." One of the roles of
thes ? experts would be to relieve certain officials of decision­
making responsibilities over issuen for which they lacked
training and experience.

128
 1975-1982 Pan African Vegetable Products, Ltd.

The new company began operations in 1975. It was composed

of two legal entities. One was the holding company Pan African
Vegetable Products (PVP) whose purpose was to process and market
dehydrated vegetables. The second was a wholly-owned subsidiary
called Pan African Vegetable Products Estates, which was to
manage nucleus farms and supply fresh vegetables to the holding
company.

From the beginning, the company's performance trailed behind

the expectations ol both the Government and the private partners.
Even with its expanded capacity and virtually guaranteed market
access, the company was n.ver able to earn an annual net profit.
Financial losses accumulated year-by-year and frequent government
subsidies were required to keep the company operating. The
company experienced severe problems in raw material procurement,
in processing, and in marketing, and continued financial losses
fed back to magnify the problems in each of these areas.

The financial picture of the company was dismal from the

start. The quadrupling of oil prices in the mid-1970s
considerably increased production costs. Fuel oil would be the
prime source of energy for the factory, used to generate the hot­
air process for dehydration. Less than one year into the project
it was estimated that even if the factory were operatirl at full
capacity, the increased costs would result in an operating profit
only 35 percent of that originally forecast in the feasibility
study. (22) In fact, the factory never even came close to
operating at full capacity. naximum capacity utilization was
reached in 1977 at approximately 70 percent and annual capacity
utilization averaged just over 50 percent.

Financial losses were generally in the range of Ksh 2-5

million per year. Accumulated losses reached Ksh 22.8 million in
1979 and Ksh 45 million in 1982. Working capital was also a
problem. In 1977 and 1978 the Ministry of Agriculture and the
Treasury provided Ksh 4 million. As accumulated losses absorbed
all finance, the company's situation was considered irreversible
as early as 1978. In that year the company began defaulting in
its repayment of overseas loans. It kept operating by delaying
payments for inputs and raw materials, by a limited injection of
fresh (government) equity, and by making full use of an overdraft
facility. By 1980 the company's bankers were refusing tu honor
its checks. In 1982 PVP went into receivership. (23)

Despite its overall poor financial performance, PVP did have

considerable developmental impact. In the late 1970s it earned
an average of Ksh 11.5 million per yeec in foreign exchange.
Also, it became the second largest employer at Naivasha with a

129
combined labor force in its factory and on its estates of 1600
people. Furthermore, the company provided a valuable source of
income for up to 3000 smallholder farming families.

We begin our review of PVP by first examining the general

market environment in which it operated in the late 1970s. We
then go on to examine PVP's marketing, processing, and raw
material procurement problems.

Market Stagnation

Pan African Vegetable Products started operations at a time

when Western Europe was in the midst of an economic recession.
The recession had been brought on partly by the quadrupling of
oil prices after 1973. Economic rates of growth were declining
and consumer demand for numerous items was down. Both the
production and consumption of soup declined in several countries.
Between 1973 and 1975, the production of canned and packet soup
in West Germany declined Irom 98,200 tons to 81,000 tons. (24)
The dehydrated vegetable industry suffered as a consequence.
Compare below the imports of several cruntries for the year 1970
with those for 1975 in the midst of the recession:

Effect of Recession on Dehydrated Vegetable Imports

(Figures are Tons per Year)

Year W. Germany U. Kingdom Netherlands Total

1970 13271 15574 6721 35,566

1975 11330 11870 6191 29,371

Sources: ITC 1972; 1981

Even with economic recovery in the latter half of the 1970s,

the market for dehydrated vegetables remained stagnant. The
combined imports for West Germany, the U.K., and the Netherlands
for 1978 was only 34,613 tons, a level below that for 1970.

Through its marketing agreement with Bruckner Werke, PVP

would be exporting most of its finished product to West Germany.
It is significant to note that West German production of packet
soups actually declined over much of the 1970s. This can be seen
in the data below:

130
 West German Packet Soup Production (tons '000)

1971 42.1
1973 43.4
1975 39.0
1978 39.1
1979 36.5
1980 37.3

Sources: Marketing! In Europe, April 1976; July 1984

The D-Mark value of production was no higher in 1979 than it

was at the beginning of the decade. This pattern was not limited
to West Germany. For example, consumption of packet soups also
declined in the Netherlands in the late 1970s, falling from 156
million liters in 1977 to 129 million liters in 1979. (25)

The mid to late 1970s was a period not only of fluctuating

and/or declining demand for soups and dehydrated vegetables in
Western Europe, but it was also a period when the countries of
Eastern Europe as well as Egypt, China, Taiwan, and Morocco were
increasing their supplies of dehydrated vegetables onto the
market. Price competition thus tightened. Several countries
heavily subsidized their dehydrated vegetable industries or used
this product in barter or compensation dears. (26)

As a result of stagnant demand and increased market

penetration by several suppliers, overall market prices exhibited
no nominal increase over the course of the 1970s. Compare, for
example, the ex-factory prices in West Germany for several
dehydrated vegetables that Kenya also supplied to that market:

Ex-Factory Prices in West Germany

Product Price (Dn/Kg.)

1970 1980

Carrots (cubes/flakes) 4.40-5.40 4.00-4.50

Carrots (powder) 4.20 2.50
Leek, white(slices) 5.70-6.00 5.70-6.00
Leek, white-green(slices) 5.00-5.30 4.00-5.50
Beetroot (powder) 7.70 4.50

Source: ITC 1981

With the exception of beans, the import prices in West

Germany for items that Kenya also exported do not show a pattern
of increase in the late 1970s which would have compensated for

131
increased production costs arising from higher energy costs.
This can bc seen in the figures below:

West German Average Import Prices (DM/Kg.)

