Documente Academic
Documente Profesional
Documente Cultură
and
Horticulture Development Center
Projects in the Tree Fruit Value
Chain in Kenya
Baseline Research Report
microREPORT #33
JULY 2005
This publication was produced for review by the United States Agency for International Development. It was prepared by
Don Snodgrass and Jennefer Sebstad with Action for Enterprise under the Accelerated Microenterprise Advancement Project
(AMAP) Business Development Services (BDS) Kenya STTA Task Order. Action for Enterprise is a subcontractor to
ACDI/VOCA under the AMAP BDS IQC.
DISCLAIMER
The author’s views expressed in this publication do not necessarily reflect the views of the United States Agency for Interna-
tional Development or the United States Government.
Accelerated Microenterprise Advancement Pro-
ject (AMAP) is a four-year contracting facility that
USAID/Washington and Missions can use to acquire
technical services to design, implement, or evaluate
microenterprise development, which is an important
tool for economic growth and poverty alleviation.
1
ACDI/VOCA is a private, non-profit international de-
velopment organization based in Washington, DC.
2
Assessing the Impact of Kenya
BDS and Horticulture Develop-
ment Center Projects in the
Tree Fruit Value Chain in Kenya
2005
3
CONTENTS
EXECUTIVE SUMMARY............................................8
REFERENCES..........................................................63
4
LIST OF TABLES AND FIGURES
TABLES
5
Table 27: Proportion of Household Income from Tree Fruit (Estim-
ated)
Table 28: Proportion of Household Income from Tree Fruit (Estim-
ated) by Intervention
Table 29: Average Monthly Consumption Expenditure per Capita
by Intervention (Ksh)
Table 30: Distribution of Respondent Households by Monthly Consumption
Expenditure per Capita
Table 31: Distribution of Respondent Households by Asset Score
Table 32: Distribution of Respondent Households by Asset Score
and Intervention
Table 33: Distribution of Respondent Households by Size of Land-
holding
Table 34: Distribution of Respondent Households by Size of Land-
holding by Intervention
Table 35: Percentage of Woman-headed Households in Study
Sample
Table 36: Selected Household Data by Gender of Household Head
Table 37: Selected household data by asset score group (total
sample)
FIGURES
6
LIST OF ACRONYMS AND ABBREVIATIONS
AGOA African Growth and Opportunity Act
BDS Business Development Services
EAGA East Africa Growers’ Association
EPZ Export Processing Zone
EU European Union
EUREPGAP Euro-Retailers Produce Good Agricultural Practices
FPEAK Fresh Produce Exporters Association of Kenya
GOK Government of Kenya
HCDA Horticultural Crop Development Authority
HDC Horticulture Development Centre (Fintrac)
HIV/AIDS Human Immunodeficiency Virus/Autoimmune Deficiency Syndrome
IR Intermediate Result
KACE Kenya Agricultural Commodities Exchange
KADI Kamurugu Agricultural Development Initiatives
KARI Kenya Agricultural Research Institute
KHE Kenya Horticulture Exporters
Ksh Kenya Shilling
KWETU Swahili for “our home”, name of a local non-governmental organization
MOA Ministry of Agriculture
MOU Memorandum of Understanding
MSEs Micro and Small-Scale Enterprises
NGO Non-Governmental Organization
SITE Strengthening Informal Sector Training and Enterprise
SMS Short Message Service
SO Strategic Objective
USAID United States Agency for International Development
7
XECUTIVE SUMMARY
THE PROJECTS AND THE the project issued tenders and holder associations, and two
IMPACT ASSESSMENT awarded contracts to eight small businesses producing
private sector and NGO part- plant stock. In the future, it in-
This study assesses1 the im- ners. The contracts were de- tends to work through input
pacts of two USAID/Kenya fun- signed to facilitate the devel- suppliers as well. This five-
ded projects that focus on de- opment of sustainable solu- year project began in late
veloping sustainable solutions tions/services that provide ma- 2003 and was in its first year
to constraints facing busi- terial inputs (agro-chemicals of operation at the time of the
nesses in targeted industries and seed varieties), appropri- baseline survey.
and the degree to which these ate technology to upgrade
solutions impact: products and production pro- The study is longitudinal, with
the competitiveness of the cesses, business and skills a baseline study including
mango, passion fruit, and training, and extension and in- both quantitative and qualitat-
avocado value chains; formation services. Several ive research. This will be fol-
contracts promote market link- lowed up in two years with a
the integration of micro
ages between smallholder pro- resurvey of the same respond-
and small enterprises
ducers and lead firm exporters ents as well as further qualitat-
(farmers and others) into
through supply contracts and ive research. The major find-
these value chains so that
lead firm provision of embed- ings on the impact of the two
they contribute to and be-
ded services, and encourage projects will emerge after this
nefit from the tree fruit in-
inter-firm cooperation through second stage of research.
dustry’s increased com-
organization of producer
petitiveness
groups and provision of em- This report presents the find-
bedded services. ings of a baseline study that
The projects are the Kenya
featured a survey of 1,947
Business Development Ser-
The HDC project focuses on a smallholder farmers who grow
vices (BDS) project implemen-
wide range of horticulture avocado, mangos, or passion
ted by the Emerging Markets
products. We study only their fruit in Central, Eastern, and
Group and Fintrac’s Horticul-
passion fruit work, which in- Rift Valley provinces. The
ture Development Centre
cludes plans to: (1) introduce sample included farmers who
(HDC) project. Both projects
new varieties of passion fruit are participating in the two
support USAID/Kenya’s stra-
for fresh export; (2) improve projects as well as a control
tegic objective of increasing
agricultural practices of local group of non-participants. The
rural household incomes in
producers; (3) expand local survey was complemented by
Kenya (SO 7). They seek to
processing capabilities for loc- qualitative research (in-depth
raise smallholder productivity,
al market products; and (4) interviews and focus group
widen market outlets, facilitate
strengthen the farm-to-market discussions) with over 60 indi-
vertical and horizontal link-
value chain, inclusive of busi- viduals involved in the tree
ages, and promote the sus-
ness services to small farmers. fruit value chain, including
tainable development of busi-
Unlike Kenya BDS, the HDC farmers, farmer producer
ness services for rural MSEs
project does not operate group leaders, input suppliers,
through contracts but carries extension workers, brokers,
Kenya BDS, a five-year project,
out activities directly through exporters, and Kenya BDS and
started in 2002 and worked
project staff based in Nairobi Fintrac project directors and
initially on the tree fruit value
and agronomists based in four staff.
chain. In its first two years,
field offices. It works with and
1
“Value chain” and “sub-sector” are used through cooperating partners,
synonymously throughout this report. including KARI, existing small-
8
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
The study design is based on a Hypotheses 3: Greater integ- local supermarkets, and ex-
causal model of impact that ration of smallholder MSEs into porters. Exporters buy fruit
shows how project facilitation productive value chains con- produced by smallholders and
activities to promote sustain- tributes to improved competit- medium-scale farmers and
able solutions can address iveness and growth of the tar- also produce some fruit on
constraints to smallholder par- geted value chains. their own plantations. Export-
ticipation and the competitive- ers have just recently begun to
ness of the tree fruit value buy directly from producer
chain2. These activities in turn FINDINGS OF THE groups and to provide embed-
lead to sustained access to the BASELINE STUDY ded services to smallholders
solutions, smallholder upgrad- through these groups.
ing, increased smallholder MSE
Tree Fruit Market
profits from tree fruit activit- Constraints to smallholder par-
Avocados, mangos, and pas-
ies, increased rural household ticipation in the tree fruit value
sion fruit are among the most
incomes, and overall sector chain include:
common fruit crops in Kenya.
growth and competitiveness Lack of information and
Most of the fruit produced is
within the value chain. knowledge of the markets
sold in the domestic market,
but all three fruits are import- Limited access to inputs
The study tests three hypo- Limited access to re-
ant and growing export crops.
theses about the impact of sources for, and/or weak
In Europe, Kenyan fruit has a
donor interventions in opening incentives for, upgrading
competitive advantage based
up opportunities for smallhold- Weak vertical and hori-
not on volume, quality, or
er MSEs in local, regional, and zontal linkages within the
price, but rather on seasonal-
global markets and in improv- value chain
ity. Avocados, mangos, and
ing the competitiveness of the
passion fruit each have a ‘win- Lack of trust among pro-
overall value chain:
dow’ when these crops are ducers, brokers, and ex-
less available from other sup- porters
Hypothesis 1: Project activit-
pliers. Kenya is also better set
ies to promote sustainable
up to meet certification stand- Governance in the tree fruit
solutions in the tree fruit value
ards than other countries. value chain is characterized by
chain contribute to better in-
a mix of market and network
tegration of smallholder MSEs
The main marketing outlets for relationships (see page ___ for
into the value chain.
tree fruit producers are traders more detail). Smallholders
and brokers, who in turn sell to have traditionally sold their
Hypothesis 2: Better integra-
both domestic and export mar- fruit to brokers on a spot basis;
tion of smallholder MSEs into
kets. In the domestic market, contractual relationships have
the tree fruit value chain con-
they sell to wholesalers, fresh been marked by distrust. With
tributes to enterprise upgrad-
fruit retailers, and small retail increasing concentration
ing, improved performance,
shops. Producers also sell dir- among European buyers and
and enhanced household well-
ectly to fresh fruit retailers and rising standards in end mar-
being.
the Horticulture Crop Develop- kets (especially Europe, but
ment Authority (HCDA). also in other international, re-
2 Brokers and traders are the gional, and domestic markets),
Sustainable solutions here refer to more the power of the retailing
main conduits for smallholders
than business services from third party
to formal and informal pro- groups to impose governance
providers - they also include sustainable
access to markets, business relationships, cessing plants and to export- rules on the value chain is in-
TA provided in an embedded fashion from ers who buy tree fruits. Medi- creasing. Horizontal linkages in
one firm to another, improved business en- um-scale growers often link the form of farmers’ associ-
vironment, capacity of industry represent- directly to processing plants, ations exist but need strength-
atives to influence policy, etc.
9
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
ening. The horticulture value accepted for farmers selling household income sources
chain has had limited govern- avocados to EAGA under the suggest an active working
ment involvement and private Kenya BDS project and those population among respond-
firms have generally been left selling passion fruit in the Fin- ents. There are no major differ-
free to organize the trade. This trac areas; remaining groups ences in earner-dependent ra-
differs from the pattern that sold their fruit predominantly tios between men and women
characterizes some other com- in spot markets. headed households or by
modity value chains in Kenya wealth level (as indicated by
-- for example, coffee, tea, and Hired labor was used fairly ex- asset scores), suggesting that
pyrethrum, for which official tensively by richer farmers, this may not be a major de-
marketing boards still control while poorer farmers relied terminant of vulnerability for
procurement and prices. As primarily on family labor. Wo- households in the sample.
producer groups form to link to man-managed farms tended to
inputs and markets, and as ex- hire more labor than compar- The asset scores and con-
porters form associations, the able farms managed by men. sumption expenditure data
patterns are shifting more to- show a significant number of
ward network relationships. Producer group membership poor households in the
was almost universal among sample, in both the participant
Tree Fruit Enterprises program participants, both and control groups. This sug-
We surveyed five interventions male and female. Moreover, gests the projects are in-
intended to promote upgrad- nearly all of the farmers who volving poor households and,
ing and raise productivity and belonged to producer groups thus, have potential for direct
income from tree fruit among characterized them as either impact on their income from
smallholder producers of avo- very or fairly useful. tree fruits. The sample also in-
cado, mango, and passion cludes non-poor households,
fruit. The MSEs included in the Few farmers had access to ir- which should provide a good
survey cultivated varying num- rigation and less than one-half basis for comparing impacts
bers of trees/vines, with avo- purchased fertilizer for use on across poverty groups at the
cado holdings the smallest on their fruit trees. A larger num- end line.
average and passion fruit the ber said they had bought
largest. For each fruit, the pesticide or fungicide sprays. Households are quite diversi-
range of holding sizes was fied in their sources of income
wide. With one exception, pro- Considerable numbers of re- and tree fruits are an import-
duction and productivity were spondents had instituted im- ant source. While these figures
higher for program parti- proved cultivation or market- may reflect an upward bias in
cipants than for controls, dif- ing methods in the past two some respondents who associ-
ferences that may reflect se- years. Large numbers in some ated the study with the tree
lection bias and/or early im- areas had planted fruit trees in fruit projects, it suggests the
pacts of program participation. the past year. Farmers looked importance of relatively small
Between the two passion fruit to a wide range of sources for amounts of cash income for
sites, Fintrac works with larger useful technical advice, in- rural households.
farmers. formation, or training.
Gender differences in the divi-
Nearly all the farms surveyed Tree Fruit Households sion of labor related to tree
sell tree fruit, primarily Household size in the sample fruit production, the control of
through traders of different is large relative to the total tree fruit income, and access
sorts, but most earn only small population, but about average to productive resources are
amounts from these sales. for poor rural households. The likely to play out in the impact
Contract sales have become number of earning members in of the projects. Producer
dominant and relatively well households and the number of groups appear to be an effect-
10
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
ive means of reaching women It is too soon to tell whether
and poorer tree fruit farmers. To varying degrees, the five in- the project activities will result
terventions included in the in "sustainable solutions" to
Role of the Projects baseline study succeed in the recurrent needs of tree
Both projects facilitate activit- reaching low-income farmers. fruit producers. This includes
ies to promote upgrading of This means that there is po- both embedded and stand-
tree fruits, primarily product, tential for direct impact by alone solutions/services that
process, and inter-chain up- raising rural household in- provide inputs, TA, or market
grading. The aim is to improve comes through the projects. access. In some cases changes
the capacity of smallholders to might take place due to direct
respond to changing market Building stronger horizontal provision by the projects but it
demand and increase rural in- linkages by grouping produ- remains to be seen if embed-
comes. The baseline research cers and achieving economies ded service arrangements, the
identified specific forms of up- of scale is an important part of commercialization of nursery
grading in the tree fruit value this potential because it helps and extension services, or the
chain and polled the views of poor farmers link to export “network broker” concept of
producers and other actors in markets – something they EAGA and Kenya BDS will last
the value chain on the incent- have very little opportunity to once the project activities end.
ives and disincentives to up- do by other means. The pro-
grade. All the sub-projects in jects have been instrumental While scrupulous efforts were
the study promote the forma- in organizing and strengthen- made to select control group
tion of producer groups as part ing tree fruit producer groups. samples for the baseline sur-
of their strategy to link small- vey that were comparable to
holders to input, service, and Vertical links to higher-value the participant samples, at the
product markets. At the time markets provide critical incent- time of the survey the parti-
of the qualitative research, ives for tree fruit producers to cipants as a group were signi-
producers groups had been upgrade. So far, only one of ficantly better-off and more
formed, but most of them (ex- the interventions studied – the productive than the controls.
cept the avocado groups) were EAGA avocado intervention – When each group is resur-
still at an early stage in their has begun to realize this po- veyed two years hence, care
actual activities. tential by forging a direct link will need to be taken in analyz-
from farmers to the European ing the results to ensure that
Baseline Research Conclu- market. This linkage has in- differences in household
sions and Implications for volved the provision of embed- wealth and other mediating
Round Two ded spraying services by the variables are taken into ac-
Smallholders are part of the exporter and negotiated MOUs count in determining the im-
tree fruit value chain, but they between producer groups and pact of the programs.
occupy a low position within the exporter. The process has
that chain. They are numerous required considerable “hand In the second round it will be
and active producers, but their holding” by Kenya BDS and crucial to review and docu-
productivity is low and they other support from USAID to ment the interventions care-
sell much of their produce un- help prepare smallholders to fully. The activities are very
der unfavorable conditions. In- meet EUREPGAP standards. different and the scopes of
come from tree fruits plays an their activities and the ap-
important role as a source of Brokers remain alive and well proaches they take are likely
household income, especially in all three fruit value chains to evolve over time. Finally, it
for the poorer farmers, but in- and continue to be important will be important to analyze
come from tree fruit and total marketing channels for many the commercialization issue,
household income are both farmers. including a careful look at the
very low in most cases. specific services/solutions pro-
11
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
moted during the course of the
projects.
12
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
I. INTRODUCTION
cipants in this activity, and A. MAIN FEATURES OF
This report presents the find- provide some preliminary in- THE PROGRAM
ings from a baseline study of dications of what impacts the ENVIRONMENT
the impact of two projects to projects may be having.
develop tree fruit value chains 1. DEVELOPMENT IN
in Kenya. The study featured a
KENYA3
survey of 1,947 smallholder
farmers who grow avocado,
Kenya achieved independence
mangos, or passion fruit in
from Great Britain in 1963 fol-
three provinces of Kenya –
lowing a nationalist struggle.