Product 1975 1976 1977 1978 1979

Carrots 3.79 3.72 3.83 4.02 3.lr

Leeks 3.96 3.69 4.31 3.98 4.23
Beans 5.54 5.86 7.95 7.92 7.20
Onions 3.36 3.25 3.79 3.64 3.27

Source: Calculated from data in ITC 1981(27)

The Marketing of PVP's Products

Was the stagnant position of the West European dehydrated

vegetable market the prime cause of the company's finailcial
problems and ultimate demise? Did the company's tied marketing
arrangements with Bruckner Werke contribute to lower returns from
exports? The evidence suggests that neither the overall market
situation nor the company's marketing arrangements were major
contributors to the problem.

Before examining PVP's marketing problems, let us first

examine PVP's performance in terms of export volumes and sales.
At full operating capacity the company had expected "o produce
2560 tons of finished product per year. As we can see in the
following figures, its maximum export level was only 53 percent
of this figure, reached in 1976.

Kenyan Exports of Dehydrated Vegetables

Year Quantity (Tons) Value

(Ksh Million)

1975 479 4.08

1976 1362 15.34
1977 1326 17.70
1978 949 18.75
1979 1340 23.81
1980 1044 18.30
1981 832 13.47
1982 385 6.97

Sourcet Kenya Annual Trade Reports

132
 During this period, between 60 percent and 80 percent of
exports went to West Germany, with the remainder going to the
U.K. and the Nethorlands. By 1979 Kenya had become the leading
supplier of dehydrated carrots, leeks, and beans to West Germany.

In the original marketing agreement with Bruckner Werke, the

latter would be responsible for all ove-seas marketing of PVP's
products. Marketing had proven to be a major problem of the
earlier dehydration company, and it was felt that Bruckner could
guarantee PVP market access and obtain for it favorable prices.
The exlusive marketing agreement held in force until December 31,
1977. Although a number of draft agreements were drawn up in
1978, no new marketing contract was signed. From that point
onward the parties operated on a quasi-contractual basis,
sometimes wishing to enforce the terms of the original agreement
while at other times seeking alternative arrangements. (28)

From the begin.I±ng the marketing links between PVP and

Bruckner were an 3rena of conflict, distrust, and dismay. PVP
management felt that Bruckner was paying insufficient prices,
that Bruckner was not providing management with sufficient market
information, and that under the prevailing marketing arrangements
several potentially promising distribution outlets were not being
properly developed. Bruckner was disturbed by the factory's
inability to maintainx high quality standards and by PVP's
inability to produce according to production plans. Complicating
the marketing situation was the fact that Bruckner was also a
shareholder in PVP and had major input into production-related
decisions.

From as early as 1976, PVP managerR were becoming concerned

about the marketing arrangements with Bruckner. PVP had little
understanding of the market and was dependent upon Bruckner to
provide all market information. Bruckner was unlikely to pass on
information that would improve PVP's bargaining position as
regards pricing. Thus, only scanty market price information was
provided. (29) PVP's information on its own production costs was
not very reliable and subject to "editang" by Bruckner. Thus,
Bruckner was virtually able to dictate prices. In addition, many
PVP shipments vere sent direct to end-users without Bruckner
taking possession at all. PVP was obtaining enquiries from some
of these end-users. This signalled to the management that PVP
could perhaps by-pass the "middleman" (i.e., Bruckner) and obtain
better prices. PVP management was almo suspicious that Bruckner
was tailoring the product mix to suit its own sourcing
requirements rather than emphasizing a mix that would obtain the
best sales return for PVP. (30)

In December 1976, PVP management examined the pattern of its

selling prices to Bruckner up until that time. It found that

133
there had been slight price increases for a fcew items, but that
the price levels for most items were below those predicted in the
earlier feasibility study. Still, management did not know
whether this was due to the depressed market or due to the
Bruckner monopsony on PVP's products. A year later the PVP
management gained access to data from the International Trade
Center that compared 1976 and 1977 import prices into West
Germany and the NethErlands for Kenyan dehydrated vegetables and
for these products from other sources. While the results were
somewhat mixed, they did show that in 1977 Kenyan leeks, beans,
and potatoes were obtaining lower prices than alternative
suppliers. Was Bruckner paying "too boy" a price? PVP
management thought so and put in a claim to Bruckner for D-Marks
293,343. With the original marketing agreement approaching its
end-date of December 31, 1977, various attempts were made to
draft a renewal contract containing revisions in certain clauses.
None of these revised agreements were actually brought into
practice, but it is interesting to note some of the proposed
changes. For example, it was proposed that the proportion of
output going to Bruckner be progressively reduced to 50 percent.
It was also proposed that Bruckner's payment be within 30 rather
than 60 days in order to improve PVP's cash flow position.
Further, it proposed that contract prices be "comparable to world
prices." The most interesting proposal was that PVP would
develop its own sales unit for direct sales both locally and
abroad and that "to enhance di.-ect marketing the company will
negotiate for a share of the markets where Bruckner Werke is
already represented. "(31)

In the late 1970s PVP did increase its level of sales on the
local Kenyan market and did begin to make sales direct to several
European companies other than Bruckner. The prices obtained on
the local market were considerably )igher than those offered by
Bruckner, converted into Kenyan shillings. Several of the orders
made by European companies were also at prices above those
offered by Bruckner. However, when PVP sent a detegation to
Europe to inquire about the scope for expanding these direct
sales, Bruckner threatened to cease its involvement in PVP
product distribution altogether.

What was Bruckner's perspective on its marketing links with

PVP? Bruckner's marketing strategy was based primarily on long­
term (i.e., annual) contracts with major food manufacturers and
institutional buyers. Based on buyer requests and the production
possibilities in Kenya, Bruckner and PVP were to develop a
production plan for the factory and shipping schedule. The PVP
operation served as one of many sources for the company and thus
the planned product mix for each year would reflect Bruckner's
expectations of supplies from other sources. It would be
difficult to argue, however, that PVP's product mix was dictated

134
by the wishes of Bruckner alone. B-uckner's largest orders were
for the lower value carrots, cabbages, and leeks. These had unit
values only 1/2 to 2/3 those of beans or capsicums. Carrots
remained PVP's main item accounting for 60 percent of exports in
the late 1970s. However, this proportion is actually lower than
the share of carrots in Kenya's exports a decade earlier before
Bruckner was involved. Bruckner found that PVP consistently
operated far behind schedule on contracted deliveries for beans,
leeks, and capsicums, and that Bruckner itself was unable to
fulfill its contracts with the customers. Bruckner contended
that it waa inappropriate simply to examine official import
statistics in order to compare supply prices. It responded to
PVP's price diiocount claim by pninting out a number of
extenuating circumstances that had influenced the annual
laverage" import prices in West Germany and the Netherlands
during the years for which the PVP claims applied, and provided
evidence that PVP was generally receiving prices above average
world prices. (32)

Bruckner's critical concerns related to the quality and

reliability of PVP products. For many sales, particularly those
destined for customers outside of West Germany, products would be
sent directly by PVP to the customers without Bruckner inspecting
the consignment. For at least four major consignments during
1975 and 1976 either the customer rejected the lot outright or
demanded a price reduction from Brucxner. On these and other
occasions Bruckner was forced to ship consignments to its own
factory for testing, reprocessing, and repacking. Sometimes the
material could only be sold to producers of dog food or to
chemical companies. Bruckner's customers complained that PVP
supplies sometimes had high bacteria counts, high levels of S02,
contained foreign matter, had vegetables of the wrong cut, or
contained rotten material. (33)

Delayed deliveries were said to have resulted in cancelled

sales contracts for Bruckner. On some occasions the customers
went on to buy elsewhere to cover their requirements and paid
higher prices. Bruckner would then receive the invoice for the
price difference. (34)

On at least two occasions, Bruckner placed claims against

PVP to compensate them for the costs associated with problems in
quality or delay. The first claim was made in 1977 for
documented cases during 1975 and 1976. The value of the claim
was DM 105,065, equivalent to about 1.3 percent of PVP foreign­
sales revenues. Several later claims were of perhaps
questionable authenticity. For example, in 1978 PVP's financial
manager transferred tn Bruckner the sum of Ksh 406,686 against
compensation for undergraded products. The products were neither
returned nor certified by an independant statutory body as being

135
"disposed of. " In addition, a clause in the marketing contract
stipulated that payments should be made after 60 days of receipt
and that any money paid by Bruckner prior to 60 days be treated
as an "advance payment," subject to interest. Even though the
marketing contract officially lapsed in 1977, over the 1978-80
period Bruckner debited PVP the sum of Ksh 635,329 for such
interest payments. (35)

Clearly PVP's marketing position was not optimal. At

certain times better prices could have been obtained if the
company had bypassed Bruckner and sold directly to end-users.
PVP was certainly not obtaining full market information from
Bruckner and thus did not know about a number of short-term
opportunities. Clearly, the exclusive marketing arrangement
limited the scope for Kenyans to learn about the market and
develop marketing expertise. PVP was thus extremely vulnerable
to strains in its trading relationship with Bruckner as PVP
lacked a credible threat of sending most of its supplies to
Bruckner's competitors or customers.

However, what Bruckner did provide PVP was guaranteed market

access. In the increasingly competitive but stagnant market of
the late 1970s, it is not at all clear that PVP would have been
able to act independently and supply the volumes that it did.
Things might have been different if PVP was supplying
consistently high quality products on a reliable scheduling
basis. The fact the quality and reliability were indeed major
prcblems made the link with Bruckner (or a similar type of firm)
absolutely necessary. It is certainly not clear that Bruckner
was paying PVP prices that were "too low." A review of
Bruckner's contracts with its customers over the 1976-1978 period
revealed wide variations in the firm's selling margins, but
certainly not a general pattern of sensational profits. For
sales contracts for carrots and beans Bruckner's margins varied
between I percent and 11 percent with the higher margins being
asszciated with lower volume sales. (36) Bruckner had little
incentive to "bleed" PVP since the latter had developed into an
important supply source for several items.

PVP's Processing Problems

Throughout the life of the project the fnctory operated at

well under its full capacity. Annual capacity for raw material
in-take was 33,000 tons. We can see in the figures below that
low rates of capacity utilization prevailed.

136
 Factory Capacity Utilization

Tons/yr (rounded) % Utilization

1976 21.000 64
19'7 '2,000 70
1978 19,000 58
1979 20,000 61
1980 13,000 40
1981 11,000 33

Average 17,700 54

Operational inefficiencies at the factory also contributed

to the poor financial performance of the overall operation.
Important inefficiencies were related to poor conversion rates
for raw material into finished product and poor quality control.
The quantitative significance of these factors can not be
assessed since the factory lacked a cost-accounting system
calculating unit costs.

Even though the new factory contained modern equipment, the

new operation obtained worue conversion rates than that achieved
in the old factory. Profitability clearly depends on achieving
the optimal ratio of raw material to end product. For carrots
this ratio should be 12:1, but the actual results were closer to
16:1 in some years. Similar poor results were being obtained for
other crops. (37) Although it was never actually admitted, this
loss of dry matter (by leaching or wastage) was a basic cause of
unprofitability.