Central, Eastern, and Rift Val-
Significant economic growth
ley. The sample included farm-
was achieved through the
ers who are participating in
1970s, but growth slowed in
the Kenya Business Develop-
the 1980s and per capita in-
ment Services project (imple-
come declined in the 1990s
mented by the Emerging Mar-
under the dictatorial rule of
kets Group, formerly known as
Daniel Arap Moi. Hopes of na-
Deloitte, Touche, Tohmatsu)
tional revival were raised by
and Fintrac’s Horticulture De-
the free election of 2002,
velopment Centre project, as
which brought to power a gov-
well as a control group of non-
ernment headed by Mwai
participants. The survey was
Kibaki. Yet the economic
complemented by qualitative
growth rate remained low and
research involving in-depth in-
per capita income in 2003 was
terviews and focus group dis-
still below the 1990 level. Pre-
cussions with over 60 individu-
dicted economic growth rates
als involved in the tree fruit
have recently been revised up-
value chain, including farmers,
ward to 3.3 percent in 2005
farmer producer group lead-
and four percent in 2006,
ers, input suppliers, extension
based on accelerated dis-
workers, brokers, exporters,
bursement of donor funds,
and Kenya BDS and Fintrac
strong performance by cash
HDC project directors and
crops and tourism, and rising
staff.
garment exports to the U.S.
under AGOA. These rates re-
This baseline study will be fol-
flect improvement on past per-
lowed up in two years with a
formance but remain far from
second survey of the same re-
the six percent annual growth
spondents as well as further
that the World Bank believes
qualitative research. The ma-
to be achievable if planned re-
jor findings on the impact of
forms are implemented in full.
the two projects will emerge
after this second stage of re-
Slow economic growth has
search. In the meantime, the
combined with HIV/AIDS to
present report will describe
smallholder tree fruit cultiva- 3
References for this section include the
tion in Kenya, note some of CIA World Factbook, the Economist Intel-
the characteristics of parti- ligence Unit, USAID, and the World Bank.
13
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
cause poverty to rise and Rapid population growth age of 1.8 hectares in the
health conditions to worsen. HIV/AIDS rainy season. 5 Just over one-
The poverty headcount in- Low levels of investment half of farmers have a deed to
creased from 49 percent of the Inefficient and dilapidated their land, while another one-
population in 1990 to more infrastructure third own the land but have no
than 56 percent in 2003. Life formal title. Many farms lack
Vulnerability to drought
expectancy fell from 57 years good access to markets. Close
Threats to Kenya’s ex-
in 1986 to 45 years in 2004 to one-half are located within
traordinary environment4
while the infant mortality rate five kilometers of a paved
rose from 63 per live births in road. The average farm house-
The government that came to
1990 to 78 in 2002. The estim- hold has 6.8 members and is
power in 2002 pledged to ac-
ated HIV/AID prevalence rate headed by a 53-year old. Male
celerate economic growth and
is currently 6.7 percent. household heads (86 percent
reduce poverty. To this end, it
of the total) average six years
formulated a poverty reduction
High fertility and rapidly in- of schooling, woman house-
strategy, known as the Eco-
creasing population have com- hold heads four years.
nomic Recovery Strategy for
pounded Kenya’s economic
Wealth and Employment Cre-
problem. The population grew
ation, and committed itself to
from 9.4 million at the time of
shift public expenditure to-
independence to 31.9 million
wards programs that benefit
in 2003, averaging more than
the poor, notably a free
three percent per annum.
primary education program.
However, the total fertility rate
Subsequently, however, polit-
has now declined to 3.3 births
ical in-fighting over constitu-
per woman and the population
tional reform and other issues,
growth rate in 2004 was only
together with signs of reluct-
1.1 percent.
ance to tackle high-level graft,
have raised doubts about the
Kenya has many natural ad-
government’s ability to reform
vantages as well as the largest
and shake off the bad habits of
and most diversified economy
the past. Kenya is heavily de-
in the East Africa. According to
pendent on donor funding,
USAID/Kenya, the country’s
which was withheld during the
perennial failure to achieve
1990s and is currently
sustained economic growth is
threatened once more by the
attributable to several factors:
governance issue.
Governance issues: lack of
democracy; over-concen-
Poverty and inequality remain
tration of power in the ex-
severe, largely because most
ecutive branch with inad-
Kenyans are still low-pro-
equate checks and bal- 5
Data cited in this paragraph derive from
ductivity farmers. Agriculture
ances the Rural Household Survey carried out by
absorbs 75 percent of the
Corruption that pervades labor force but produces less Egerton College, Tegemeo Institute, and
public administration than 20 percent of GDP. Farms
Michigan State University in 2000, as re-
Inconsistency in policies, ported in Nicholas Minot and Margaret
are small on average. Most Ngigi, “Are Horticultural Exports a Rep-
laws, and regulations that farmers are heavily dependent licable Success Story? Evidence from
adds significantly to costs on rainfall and plant an aver- Kenya and Cote d’Ivoire.” Paper presented
of doing business and dis- at the InWEnt, IFPRI, NEPAD, CTA con-
courages investment 4
USAID/Kenya. 2000. Integrated Stra- ference, “Successes in African Agriculture,
Low productivity tegic Plan 2001-2005, pp. ii-v. Pretoria, December 1-3, 2003.
14
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
2. USAID KENYA’S STRA- fore heavily dependent on (USAID/Kenya 2000, pp.
TEGIC OBJECTIVES its agricultural productiv- 65, 83)
ity. Over the past decade,
Based on U.S. foreign policy in- however, agricultural pro- Kenyan agriculture is primarily
ductivity has declined and organized in smallholdings and
terests and Kenya’s develop-
poverty has increased. is almost exclusively rain-fed.
ment constraints, USAID/Kenya
identified four strategic object- According to the Mission’s ana-
Over the same period,
ives and one special objective drought has plagued lysis, factors contributing to
for its programming in 2001- Kenya on an increasingly low and falling agricultural pro-
2005: frequent basis, affecting ductivity include HIV/AIDS, a
Strategic Objective 6: ‘traditionally’ drought- confused policy environment,
Sustainable reforms and prone areas, as well as the survival of marketing
accountable governance many other agro-ecologic- boards for a few key commod-
al zones (AEZs) of the ities (coffee, tea, and pyr-
strengthened to improve
country. While poverty is ethrum), poor access to credit
the balance of power found in both urban and
among the institutions of and extension services, and
rural areas, 75 percent of
governance weak smallholder organiza-
the poor are in rural areas.
Strategic Objective 7: USAID/Kenya will, there-
tions.
Increased rural household fore, focus on increasing (USAID/Kenya 2000, pp. 68-75)
incomes the incomes of rural
households in selected The results framework adop-
Strategic Objective 3:
high and medium poten- ted by USAID/Kenya for SO 7
Reduce fertility and the
tial and arid and semi-arid includes four high-level inter-
risk of HIV/AIDS transmis- lands, most of which mediate results (IRs). IR 7.1
sion through sustainable, already rely on a combina- calls for increased productivity
integrated family planning tion of on- and off-farm in three targeted agricultural
and health services activities...
sub-sectors: dairy; horticul-
Strategic Objective 5:
Increasing rural household
ture; and maize. IR 7.2 aims to
Improved natural resource
incomes is essential to increase the volume and value
management in targeted
achieving a prosperous of traded agricultural commod-
biodiverse areas by and
and democratic Kenya. To ities, especially dairy and hor-
for stakeholders
sustain and improve pub- ticultural products. IR 7.3
Special Objective 4: lic services and build seeks increased access to
Critical needs met for democratic institutions, business support services
Kenyans affected by the Kenyans must have higher (credit and savings; appropri-
bombing of the Nairobi incomes. A population ate technology; skills, and
Embassy in 1998 and ca- with higher incomes is a business training) for micro
pacity built to address fu- population with higher ex-
and small enterprises. IR7.4
ture disasters pectations for its future
and the future of its chil- targets increased effective-
dren. When people are ness of smallholder organiza-
The activities covered in this tions in providing business ser-
able to pay for health and
report fall under SO 7. The vices to members and repres-
education services, these
Mission justifies this strategic services can be sustained enting their business interests.
objective as follows: and improved. Likewise, Below these four IRs, 15 sub-
economic growth will cre- IRs are specified.