Previously we discussed Bruckner Werke's concerns about the

quality of PVP's final product. Factory breakdowns, absence of
spare parts, poor maintenance, and frequent management turnovers
were all characteristics of factory operations, particularly once
the company's financial position reached crisis point. An
analysis of 1980 factory production showed that only 50.8 percent
of output had a microbiotic content below legal standards. Of
986 tons produced, 186 tons or 18.5 percent was referred for
repick. Thus nearly a fifth of factory picking effort was spent
on repicking operations. The management report noted that "this
high percentage is not explainable or acceptable by standard
manufacturing practices."(38)

However, problems of quality control date to the beginning

of new factory operations. For example, in 1976 four containers
of carrot flakes were sent to West Germany together with
satisfactory PVP laboratory quality-control test results.
Bruckner noted that "the control in the laboratory of our

137
customer shoved results whicn were really horrible. Not only the
total bacteriological r-unts were extremely high but there were
found such high countE f coliform germs and E-Coli that all the
carrots of the four conLainers were rejected."(39)

Raw Material Procurement Problems

The project's raw material procurement system was to be

based on "three legs." One leg was the smallholder farmers in
the Kinangop and elsewhere who would provide root crops
(primarily carrots) under rainfed production conditions. These
farmers would plant in April, May, and June for harvesting from
September until March. The second leq was to be Lake Naivasha
private farmers. They would supply specialist crops such as
beans and capsicums year-round while supplying root crops during
the Kinangop's off-season. The third leg would be company
estates on lind owned or leased by PVP. The estates would
concentrate on the specialist crops, but also do some root cropv.

Smallholder Contract Farming

The first leg, that of smallholder contract farmers, served

its function fairly reliably up until the project neared
financial collapse. During the 1970s the project expanded its
geographical scope of smallholder contracting bringing in
cooperatives as far north as Nyahururu and as far south as
Uplands. At one time or another some 30 cooperatives (or Forest
Department employee groups) were active in the project with as
many as 3000 farmers under contract.

A contract document between PVP and cooperative society

committees was prepaxed annually, and subsidiary agreements were
provided with each issue of seed. Seeds were provided on credit
to the cooperatives for distribution to members. Each farmer
taking seed made a written statement acknowledging his/her
receipt of seed and issuing a "guarantee" to supply the company
with a certain tonnage of produce. For carrots, this guarantee
generally varied from 5 to 10 tons per kilo of seed. A pre­
emergence herbicide, afalon, was used by some farmers. It was
provided on credit to the cooperatives and then sold to farmers.

Producer prices were decided at the beginning of each year

at meetings between the company and cooperative society
committees. These prices were then offered on a "take it or
leave it" basis to farmers. (40) The producer price consisted of
a basic rate and a bonus rate. The basic rate was paid for all
deliveries, subject to deductions for produce that was not first
quality. (See below.) Farmers delivering quantities at least as
large as their "guarantees" would then receive a lump sum payment
calculated by multiplying a bonus rate by the guaianteed

138
quantity. Crops delivered before reaching the tonnage guarantee
or accepted after the guaranteed quantity had been reached would
be payable at only the basic rate. The bonus rate was generally
40 percent or mor . of the basic rate. For example, in 1977 the
basic rate for carrots was Ksh 195/ton while the bonus was Ksh
80/ton.

The grading of crops was on the basis of acceptable material

delivered. Grade I consisted of 0-5 percent unacceptable produce
and the lull price was paid for this crop. Grade II consisted of
6-10 percent unacceptable produce. For these deliveries the
farmer would be docked for the weight of the reject material and
would receive 95 percent of the full price for the balance.
Produce was denoted es Grade III if 11-20 percent was
unacceptable. Farmers would be docked for the weight of reject
material and paid 90 percent of the full price for the balance.
Deliveries with more thin 20 percent of unacceptable material
were totally rejected and the owner was given the option of
having the delivery returned at his expense, collecting it
himself, or leaving the factory to dispose of it.

The company provided field assistants and placed them in

each major growing area. The field assistants were to work
closely with the cooperatives to ensure proper plantin~g and
cultivation, to determine the timing of harvests, and to organize
collection. They provided information to company management by
preparing monthly reports on individual production areas. These
reports provided information on seed distribution and planting,
use of herbicide, weather, incidence of disease, harvesting
patterns, demonstrations given, and the illegal sale of the crop
to alternative outlets. (41)

The farmers in the Kinangop had ample land and generally

grew vegetables in a shifting patter:. without the use of
fertilizers. Initial ploughing would be carried out by local
enterprises and paid for in cash by the farmer. Planting was
done during thE long rains (i.e., April-June) and harvests took
place over the September to March period. While carrots take 4-6
months to reach maturity, they can remain in the ground before
harvesting for up to nine months. Most field activities were
performed by family labor although some paid labor was used for
harvesting. The use of resistant seed varieties made it
unnecessary to apply insecticides or fungicides. On the basis of
an average yield of 10 tons of carrots per acre, the smallholder
farmer could expect a net profit of Ksh 1000 to 1500/acre. A
sample 1980 income estimation can be seen below:(42)

139
 Smallholder Production Cost and Income Per Acre(1980)

Seed Ksh 170

Ploughing 140
Harrowing 130
Sowing 40
ufalon (I kg) 94
Spraying 40
Hand Weeding and Thinning 400
Harvesting 500
Transport (Ksh 50 per ton) 50

Total Costs 2014

Income 10 tons @ Ksh 330 3300

(Includes Bonus)

Net Income Ksh 1286

Demand for seed in the rain-fed areas generally far exceeded

what the factory was prepared to issue to meet its requirements.
One key problem was to spread this requirement over an extended
period. In the early years of the project smallholder supplies
were heavily concentrated in only four or five months of the
year. PV? attempted to lengthen the smallholder supply season by
issuing seed supplies to the cooperatives in three phases over
April, May, and June. (43) During several years smallholder
deliveries were indeed extended over seven or eight months. For
example, during both the 1976-77 and 1977-78 seasons smallholder
vegetable supplies exceeded 1000 tons in each month from
September to March.

The contracts were theoretically legally enforceable

although in fact legal action was never resorted to. Instead,
PVP depended on close supervision and disciplinary action by the
cooperative committees. Field assistants also monitored the
progress of a crop and provided frequent estimates of the
standing crop and the crop being harvested. For cases where the
farme sold the produce on the fresh market, the cocperative
would issue fines. For persistent cases the farmers would not be
issued further seed. Where cooperative support was lacking,
contracts with the offending cooperatives were withdrawn. During
periods when produce "leakaqe" was very strong, police checks
were established to inspect trucks leaving the smallholder
areas. (44) The extent of "leakage" was partly controlled by the
selection of a particular carrot variety for processing. The
factory distributed seed of the red-cored Chantenay variety. The
Nantes variety, favored by the fresh market, was unsuitable for

140
processing. The Crsixtenay variety was not well liked on the
fresh market.

The extent of "lea;:age" differed by production area. Areas

close to Nairobi and well served by all-weather roads were more
vulnerable to "leakage" than areas far into the settlement
schemes having very poor feeder roads. During periods of glut
Nairobi traders avoided the Kinangop altogether as supplies were
sufficient from areas adjacent to Nairobi. Some parts of the
Kinangop had extremely poor feeder roads and the company needed
to hire tractors and even army trucks to collect produce during
the rains. Traders would normally avoid these areas.

Part of the attraction of the fresh market was price.

Raikes (1978) compared factory prices with those offered by
"lorry-traders" and found that while the factory's prices were
"marginally" higher during the peak season, they were as little
as 1/6 the market price during the off-season. There is no doubt
that during certain times of the year market prices were
considerably above those of the factory. However, it does appear
that the traders' buying procedures reduced the price advantage
of selling on the fresh market.

Carrots are sold by producers by the sack, which, when full,

should weigh 60 kgs. Generally the transporters would force
farmers to overload the sacks. Woven cord then held the surplus
produce in place. Both transport and market levies are charged
per unit container irrespective of weight, so t.4e traders have
the incentive to maximize the loaded weight of their containers.
It was not uncommon for carrots sacks to be overloaded by 100
percent. In fact, a 1983 survey found the average weighL of a
carrot sack brought to Nairobi's wholesale market was 103 kg. (45)
Thus, farmers were providing two sacks of produce while receiving
payment only for one sack. Considering this, the prices that the
farmers received from the factory may have actually exceeded
those for the fresh market over much of the main harvesting
season. 146)

Perhaps a more important advantage of selling to local

traderE- was the fact that farmers could avoid paying certain dues
and outstanding debts to their cooperatives. Some of these debts
were related to inputs for the vegetable project, while other
debts would have been related to the other services provided by
the cooperatives. Farmers could get ready cash in hand from the
traders while payment from the cooperative might have been
delayed until all "cooperative expenses" were covered first.
Delayed payments became more problematic over time due to the
worsening financial position of PVP.

141
 Still, in general terms the smallholder scheme generated a
fairly consistent flow of raw materials to the factory throughout
the late 1970s. Unlike for largeholder production, raw material
supplies from smallholders were not far below the long-term
production plan set out in the feasibility study. The following
figures represent the factory intake of carrots from the
cooperatives:
1976 8961 tons
1977 10281
1978 8195
1979 9141
1980 4849

Performance was generally good until 1980. In that year

various problems contribut-d to a considerable decline in
deliveries. One problem was that a large quantity of seed that
was provided to the farms was of poor quality and had low
germination rates. A second problem was that due to mechanical
faults and inadequate fuel supplies the factory was unable to
operate during part of the peak harvesting period. At the same
time PVP had inadequate working capital, and payments to farmers
were being delayed for several months. With the factory broken
down and with payments being delayed, some of those farmers who
did have a crop sold it to Nairobi traders. Vegetable producing
areas in close vicinity of Nairobi experienced a drought in 1980,
and market prices rose considerably. The smallholder scheme did
recover for the 1980-81 season with raw material deliveries
topping levels for several years in the late 1970s.

Large-scale Farmezs

The second eg, that of the Lake Naivasha farmers, never

fulfilled the company's expectations, and by the late 1970s raw
material supplies from this source had virtually ceased. As
early as 1976 it was becoming clear to PVP management that the
Lake farmers would not be a reliable source of supply and that
greater reliance would have to be put on the company developing
its own estates. (47) Horticultural production was expanding
around Lake Naivasha, but costs per acre had risen considerably
from a decade earlier. In addJtion to rising fuel costs (for
irrigation pumps) these farmers were facing rising aprochemical
costs. Furthermore, with the rapid development of the flower
sector and with B large nLmber of farmers going heavily into
labor-intensive french-bean production, a labor shortage existed
in the area and the cost of labor was rising.

Many of the Lake farmers who did not have large acreages
found that with normal yields it was only marginally profitable
for them to grow for the factory. Even a small reduction in

142
yields brought about by weather, nematodes, or disease would
result in losses. (48) These smaller farmers felt that it would
be useful to perhaps grow for the factory during the export off­
season. Alternatively, they were inclined to send their third­
quality produce to the factory after fresh produce exporters and
Nairobi greengrocers were provided first and second grades.
Neither of these two practices were acceptable to the factory.
The factory needed raw materials all year long and not simply
during three or four months. It was also impossible for the
segmented marketing procedure to work. The factory actually
needed first grades and applied its price discounts for any other
deliveries. In addition, the factory required particular
varieties and these were not the varieties preferred by the local
or export fresh market. For example, beans for export are mainly
the Monel or Nasterpiece varit ties which at an early stage
develop fibrous strings and are thus unsuitable for dehydration.
The Saxa or Contender varieties were rEquired by the factory.
Some indication of future trends was seen in 19'/6 when the Lake
farmers absolutely refused to grow leeks on the basis of the
prices and grading arrangements offered.