Since 80 percent of the ate financial stability and
Kenyan population lives in allow Kenyans to take a
rural areas, and 75 per- As discussed in the following
more constructive interest
cent are somehow in- section, tree fruit cultivation
in the political environ-
volved in agriculture, ment that affects their and other forms of horticulture
Kenya’s economy is there- economic well-being. are important activities for
15
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
Kenyan smallholders. Raising production in Kenya include a lackluster economy (Minot and
productivity in these activities climate that allows for year- Ngigi 2003, pp. 3-8). Domestic
and the revenue earned from round cultivation, fertile soils, sales through traditional retail
them should therefore contrib- and a competitive labor force and public wholesale outlets,
ute significantly to the in- with good education and tech- by far, dominate the market.
creases in average rural nical background. While there are two large su-
household income that are permarket chains, they com-
sought by the GOK and According to the 2000 Rural prised less than five percent of
USAID/Kenya. Household Survey carried out domestic market horticulture
by Egerton College, Tegemeo sales in 2003. Much of what is
Curiously, given the emphasis Institute, and Michigan State sold in these supermarkets is
on raising average rural University, almost all farmers procured directly from pre-
household incomes, little in- in Kenya (98 percent) grew ferred growers – mostly com-
formation seems to be avail- some fruits and vegetables mercial farmers and a small
able on the actual levels of and 35 percent of fruit and ve- number of organized small-
these incomes. Tegemeo Insti- getable production was sold in holders (Tschirley et al 2004).
tute, on behalf of the market. Overall, fruits and Only two percent of farmers
USAID/Kenya, does track annu- vegetables contributed 18 per- currently produce for export
al movements using a proxy cent of average household in- markets. The Horticulture
method that it developed in come. Over 90 percent of Crops Development Authority
partnership with Michigan households across income (HCDA) estimates that 40 per-
State University.6 groups grow fruits and veget- cent of exported fruit is pro-
ables, although richer house- duced by smallholders (cited
3. BACKGROUND OF THE holds market a larger share of by Minot and Ngigi 2003, pp.
HORTICULTURE SECTOR their output and account for a 10-11), with the remaining 60
large proportion of total sales. percent produced by commer-
Kenya’s tropical and temper- According to a study by the In- cial farms.
ate climate zones favor cultiv- stitute of Development Studies
ation of a wide range of horti- at the University of Sussex, Horticulture Exports from
cultural crops. In the coastal households involved in the Kenya
lowlands, farmers grow man- production or processing of ex- Over the past two decades,
gos, citrus fruits, cashews, ba- ported horticultural crops export horticulture in Kenya
nanas, hot peppers, brinjals, earned higher incomes than has grown in importance, al-
and melons. In the middle alti- households that are not, other most tripling in value between
tudes, crops include bananas, things being equal. This sug- 1996 and 2001 (Table 1). Hor-
mango, avocado, pineapple, gests that enabling more ticulture (comprising fresh
grapes, passion fruit, pawpaw, households to participate in fruits and vegetables and cut
citrus, flowers, onions, garlic, the sector could reduce flowers) has become the na-
tomatoes, kale, cucumbers, poverty substantially in both tion’s third most important for-
pepper, okra, and French rural and urban areas (McCul- eign exchange earner after
beans. At high altitudes, avo- loch and Ota). tourism and tea.
cado, pears, apples, plums,
carrots, cabbages, peas, pota- While horticulture products Kenyan horticulture products
toes, and flowers are grown. have long been grown for are exported primarily to
Factors that favor horticulture home consumption, production Europe and the Middle East7
for sale in domestic and export
6 7
See David Tschirley and Mary Math- markets began in the early Regional exports, especially to neighbor-
enge. 2003. “Developing Income Proxy 20th century and has recently ing Tanzania and Uganda, are minimal.
Models for Use by the USAID Mission in Overall, Kenya is a net importer of horti-
become one of the few suc-
Kenya: A Technical Report.” Tegemeo culture from these countries (Tshirley, et al
Working Paper No. 7.
cess stories in an otherwise 2004).
16
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
Table 1: Value of Horticulture Exports (in millions of Kenya Shillings)
Year Fruits Vegetables Cut Total Horticulture
Flowers
1996 770 2,577 4,366 7,713
1997 805 3,116 4,888 8,809
1998 820 4,025 1,856 9,728
1999 1,256 5,713 7,235 14,204
2000 1,098 5,293 7,166 13,557
2001 1,560 8,035 10,627 20,221
Source: Cited in Dolan and Sullivan
where they compete with pro- are the main vegetables sup- getables and fruit are sold on
ducers from EU countries as plied to Europe. The leading export contracts that specify
well as from other African, destinations for fresh fruit ex- quantities and prices. British
Middle Eastern, and Southern ports (mango, avocado, and supermarkets took an increas-
European countries. Consign- passion fruit) are France, ing role in the vegetable trade
ments of fresh cut flowers, Dubai, the Netherlands, and during the 1990s as a way of
fruits, and vegetables are air the UK. Overall, nearly 90 per- ensuring the quantities,
freighted daily to various des- cent of Kenyan horticultural safety, and qualities that they
tinations from Kenya’s two in- exports go to Europe. The wanted. This shifted the trade
ternational airports. Some Middle East is a significant from Kenyan wholesale mar-
bulky produce is shipped from market for mangoes. Fruit ex- kets, where Asian traders are
the port of Mombassa. The ports grew rapidly from 1996 active, to contracts with large
European Union is the princip- to 2001 but remained much exporters that obtain their pro-
al importer of Kenya’s fresh smaller in value than either duce primarily from their own
produce. The bulk of flower ex- cut flowers or vegetables. The farms and large contract
ports go to the Netherlands for official figures are shown in farms. The move hurt small
sale by auction. By 1999, Table 1. out-growers. Pre-packs for the
Kenya had become the leading supermarkets and Asian veget-
supplier of flowers to the EU, The demand for horticulture ables became increasingly im-
followed by Israel, Costa Rica, products in the European mar- portant products during the
Colombia, the USA, Ecuador, kets is increasingly concen- 1990s.
and Zimbabwe. Britain, France, trated on fresh produce distri-
the Netherlands, and Germany bution channels in supermar- Some 10-15 major export-
are the major importers of ve- kets. Another important factor ing companies dominate
getables. Kenya has been de- influencing demand is increas- the sector. These compan-
scribed as one of the worlds’ ing importance among con- ies are very well organ-
sumers of food safety and the ized, often with an integ-
most successful exporters of
rated system of produc-
fresh vegetables to EU coun- environmental and social di-
tion/processing/
tries; in 2002 it ranked second mension of the food supply transport/marketing.
among non-members in the chain. As a result, the regulat- There is also a quite well
value of fresh vegetables (Jaf- ory environment is becoming developed small/medium
fee 2003). 8 Beans and peas more stringent, raising the bar size exporter sector who
for new entrants and posing are well organized on pro-
8
The $6 billion annual fresh vegetable new challenges for existing duction/transporting level,
market in the EU was supplied largely by suppliers (Jaffee 2003). but less on processing and
EU producers. Among the $950,000 mil- marketing due to their
lion of vegetables imported from non-EU Export marketing systems for size of operation. There is
members, Kenyan exports account for a third level of exporters
horticulture differ by crop.
$100 million (Jaffee 2003). Similar data who still perform more or
While cut flowers are sent for less in an ad hoc manner,
to show the position of Kenyan fresh auction in the Netherlands, ve- and rely on the prevalent
fruits was not found.
17
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
market situation and by increasing employment in supply chain. In response,
brokers for their existence. production, transport, input the Fresh Produce Exporters
However, the latter group supply, processing, sorting; by Association of Kenya (FPEAK)9
has almost disappeared increasing jobs for unskilled adopted a Code of Practice
from the flower export
workers, especially women; by for growers in 1999. The
sector in the last five
years, and will, most prob- increasing employment on Code includes a 14-step doc-
ably, decline also in the large farms and plantations; umentation procedure for en-
vegetable sector in the and by building new know- suring the traceability of pro-
next five years due to the ledge and technology that is duce handled by the export-
effects of the Code of valuable in producing and er. “This is an important step
Practice to be implemen- marketing other high value in establishing a common set
ted. However, brokers products. of standards regarding safe
make out an essential part handling of fresh fruits and
of the fruit export sector Constraints to Horticulture vegetables and disseminat-
and will continue to be im-
Exports ing the information. However,
portant if Kenya is going
to remain a fruit exporting
some aspects of the Code im-
country in the future. Demand Side Constraints: ply significant costs and
(FKAB Feldt Consulting According to Minot and Ngigi there are currently no en-
2001, p. 8) (2003, pp.9-10), the trans- forcement mechanisms.”
formation of food retailing and (Minot and Ngigi 2003, p.10)
According to a sector study changes in the structure of More recently, EUREPGAP has
contracted by USAID/Nairobi consumer demand in Europe significantly raised the stand-
(FKAB Feldt Consulting 2001), are serious challenges for ard that Kenyan produce
Kenya has several competitive Kenya and other horticultural must meet to enter the
advantages in export horticul- exporters: European market, as well as
ture: The rise of supermarkets: the cost of compliance.
A strong and well organ- The share of fresh fruits and Competition from other sup-
ized private sector vegetables sold by supermar- pliers: Kenyan horticulture
A variety of suitable cli- kets in the UK rose from 33 enjoys duty-free access to
mates for different species percent in 1989 to 70 percent European markets. If and
A rather good main road by 1997. Increasingly, super- when this preference is ter-
infrastructure and good market chains bypass whole- minated, Kenya will face in-
local supplies of inputs salers and buy directly from creased competition from
and implements exporters in Kenya and other countries such as Egypt,
Access to good air cargo countries. To protect their South Africa, Chile, Brazil,
handing facilities and air- reputations, the chains im- and Thailand. Even without
port services with ad- pose new restrictions and such a change, horticultural
equate cargo space to ma- even organize production in markets are highly competit-
jor destinations developing countries. ive and subject to rapid shifts
Rather simple export doc- Increasing concern over food
umentation procedures safety: The demand for horti- 9
FPEAK is an organization that represents
Incentives for exporters culture products in the than 140 members who are active export-
European markets is shifting ers and other interest groups. Besides for-
(VAT reimbursement and
with consumers increasingly mulating and implementing a Code of
duty-free imports of most Practice to ensure quality produce grown
inputs and implements) aware of the health con-
and shipped in an ecology- and worker-
sequences of pesticide
friendly environment, FPEAK maintains a
Horticulture production for ex- residues and placing more database of local products and suppliers
port has potential to benefit importance on food safety and provides market leads and contacts to
poor people in several ways: and the environmental and members. Its secretariat is assisted by
social dimension of the food USAID.
18
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
in export competitiveness. A general shortage of skilled ley et al 2004; Muendo, Tschir-
Kenya lost the European mar- labor and qualified manage- ley, and Weber 2004).
ket for fresh pineapple to ment staff Moreover, recent data suggest
Cote d’Ivoire in the 1980s, High air freight rates and a a downward trend in the share
was squeezed out of avocado need for more cargo capacity of smallholder production in
exports to Europe by higher to London, Paris, and Frank- these markets. Smallholders’
quality products from Israel furt share in export horticulture
and South Africa, and also has fallen from 75 percent in
Inadequate communications,
lost the market for several the early 1990s to perhaps 45
power supply, and rural feed-
temperate vegetables. It re- percent today, indicating a
er roads. Failure to exempt
sponded by finding new mar- “clear decline and rough chal-
contract farmers and out-
kets and expanding exports lenges ahead” (Muendo,
growers from VAT (because
of French beans, Asian veget- Tschirley, and Weber). Because
their products are exported
ables, and cut flowers. Export exports have soared, this does
through a third party)
competitiveness evolves con- not necessarily imply an abso-
tinuously in response to lute decline in the quantities
The industry has perceived
changes in markets, techno- that smallholders supply to the
threats to its prosperity from
logy, and competitors (Ibid). export market, but it does sug-
both the Kenyan government
gest limitations on new oppor-
and the EU. Recently there
Supply side constraints: En- tunities.
was a general fear that the
hancing the capacity of the government might raise taxes
Kenyan horticulture industry to In the context of this dualistic
and fees that impact export-
respond to changes in market market, smallholders particip-
ers. There was also a move to
demand is critical to remain ate primarily in traditional
increase government control
competitive in export markets. markets, which at present are
of horticulture by broadening
Small farmers need to become not competitive even on a re-
the role of the Horticulture
more competitive, not only gional basis. There are relat-
Crops Development Authority10
today but also tomorrow. Pro- ively few regional exports,
– from being a facilitator to a
jects such as those reviewed largely because of high trans-
more active role in buying and
here need not only to create portation costs. Kenya is in
selling commodities like a mar-
competitive advantage but fact a net importer of horticul-
keting board -- but this seems
also to sustain it. At present, tural products from Uganda
to have been withdrawn fol-
however, challenging con- and Tanzania. While Kenyan
lowing the change in govern-
straints exist on the supply exports have been competitive
ment. (Minot and Ngugi, 2003,
side. Among the most critical in international markets, the
p. 5)
are shortages of the seedling stringent quality standards
varieties needed for participa- that are being introduced in
Constraints to Smallholder
tion in exports and the lack EU and other export markets
Participation in Horticul-
(on the part of smallholders) of are likely to further raise the
ture Export Markets
the knowledge, skills, and fin- bar for small scale producers –
Despite the potential of horti-
ance needed to grow fruit in thus further limiting their par-
culture sales to increase
ways that will safeguard qual- ticipation in these markets. As
household incomes and reduce
ity and protect them from dis- stated in an article from The
poverty, a large majority of
ease. The same study identi- Financial Times:
Kenya’s smallholder horticul-
fied several important con- ture producers remain outside
straints and areas requiring In the wake of mad cow
the more lucrative export and disease and other scares,
improvement: supermarket segments (Tshir- European authorities de-
A shortage of irrigation water
mand ever tighter food
in many areas 10
The HCDA was formed in 1967 and car- quality controls. A bewil-
ries out a variety of promotional activities.
19
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
dering array of these Increase outreach and sus- cess and incentives for small-
already apply. There are tainability of solutions/ser- holders and enhancement of
more than a dozen quality vices offered by multiple pro- their capacity to respond to
standards across the EU, price incentives through em-
viders to large number of mi-
usually set up and mon-
croenterprise clients bedded services provided by
itored by the trade…For
poor countries like Kenya, Foster a better-skilled and lead firms and input stockists.
the question is whether more competitive MSE sector The Fintrac project includes a
the regulations, or non- technical component (im-
tariff barriers, are becom- The projects seek to: proved planting stock and cul-
ing incompatible with the Raise productivity through tivation methods) and tries to
vision of development that market intervention by pro- improve the international en-
sees small-scale crop pro- moting the production of abling environment through its
duction of export crops as work with European retailers
higher grade, better quality
central to poverty reduc- to shape their standards so
tion. (Wallis)
fruit by facilitating access to
improved stock and seed- that Kenyan farmers can meet
lings, productive inputs, them and by helping produ-
As the Financial Times article
training, extension and in- cers meet the retailer’s rising
suggests, large producers and
formation services quality standards.
exporters find it easier and
Increase market outlets in
cheaper to comply with such
regulations than do small and selected areas by facilitating
medium firms because large direct links between small- 1. KENYA BDS PROGRESS
firms can spread the cost of holder producers and lead TO DATE
compliance, which is substan- firms involved in fruit export
tial, over a larger volume of and processing and promot- Kenya BDS, a five-year project
sales. ing the formation of producer that started in 2002, was in-
groups tended to work in three sub-
Facilitate inter-firm cooper- sectors; tree fruit was the first
ation and organization within sub-sector selected.11 During
B. PROGRAM DESCRIP-
the overall value chain, its first two years, the project
TIONS between producers, input issued tenders and awarded
suppliers, producers and buy- contracts to eight private sec-
The two projects covered by ers, by organizing and build- tor and NGO partners active in
this assessment are designed ing the capacity of tree fruit the production and marketing
to promote growth in Kenya’s producer groups, linking of tree fruit. The contracts
tree fruit agriculture and en- smallholder MSEs to lead were designed to facilitate the
courage smallholder participa- firms that provide embedded development of sustainable
tion in the tree fruit value services, and facilitating oth- business solutions that provide
chain. USAID/Kenya funds er business arrangements material inputs (agro-chemic-
both in support of their stra- and relationships als and seed varieties), appro-
tegic objective to increase rur- Promote the development of priate technology to upgrade
al household incomes in Kenya sustainable business solu- products and production pro-
(SO 7). tions/services for rural cesses, business and skills
MSEs training, and extension and in-
The overall goals of the Kenya formation services. Several
BDS and Fintrac HDC projects contracts promote market link-
The Kenya BDS project focuses
are to: ages between smallholder pro-
on vertical linkages, especially
Increase small farmer and the link connecting farmers to 11
household incomes The second sub-sector chosen was Lake
lead firms. It is essentially a Victoria Fish; the third sub-sector has not
Promote growth in final sales business model, which em- yet been identified but is likely to be a
in selected commodities phasizes improved market ac- non-agricultural activity.