The factory did havE a different price structure for the

smallholder farmers growing under rain-fed conditions and the
larger farmers growing under irrigation, but due to its
accumulation of losses it was unable to increas_ the prices it
paid to the large growers. For most items there was absolutely
no price change between 1977 and 1979. By the latter yea­
factory prices had become well out of line with priceL for
comparable products in the fresh market. Compare below the
factory's prices with those offered by a leading exporter:

Producer Price Comparison (1979)

Item Factory (Ksh/ton) Exporter (Ksh/ton)

Beans 1020 5600
Leeks 515 2000
Chillies 500 2000
Capsicum 450 750

Some factors were clearly outside of PVP's control. This

can be seen in the cases of onions and capsicums. Growing onions
for the factory could not be economical given the very low yields
that are obtained for this crop in Kenya. No short-day white
onion variety of high solid content was available. At the same
time a protected market fox onions was being established by the
Horticultural Crops Development Authority in order to maintain
their statutory monopoly on onion wholesaling and in order to
protect smallholders in the Perkerra Irrigation Scheme whose
yields were less than half those of the Naivasha farmers. While

143
production licensing deterred some growers it encouraged others
to grow the product and find grey market outlets. (49)

Farmer deliveries of capsicums subsided with the dying

export traoe in this product. While peak export levels were
reached in 1972 at 1128 tons, thereafter competition from Spain,
Israel, and the Netherlands cut into Kenya's market, and exports
were down to just 333 tons by 1977. Many of the Italian farmers
at Naivasha who had grown capsicums reduced their plantings.
Given the factory's price structure, it was not economic for
these farmers to grow exclusively for processinq.

The Lake farms that were of larger size retained somewhat

more interest in growing for the factory. They were in a better
position to risk possible losses and in any case wanted to spread
their overhead costs among a larger cropped acreage. However,
when PVP's financial troubles prevented the company from
increasing producer prices in line with changing production
costs, some large farins found that they could get better returns
by growing for other Kenyan processors. Tomatoes sold to canning
companies generally brought better returns than various
vegetables sold to PVP.

Company Estates

The third leg, estate production, was required to compensate

for the declining supplies from the private Lake farms. Since
1970 it had become apparent that the factory could not hope to
operate effectively on a continuous basis without a nucleus farm
under its full control. Until the new project was started in
1975, funds had not been available for such a farm. In any case
it had been tne policy to depend on the support of local farmers,
particularly those in the Kinangop settlement schemes. While not
explicitly stated in the feasibility study, the development of a
nucleus farm was viewed as a central part of the new project's
crop-production component. (50)

Initially an agreement was entered into with Marula Estate

to place 400 acres at the full disposal of PVP with the option of
a further 400 acres. A contractor was hired to clear the land,
but at the last moment the owner withdrew from the agreement.
Alternative areas were sought. A plot of land between the
factory and Lake Naivasha, owned by the company, was brought into
use under irrigation by factory waste water. Although the soil
on this 200-acre plot was sodic and restricted in use and also
subject to flooding during heavy rains, the company found that it
could get a good leek crop from it. Two plots were leased at
Morendat. One section of 200 acres was already fully developed
and under irrigated lucerne, while Enother 200 acre plot was
developed with irrigation installed by PVP. One additional 200

144
acre plot was leased on the South Lake side. By 1978, the
company thus had a total of 800 acres of land under its direct
management.

During the late 1970s the factory's raw material procurement

from company-operated estates did increase considerably. In 1976
company estates provided 3185 tons accou.-ting for 17.3 percent of
intake. By 1980 company estates provided 7889 tons, accounting
for 60.7 percent of raw material supplies. These nucleus estates
operated at a continuous loss, however, and by the end of 1979
they had run up an accummulated loss of Ksh 5.8 million.

Part of the poor financial performance of the estates can be

attributed to the accounting prices offered by the factory. PVP
and its estates subsidiary operated separate accounts and issued
independent financial statements. The estates department
essentially absorbed some of the losses of the overall holding
company. (51) Farm production costs were estimated in 1976 and
accounting prices were set then. These prices remained constant
over the next five years despite changes in production costs. A
second factor that contributed to losses was the relatively poor
yields obtained on the farms. These low yields were attributable
partly to inadequate finances that caused problems for the timely
application of material inputs, and partly to the poor quality of
several plots of land. Over the 1978-80 period average bean
yields were 2.2 tons/acre while the estimated break-even point
was 3.7 tons/acre. For leeks actual production averaged 8.4
tons/acre while the break-even level was 11.3 tons/acre. (52)

A third factor contributing to the poor financial

performance of the estates was their excessive productio costs.
An FAO mission examining the finances of the estates found
excessive costs attributed to permanent employee salaries that
were increasing production costs per crop area by as much as 20
percent.

The estates department reacted to the financial losses and

the low factory prices in a rational way. The estates began
selling increasing quantities of their vegetables on the fresh
market. In 1980 comparisons of estate break-even points for
sales on the fresh market ver;-- sales to the factory (at
accounting prices) found that factory prices as a proportion of
average fresh market prices were as follows: beans 35 percent;
tomatoes 30 percent; onions 33 percent; cabbage 40 percent, leeks
30 percent. In 1979 and 1980 estate sales to the fresh market
totaled 15 percent and 19 percent of quantities sold and 34
percent and 42 percent of revenue earned. In 1980 the estates
sold 2400 tons of vegetables on the fresh (local and export)
market at an average price of Ksh 1100/ton, which was 130 percent

145
higher than the average accounting price of Ksh 420 for factory
intake.

In 1981 two of the leases held by the company were

discontinued by the landowners. By this time the price of lnd
around Naivasha had begun to rise rapidly as Kukuyu farmers uho
had greatly benefited from the tea and coffee price booms of the
late 1970s were seeking farms at Naivasha. The company could
neither purchase nor lease land at a cost that could be recovered
by growing vegetables for the factory.

As .he company could not afford to pay the commercial rates

for land at Naivasha it sought to have the Commissioner of Lands
compulsorily acquire 2000 acres from the European-owned Marula
Estate on the strength that the farming activities constituted a
public use. A High Court ruling went against the company's
position. A proposal was later submitted whereby the Ministry of
Agriculture would provide the company with at least 1000 acres
currently being used by the Naivasha Livestock Research Unit and
then acquire the 2000 acres from Marula Estate, since livestock
x-.search falls within the definition of public use. After the
acquisition the Ministry would provide an additional 1000 acres
to the company. The Ministry rejected these proposals and could
not provide any land from the research station.