20
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
ducers and lead firm exporters standing with East Africa Contracts were drawn up
through supply contracts and Growers Association (EAGA), a between the producer groups
lead firm provision of embed- large horticulture export firm and EAGA. The producer
ded services, and encourage in Kenya, to link with avocado groups agreed to sell exclus-
inter firm cooperation through producer groups in two loca- ively to EAGA, to upgrade
organization of producer tions in Central Province. In their avocado production,
groups and provision of em- the first year, Kenya BDS mo- and to follow a good agricul-
bedded services. Embedded bilized 803 avocado farmers tural practices protocol.
services are who organized producer
products/services/solutions groups and, with the help of a
that are provided on a non-fee facilitator hired by Kenya BDS,
basis by one firm to another as negotiated a contract with
part of their commercial trans- EAGA to supply avocados that
actions. Examples include: 1) meet agreed upon standards.
buyers/exporters who offer EAGA provides embedded
pre-financing, technical ad- spraying services, grades the
vice, or inputs to their produ- fruit, and transports it to their
cers in order to ensure a qual- warehouses. Group members
ity product that meets market have been trained in the ap-
standards; 2) input suppliers plication of manure and fertil-
who provide training/technical izers, pruning and orchard hy-
advice to MSEs in the use of giene to upgrade the quality of
the product they sell in order their fruits. Near the end of
to ensure correct/successful the first year, Kenya BDS mo-
usage of the product. bilized an additional 283 avo-
cado farmers in 10 groups in
The box below shows the several new locations. These
range of activities undertaken producer groups are working
under the Kenya BDS tree fruit with another lead firm, Kenya
contracts. Annex A details Horticultural Exporters (KHE).
activities by fruit, partner, and Farmers began basic pruning,
location. and spraying and had plans to
negotiate contractual arrange-
Kenya BDS initiated on-the- ments with KHE. Kenya BDS
ground project activities in also has plans to work with an-
2003, almost a year before the other exporter, Indu Farm
first round of data collection (EPZ) Limited in the next
for this impact study. Over the quarter.
course of that year, project
staff observed a number of At the survey site in Kandara,
changes in the tree fruit value EAGA activities with avocado
chain as a result of project groups during the year prior to
activities. In order to capture the baseline survey included
the full impact of the project, it the following:
is important to document EAGA, with Kenya BDS sup-
these activities and the ob- port, initiated the organiza-
served changes. tion of avocado producer
Avocado groups (the farmers were not
Kenya BDS in 2003 negotiated previously organized).
a memorandum of under-
21
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
EAGA purchased avocados at
a negotiated price, provided
spraying and grading ser-
vices for a fee deducted at
the time of sale, transported
fruit to their warehouse in
Nairobi, and paid farmers
through their group accounts.
According to farmers, spray-
ing was not done on time
during the first year, which
reduced the yield of high-
grade fruit.
Kenya BDS encouraged the
revival of a processing fact-
ory that will buy lower grade
fruit to process avocado oil
and promoted a linkage
between EAGA and this fact-
ory. If this plant becomes op-
erational, EAGA plans to buy
all grades of fruit from produ-
cers and drop off lower grade
fruit at this processing fact-
ory on their way to Nairobi.
positive feature of the project broker.” Its follow-on Memor- Fintrac’s HDC project fo-
in that it provides flexibility andum of Understanding with cuses on a wide range of horti-
and responsiveness in their EAGA is addressing this issue. culture products, one of which
approach. In the case of KACE, One question is the potential is passion fruit. This impact
for example, Kenya BDS had a for EAGA or producers groups study focuses only on their
one-year contract to promote to absorb the costs of this passion fruit work, which in-
market information on tree function as an embedded ser- cludes plans to:
fruits through SMS technology. vice. Develop Kenyan varieties of
They discovered that KACE passion fruit for fresh export
had a number of problems and From the perspective of one Improve agricultural prac-
a business model that they did lead exporter, forward plan- tices of local producers
not really agree with, and after ning is one of the biggest chal- Expand local processing cap-
a year they decided not to lenges in the horticulture ex- abilities for local market
continue with this sub-project. port business. Exporters have products
Other USAID projects are typ- forward contracts so they Strengthen the farm-to-mar-
ically longer (up to five years) must plan ahead for the uplift ket value chain, inclusive of
and provide much less flexibil- of fruit and cannot operate ad business services to small
ity to cut losses. hoc. A challenge in working farmers
with smallholders, from this
Another lesson is the import- exporter’s perspective, is pro- Unlike Kenya BDS, the HDC
ance of Kenya BDS’s “hand jecting a timeframe of produc- project does not operate
holding” role in building trust tion, sales, and returns. They through contracts and MOUs
between producers and the need to establish ground rules but carries out activities dir-
exporters. Kenya BDS staff and work together to develop ectly through project staff
have brokered the relationship a commercialization strategy based in Nairobi and agronom-
and facilitated communication from day one. ists based in four field offices.
to help each party to under- It works with and through co-
stand the needs of the other. EAGA’s experience in linking operating partners, including
In the future, as part of its exit directly to avocado farmers the Kenya Agricultural Re-
strategy, Kenya BDS wants to has had a demonstration ef- search Institute (KARI), exist-
explore the potential for com- fect within the company. While ing smallholder associations,
mercializing the role that its the costs of interacting directly and two small businesses pro-
staff currently plays - referring with groups and providing em- ducing plant stock. In the fu-
to this role as a “network bedded services is more for
25
ASSESSING THE IMPACT OF KENYA BDS AND HDC PROJECTS IN THE TREE FRUIT VALUE CHAIN: BASELINE RESEARCH REPORT
ture, it intends to work will train 600 producer groups
through other input suppliers (with 20 members each).
as well. This five-year project
began in late 2003 and was in Progress to Date
its first year of operation at Fintrac’s HDC activity with pas-
the time of the baseline sur- sion fruit growers in the El-
vey. doret area prior to the baseline
survey focused primarily on
The project decided to focus promoting input supply and
on passion fruit because it is extension activities. These in-
regarded as a relatively cluded:
friendly crop for smallholders. Identifying farmer groups in-
Production carries low risk and terested in planting grafted
the market potential is high. passion fruit, which has
Initial project activities related stronger root stock and more
to passion fruit focus primarily disease resistance
on product development by Linking farmer groups with a
addressing two key constraints nursery operator who pro-
to smallholder production: pro- duces grafted passion fruit
duction technology and farmer plant stock
knowledge. To this end, Fintrac Coordinating with HCDA and
HDC is cooperating with KARI the MOA in linking farmer
on training and plant produc- groups to public training and
tion, and with various small extension resources, some of
businesses in Eldoret, includ- it related to compliance with
ing input suppliers and nurser- EUREPGAP standards
ies. The HDC project hopes to
establish 30-40 good demon- Activities within the project
stration plots and, through area were just getting under-
them, have a ripple effect on way when the baseline re-
passion fruit production search began in October 2004.
throughout Kenya. It is trying Fintrac HDC had identified pro-
to develop and produce fruit ducer groups and had linked
varieties that will yield more these groups to a nursery op-
juice, including the introduc- erator supplying grafted seed-
tion of the jumbo variety from lings. Producer groups began
Uganda. It also hopes to find a receiving grafted seedlings to
good investor to build a pro- plant in demonstration plots in
cessing plant that would re- October 2004.
quire input of 50 to 100 tons
per week.