Concluding Remarks

We have reviewed various features and historical segments in

the development of Kenya's vegetable dehydration industry and
have related these to changing conditions in both domestic and
international markets. The industry was born largely in the
pursuit of social and political objectives related to the
smallholder settlers in the Kinangop. While the project received
strong political backing in its early years, it lacked strong
technical management and an economically viable production
program. The industry was thus unable to take advantage of the
expanding West European market for dehydrated vegetables.

The industry's transition into economic viability was

increasingly seen to depend upon the injection of international
capital and the involvement of a multinational firm with
technical and marketing expertise. A marriage, worked out
between Government investment, international loans, and
multinational management, appeared to provide an optimistic
future for the industry.

A major assumption of the reformulated project was that it

would be economically rewarding for the Naivasha farmers to grow
vegetables for the factory. In fact, it was supposed to be the
large farmers who would play the key role in raw material

146
production. Changes in factor and commodity markets combined
with the factory's grading standards and inability to raise
producer prices, however, made sales to the factory increasingly
unattractive for these farmers.

The collapse of large farmer supplies considerably

undermined the viability of the project. The factory was
operating at well below its capacity, with overhead costs thus
pushing up unit production costs. The company's product mix was
dominated by the low value root crops grown by the smallholders.
In response, the company needed to dc.4elop its own estates. In
this effort there were several constraints. With the Naivasha
area experiencing increased horticultural production, less land
was available and at a higher cost. The plots of land that the
company could operate were not of first-rate quality. Company
financial problems constrained the farm-level activities of the
estates department. Government political backing for estate
acquisition remained weak as an incipient fear that estate
production would marginalize the smallholder farmers prevailed.
The extent of political opposition to estate production requires
further study.

Thus, the raw material component of the project remained

problematic. Only the smallholder contract production scheme
provided fairly reliable supplies. With the available
information we have been able to examine in only general terms
the organization of production, the form of contract, and a few
indicators of performance. We have not been able to examine the
evolution of the contractual structure and Lhe changing coles of
the company and the cooperatives in supporting producers and in
enforcing the contracts. Horticultural cooperatives have not
generally been successful in Kenya, least of all in relation to
contractual arrangements with processors or exporters. It would
be useful to understand the wider relationship between the
farmers and the cooperatives (i.e., for milk) in order to
discover reasons for the apparently useful role of the
cooperatives in the case of the dehydrated vegetables project.

It would also be interesting to compare the services

provided by PVP and the cooperatives with those offered by the
"lorry-traders" in order to gain a better understanding of the
"leakage" issue. Was price the main factor? Was the escape from
cooperative dues or delayed payments a more important incentive?
Was the provision of technical services important tc the farmers?
Presently we have only limited information on farmer yields and
incomes. It would be useful to examine in greater detail actual
farmer yields and their variance by area and over time. One
would also wish to place the income earned in the vegetable
project within the context of the wider sources of income for the

147
smallholder farmers and to trace the uses to which this income
was put.

In this study we have noted some of the problems associated

with the processing operations. Again, limited information has
been available. Was the problem largely one of management? Was
any of the equipment inappropriate for the tasks being
undertaken, or were inefficiencies in operating procedures and
poor maintenance of equipment the dominant factors? Why couldn't
quality or conversion-rate results be improved through the
provision of technical assistance by Bruckner?

We have examined aspects of the marketing arrangement

between Bruckner and PVP, including the conflict over the issue
of price. Relatively little information was provided about the
marketing procedures and terms of trade that characterized the
1964-1972 period before a major multinational company was
involved. Access to relevant material could provide for an
interesting comparison with the later period. While it does not
appear that Bruckner generally paid prices to PVP which were "too
low," it would be interesting to examine further the general
nature of PVP's relationship with Bruckner. Particularly
important issues would be information flows, alternative sales
outlets, payment procedures, and qualit control and claims.

Our analysis of the Kenyan dehydration industry has shown

the critical links between production, processing, and marketing,
in any export operation. It demonstrates that contract farming
schemes should not be examined in isolation from world market
patterns for the final product or from changes in domestic factor
and product markets. Although apparently successful itself, the
smallholder contracting scheme was first undermined and then
terminated entirely because of operational inefficiencies in
other components of the project.

Notes

1. Philip Raikes(1978) obliquely discusses the project, but is

most interested in espousing a general argument about the
relationship between international capital and "middle peasants."
He provides few actual facts about the project. Dinham and Hines
in their book Agribusiness in Africa (1983) mention the project
as an example of collaboration between governments and
multinational companies in the development of "new luxury crops."
They provide a few facts about the case, drawing basic
information from a few issues of the government magazine, Kenya
Export News.

148
2. Tropical Products Institute, p.7.

3. Ibid.

4. ITC 1981, p.ll1;50.

5. Ibid., op cit.

6. Tropical Products Institute, p.8.

7. ITC 1981; Interview with David Hirst, former Agricultural

Manager of the dehydration project (October 17, 1986).

8. SIFIDA (1973), p. 2 2 .

9. David Hirst interview.

10. SIFIDA, p.22.

11. Some farmers initiated vegetable production during World War

II in order to supply the prisoner of war camp at Naivasha which
was holding Italian soldiers.

12. Interviews with Naivasha farmers including Dorian Rocco and

Cesare Bellyngeri.

13. Ministry of Agriculture (1981), p.1; SIFIDA, p.22.

14. SIFIDA, p.22.

15. Calculated from data in SIFIDA, p.17.

16. Data in SIFIDA and from the Horticultural Crops Development

Authority.

17. Information drawn from SIFIDA, Ministry of Agriculture

(1970), and East Africa Excise and Customs data.

18. ITC (1972).

19. This discount cannot be accounted for by Kenya's "low costs."

Several high wage countries had supply prices below the average
price while a low wage country such as China had supply prices
above those of Kenya.

20. As discussed in Ministry of Agriculture correspondence on the

project and reported by Makanda (1986) pp.17-24.

21. ITC 1972.

149
22. Letter from PVP Managing Director D.M. Watene to PVP Board
of Directors dated May 29, 1976.

23. Financial data drawn from FAO (1981) and Ministry of

Agriculture (1981).

24. Marketing In Europe, April 1976.

25. Marketing In Europe, January 1984.

26. ITC 1981.

27. However, even slightly declining D-Mark import prices should

not have adversely affected the company's ability to repay its
loans. The company's foreign currency loans had been denominated
in US$. During the 1970s the dollar devaluated against the Mark
by approximately 50 percent and in the second half of that decade
the level of devaluaticn was about 26 percent. In 1970 $1=3.68
DM. The rate for 1975 was $1=2.46 DM and in 1979 $1=1.83 DM.
Thus, in the second half of the decade even if prices in DM did
not show a favorable trend, their dollar value certainly did as
seen below:

US$ Equivalent of West German Import Prices (Per kg.)

1975 1976 1977 1978 1979

Carrots 1.54 1.48 1.65 2.00 2.19

Leeks 1.61 1.46 1.86 1.98 2.31

Source: Calculated from data in ITC 1981; IMF

International Financial Statistics, 1980

28. This is the general impression provided by correspondence

between Bruckner and PVP management in the late 1970s.

29. "We get no information whatsoever on the prices paid by the

end-user; such information is vital in conducting price
negotiations with Mr. Bruckner since it would enable us (to) take
advantage of favorable demand conditions." This statement was
made in a PVP management memo dated December 20, 1976 entitled
"Management and Marketing Agreement: Main Aspects Requiring
Review."

30. Ibid; also KETA 1981.

31. Draft Cooperative Agreement Between Pan African Vegetable

Products and Brueckner-Werke KG.

150
32. Letter from H. Glockner (Bruckner) to General Manager Watene
(PVP) dated September 15, 1978.

33. Ibid; also letter from Glockner to PVP on December 8, 1977

concerning quality claims.

34. See note 32 above.

35. Letter of November 14, 1980 from General Manager H.A. Odour
of PVP to the Chairman of PVP's Board of Directors.

36. "A Review of Bruckner Werke's Sales Contxicts" dated June 21,
1978 and carried out by PVP's Financial and Administrative
Director. It examined contracts over the 1976--78 period.

37. Correspondence with David Hirst; Also a document called PVP

Review 1980 showed conversion rates for most items considerably
less than "standard" rates.

38. PVP Review 1980, p. 3 1.

39. See note 33 above.

40. Correspondence with David Hirst.

41. Included in field assistant reports were discussions of

cooperative committee meetings that had taken place and what was
said about PVP at these meetings.

42. Revised from an estimate found in PVP records. The PVP

calculation was based on a yield of 14 tono/acre. Davi.d Hirst
reports that 10 tons/acre was more the norm.

43. Raikes (1978) contends that because farmer yield was affected
by the time of planting, the company had "an intermediate form of
control over the producers through the ability to reserve "prime­
time" contracts to producers who are "cooperative. " Raikes
admits that he has no evidence for this "but it is almost certain
that the extension agents of the company, who implement the
regulations, do so to their own benefit even if the company does
not." I have come across no evidence for this discrimination in
seed distribution and this issue is not one that farmers have
raised when rendering complaints about the project.

44. Interview and correspondence with David Hirst.

45. "An Analysis of all Fruits and Vegetables Sold at Wakulima

Wholesale Market During 1983" FAO/Ministry of Agriculture.
Horticultural Marketing Proj.

151
46. It should be noted that farmers sometimes complained that
company staff were underweighing produce at times and taking part
of the supplies for their own use or sale.

47. See note 22 above.

48. "Minutes of Meeting at PVP with Naivasha Farmers" dated May

1, 1976.

49. This paragraph is based on FAO (1981) and interviews with

Hirst and with Nalvasha farmers.

50. Correspondence with David Hirst.

51. Argued in FAO (1981), financial section.

52. Ibid.

Selected Sources

FAO/Bankers' Programme Investment Centre. "Rehabilitation of Raw

Material Production in Pan African Vegetable Products Ltd."
Naivasha. Project Brief. 1981.

Government of Kenya. Report on a Working Party on the

Horticultural Industry in Kenya. 1970.

International Trade Center, UNCTAD/GATT. The Market for

Dehydrated Vegetables. Geneva. 1972.

The Market for Dehydrated Vegetables in Selected West

European Countries, the United States, and Japan. Geneva.
1981.

Kenya External Trade Authority. "Mission to Seven Selected

European Countries for PAVP by the Industrial and Commercial
Development Corporation and KETA." March/April 1981.

Makanda, D. W. "Contract Farming in Kenya's Horticultural

Industry: A Case Study of Pan African Vegetable Foods Ltd."
Nairobi, 1986.

Marketing In Europe, "Canned and Packet Soup in Germany," April

1976 "Canned and Packet Soup in the Netherlands," Jan 1984
"Canned Soup in West Germany," July 1984.

Ministry of Agriculture. "Current Status and Support to Pan

African Vegetable Products," 1981.

Raikes, Philip. "The Development of a Middle Peasantry in Kenya:

Discussion of Some Hypotheses and Examples from the Dairy

152
 and Horticultural Sectors," Center for Development Research.
1978.

SIFIDA. Feasibility Study for the Establishment of a Vegetables

Dehydration Factory at Naivasha, Xenya. June 1973.

Tropical Products Institute. The Market for Dehydrated

Vegetables with Particular Reference to Selected West
European Countries. November 1971.

153