2. SMALLHOLDER MSE
HOUSEHOLDS
While mangos have long been grown for recorded by HCDA in 2003, and 14 per-
A. THE TREE FRUIT VALUE CHAINS home consumption, new export varieties cent of all recorded horticulture exports.
were introduced in the 1980s. Avocados Moreover, avocados and tree fruits in
Within the horticulture sector in Kenya, were introduced primarily as a cash crop general are growing in importance in the
avocados, mangos, and passion fruit are for export in the 1990s. Passion fruit has export market.
among the most common fruit crops a long history of cultivation in Kenya. It
grown. Mangoes are grown primarily was introduced in the first half of the 20th Kenya is not among the main suppliers of
along the coast and other low land areas, century but due to disease never took off avocados in the world market. The top
while avocado and passion fruit are cul- as an export crop. It is grown primarily for suppliers in rank order of importance in-
tivated primarily in the highlands. While home consumption, with some limited clude Mexico, Chile, South Africa (the
most of these fruits are produced for sales in domestic, regional, and export main competitor for European and Near
home consumption or sale in traditional markets. East markets), Spain, Israel, and a mix of
markets, recent data suggest their grow- other countries such as the Dominican
ing role among exports and as a source 1. DEMAND Republic and Indonesia. To give a relative
of foreign exchange. These three fruits sense of volumes produced and exported
make up the top three fruit exports from Smallholders are the by far the major pro- in 2003, Mexico, the top supplier, pro-
Kenya – comprising 85% of fruit exports ducers of tree fruits in Kenya. Most of the duced approximately one million metric
in 2003 (Table 4). Hass avocados are ex- avocados, passion, and mangos produced tons and exported 135,000 tons; South
ported primarily to Europe (many are by smallholders continue to be sold in do- Africa produced 70,000 tons and expor-
shipped to Marseilles for sale in France mestic markets. However, all three fruits ted approximately 38,000 tons; and
and Germany); Fuerte avocados to the are important exports crops and they are Kenya exported approximately 20,000
Middle East; mangos to Dubai; and pas- growing in importance.19 Some points metric tons (USDA, 2005)
sion fruit to specialty markets in Europe, with regard to the demand for Kenyan
especially France. On average, growth in tree fruits in the export markets: The export market for avocados is large,
the volumes and earnings for fruits is but competitive. In the past, Kenya has
higher than for horticulture as a whole; Avocado supplied primarily Fuerte avocados, a
tree fruits in particular are seen as an al- Avocados are an important horticultural large, smooth skinned variety with some-
ternative to primary commodities like cof- export crop in Kenya, comprising approx- what limited demand to the Middle East.
fee and tea as export crops, given uncer- imately 75 percent of fresh fruit exports However, Hass avocados, which are a
tainties and trends that place Kenya at 19 smaller, rough skinned variety, are
Between 2002 and 2003, HCDA data shows that the
risk in these markets. volume and value of tree fruit exports in Kenya grew at favored in the European export markets
Exports of mangos and avocados started higher rates (17 percent and 36 percent, respectively) because they are disease resistant and
in the 1980s and in 1990, respectively. than the rates for horticulture exports as a whole (10 per- less risky to ship.
cent and 8 percent, respectively) (HCDA 2003)
Table 4: Export Statistics for Avocado, Mango and Passion Fruit, Kenya 2003
Fruit Volume Value
Kgs. Metric Tons KSH USD*
Avocado 19,020,028 19,020 892,871,043 11,787,076
Mango 2,226,550 2,227 273,611,917 3,612,039
Passion Fruit 1,505,630 1,506 393,783,047 5,198,456
All fruit 60,982,885 60,983 1,752,645,572 23,137,227
All horticulture 133,232,517 133,233 28,839,583,186 381,380,063
Source: GOK, 2003 * Ksh. 75.75 = USD$ 1 as of Dec. 31, 2003
According to one main exporter (EAGA), local investor is re-opening a factory that major player. Kenyan mangos have a
one million boxes of avocados are expor- will process avocado oil in Central somewhat flawed reputation because of
ted from Kenya each year and they sup- Province (in an area not far from Nairobi poor quality. Deliveries of diseased and
ply one quarter of this amount: 250,000 and close to the avocado groups facilit- spoiled fruit in the past have made buy-
boxes per year, including 20% Haas and ated by Kenya BDS). It is anticipated that ers wary. Improved, disease resistant
80% Fuerte varieties. Haas avocados are this firm will buy lower grade avocados varieties were introduced in Kenya about
sold primarily to Europe and Fuerte are not suitable for export or sale in the do- 8 years ago, but are still relatively new.
sold primarily to the Middle East. Accord- mestic fresh fruit market. This will Kenyan mangos have not yet made major
ing to this exporter (EAGA), Kenya avoca- provide an incentive to smallholders to inroads into the European markets and
dos have a tarnished reputation in export grow avocados and increase their sales are sold primarily in the Middle East.
markets due to past shipments of dis- and income. However, these countries place restric-
eased fruit and mistiming of shipments tions on mangos from Kenya due to a his-
that resulted in fruit being either under- The main competitive advantage of tory of weevils. If they find one weevil in
ripe or spoiled. EAGA, a large horticul- Kenyan avocados is their availability in a container, the whole container is
ture export firm in Kenya, sells mostly to late January and February when the sup- thrown out and exporters loose the value
the speculative market and, more re- ply of avocados from other countries is of the entire container.
cently, to wholesalers with links to super- low (counter-seasonal supply). It has a
markets. It is also penetrating fair trade four-week jump on South African avoca- Nevertheless, there is potential for fur-
markets, where the price premium is 12 dos – a main competitor -- in the world ther development of mango exports, es-
percent. markets. pecially for the apple mango. There also
appears to be potential for developing
The local market, to date, has primarily Mango the industry through branding; however,
involved wholesale and retail fresh fruit In the context of a highly competitive these efforts are still very under-
markets. In addition to these markets, a world market for mangos, Kenya is not a developed.
HIV/AIDS. RI staff in Kampala interviewed able from other suppliers. Another ad-
Passion fruit brokers selling Kenyan passion fruit in the vantage that Kenya has in horticulture in
The passion fruit market is unique among local market and found it to be a highly general is that it is better set up to meet
these three fruits in that there is unmet lucrative market. The Ugandan brokers certification standards than other coun-
demand in both domestic and export are seasonal buyers – they come only tries. Considerable efforts have been
markets. At this point, while passion fruit during the off season for passion fruit in made to encourage exporters to work
is grown in countries throughout the Uganda and Rwanda. with smallholders through MOA, HCDA
world, no one country appears to have
the competitive edge with passion fruit. HCDA buys the highest grade passion fruit, which it sells to exporters in Nairobi.
Zimbabwe was a competitor in producing
and exporting passion fruit in previous and a number of bilateral donor-suppor-
years, but land redistribution activities in In terms of domestic demand, producers ted projects.
recent years have disrupted production. historically supplied passion fruit to a
Both fresh and processed passion fruit government-supported passion fruit pro-
(e.g., juice concentrate) are seen to have cessing factory located first in Kitale, 2. SUPPLY CHAINS – PRODUCERS TO
significant, but as yet untapped, potential then in Sotik, and finally in Thika before it MARKETS
in world markets. It is a specialty market, closed down a few years ago in part due
however, and considered to have less to lack of supply. This may in part relate Producers to markets: overview of the
mass based appeal as a fresh fruit than to the fact that prices in export markets tree fruit value chain
as a pulp used in marinades, sherbets, are much higher (Ksh. 50/kg) than local
and juice. Passion fruit has many positive prices for juice grade passion fruit (Ksh. Figure 3 shows key actors in the tree fruit
attributes: it is easy to ship and a high- 7-15/kg). Moreover, passion fruit has not value chain in Kenya and how they are re-
value crop. Most of the fresh passion fruit been actively promoted. lated.
exported from Kenya goes to Europe, es- Tree fruits are produced primarily by
pecially France. In sum, the competitive advantage of smallholders and medium-sized growers.
Kenya tree fruits is not volume, quality, or Small-scale chemical stockists (referred
The research found a regional value chain price but rather seasonality.20 Avocados, to as agro-vets) sell fertilizers, insect-
for passion fruit extending from Kenya to mangos, and passion fruit each have a icides, and other chemicals to producers
Uganda. Kenyan passion fruit is in high ‘window’ when these crops are less avail- through privately owned shops located in
demand in Uganda at certain times of the 20
towns throughout Kenya. Many of these
See Steven Jaffee (2003). “From Challenge to Oppor-
year (October through December). There shops also offer advice to farmers on
tunity: Transforming Kenya’s Fresh Vegetable Trade in
is a ready market in Kampala for fresh the Context of Emerging Food Safety and Other Stand- what chemicals to use as well as their
fruit and juice sold in hotels and restaur- ards in Europe.” Agriculture and Rural Development Pa- proper application and safe handling. Big-
ants. Household consumption is also up per 1. World Bank; Steven Jaffee and Spencer Henson name chemical wholesalers (such as Bay-
in Uganda because it is seen as a healthy (2004). “Standards and Agro-Food Exports from Devel- er) provide periodic training for stockists
drink, especially for people affected by oping Countries: Rebalancing the Debate.” World Bank and farmers to raise awareness of the be-
Policy Research Working Paper 3348.
nefits of using their products. This train- bedded services to smallholders through Historically, smallholders in Kenya have
ing is provided through existing farmer these groups. grown tree fruits largely for home con-
groups and meetings organized by lead sumption or small-scale local trade. They
farmers in local communities. Publicly Key Problems in the Tree Fruit Sup- have not considered tree fruit a major
supported agricultural research centers ply Chain at the Time of Baseline Re- cash crop or a business activity to invest
and government organizations, such as search in. A repeated theme in the qualitative re-
the Kenya Agricultural Research Institute search is that these crops that have not
(KARI), the Horticulture Crops Develop- Constraints to smallholder participation in been taken seriously in the past.
ment Authority (HCDA), and the Ministry the tree fruit value chain include:
of Agriculture (MOA) produce and distrib- Lack of information and knowledge of Avocado
ute tree fruit plant stock. They link to the markets Avocado production in the Kandara area
farmers through demonstration centers Limited access to inputs experienced almost total collapse in the
and farmer groups. Private nurseries also Limited smallholder access to business 1990s as a result of several factors:
have begun to produce and sell tree fruit solutions and services Disease, specifically anthracnose, which
plant stock. A long and inefficient supply chain with causes black spots on the skin of avoca-
poor vertical and horizontal linkages. dos
The main marketing outlets for tree fruit Small producers are not well organized The collapse of the government para-
producers are traders and brokers, who in in terms of access to inputs and mar- statal that bought avocados prior to the
turn sell to both domestic and export kets; buyers face high transaction costs liberalization of the economy
markets. In the domestic market, they when purchasing fruit from dispersed The related collapse of a private sector
sell to domestic wholesalers, fresh fruit smallholders. plant to process low grade avocados
retailers, and small retail shops. Produ- Lack of trust among producers, brokers, into oil
cers also sell directly to fresh fruit retail- and exporters, related to problems in A tarnished reputation in export markets
ers and to HCDA. Brokers and traders are the past in the enforcement of supply due to the poor quality of Kenyan avoca-
the main conduits for smallholders to the contracts; from the perspective of pro- dos and the mistiming of deliveries res-
formal and informal processing plants ducers, buyers are not always depend- ulting in spoiled fruit.
and to exporters who buy tree fruits. Me- able or honest.
dium-scale growers often link directly to
processing plants, local supermarkets, These constraints result in low yields (es-
and exporters. Exporters buy tree fruits pecially of export quality varieties), low
produced by smallholders and medium- quality, low sales volumes, low selling
scale farmers and also produce some prices, high rejection rates, and excessive
fruit on their own plantations.21 Exporters post harvest waste – all of which affect
have just recently begun to buy directly the competitiveness of Kenyan fruits in
from producer groups and to provide em- export markets.
21
We were able to get very little information on this.
Markets Figure 5: Kenya Tree FruitDomestic
Domestic Value Chain22 Export
Brokers continued to purchase avocados Fresh fruit
grade due to lack of oil processing
Processed ards Fresh
(chemicals, protective clothing, dig-
during these years, but in limited facility ging wells, insecticides, sprayers.
volumes and at very low prices (50 cents
Functions Limited emphasis on inputs and services Limited supply of newer varieties with
to Ksh.1 per piece). Because of disease related to avocado production and up- more market potential
Exporting
and neglect of trees, the fruit produced grading by government or private sector Limited incentives among mango produ-
was poor quality. Most farmers grew Limited information Small
on how to prevent cers because of low prices and erratic
Fresh
Fuerte avocados, which are in demand in
Retailing disease
Fruit Retailers
Retail Supermarkets
market
Shops
the Middle Eastern markets, but less so in View among producers that avocados do
European markets not have good potential as a cash crop. A mango producer group leader in Masii
Informal
describes the problem of linking produ-
Exporters
small-
Resulting problems in the supply chain in-
Wholesaling Mango Wholesale
markets
scale
cers to buyers:
Process
clude: Mango production in the areas ofLarge
Eastern
Scale ors
Supply of low quality fruit by producers province covered by the study isProcessors
spread
Brokers/Traders
… if you get a good buyer like the one I
<400 acres
<50 acres
Supply of ungraded fruit lets and are dependent on brokers who cannot meet the required standards.
Assembly
Supply of Fuerte avocados, a variety come to their farms to pick the fruit. They When you try to tell them to meet those
more vulnerable to disease and with buy at low prices, only buy the best fruit, standards, it’s when they give you
Grading things like, "we cannot afford these
little demand in the European export and do not sell on contracts. In general,
chemicals, and we can’t afford these
market the mango farmers in these areas lack standards because we have no money.”
Limited incentives for Production
producers to up- bargaining power with brokers and feel So you get a challenge because you are
grade exploited by them. torn between two forces one is the ex-
because pectation from farmers, which they ex-
Input Supply Chemical
of low Avocado: “I believe there is a Key problems
Private in the mango
Stockists supply pect you to find them good market and
farm way they [brokers] talk to these chain include:
Nurserie
KARI when a buyer who needs standards
Nurseri
gate exporters in aExtension
way that we do Droughts and lackes of irrigation Product
comes he finds that farmers cannot
F- meet the standards you see the two
prices not…. So they could even buy Disease (parast MOA Research PEA
paid by from us at a very R&Dlow price telling atal) (HCDA) Uof N, forces. (a lead farmer, Masii)
us that the market is bad. Brokers
High price of chemicals to improve Utaali
K
brokers quality
created a barrier between the ex- Passion fruit
porters and us, so we as farmers Lack of credit to help cover costs of
Passion fruit is widely produced, but
Wastage became like blind men because complying with international stand-
of lower we could not tell what the export- treated primarily as a garden crop for
LEGEND
22 er had to say.”
From: Kula, Olaf, “Activity Status Report: Holding Hands with Folded Arms: Upgrading Kenya Tree Fruit Value= Chains”.
Domestic Deloitte, Touche, Tomatsu,
Final Market
No date. -Kandara Wholesale
avocado grower Fresh
markets = Participant in
the Value Chain
Brokers
= Participant in the
Value Chain, broken Exporting
= Transactions = Market functions
line indicates
skipped functions
= Services not captured in = Pillar Three
value chain Industry Leaders
home consumption. While there is unmet Another
demand for passion fruit in domestic, re- challenge is
gional, and international markets, the vul- the seasonal
nerability of passion fruit to disease and nature of
the lack of technical know-how to man- the market
age and prevent disease have limited the for fresh
ability of farmers to produce the volumes passion
demanded. fruit, with
fluctuating
The low volumes produced relate to a dis- demand
ease that attacks the roots of vines, re- during the year. Passion fruit has poten- 3. GOVERNANCE PATTERNS IN THE
duces yields, and kills the plants. Efforts tial for year-round production if managed TREE FRUIT VALUE CHAIN
to grow large volumes of passion fruit on properly, but farmers have little incentive
plantations in Kenya in the 1940s and to increase production due to low de- The relationships among firms in the tree
1950s were thwarted by rapid spread of mand in certain seasons. Currently, there fruit value chain reflect different types of
disease. This disease has periodically is no passion juice processing facility to governance patterns. Dunn and Villeda
wiped out Kenya’s passion fruit root buy passion fruit in the off seasons and (2005), describe three general types of
stock. Waves of disease over the years the fruit rots. governance patterns:
have relegated passion fruit mostly to a Market relationships characterized by
dispersed garden crop (the risk disease is arms-length transactions and little in-
lessened if plants are spread out). formation exchange between firms.
Firms in market relationships may en-
Because of its vulnerability to disease, gage in repeat transactions, but their in-
passion fruit involves more technical teractions are limited to the exchange of
management than avocados or mangos, goods or services for money.
especially if the aim is to reduce disease Network relationships characterized by
and have the plant bear fruit all year more extensive information flows
long. Grafting purple passion onto yellow between firms than in market relation-
passion root stock is one way farmers can ships. Some firms in the chain exert a
reduce the risk of disease. Another issue degree of influence or control over the
for this fruit is the safety of some chemic- operations of other firms in the chain
als that are used to control disease and (Humphrey and Schmitz 2000). Suppli-
keep the plants flowering, especially if ers in network relationships typically
there are residuals on the fruit at time of supply products according to buyers’
sale. specifications, including what is to be
produced, when it should be available, ize some other commodity value chains the exporters, who are primarily Asian, in
and how it should be produced.23 in Kenya -- for example, coffee, tea, and the right way.
Hierarchical relationships: value-added pyrethrum, for which government mar-
functions that are vertically integrated keting boards still control procurement The Kenya BDS/EAGA project links produ-
under the ownership of a single firm. An and producer prices. As producer groups cers directly to EAGA’s export market
example of a hierarchical relationship in form to link to inputs and markets, and as through the formation of producer
a global value chain would be a retail exporters form associations, the patterns groups. The producer groups have negoti-
chain in one country that manufactures are shifting more toward network rela- ated a Memorandum of Understanding
products in facilities it owns in a differ- tionships. (MOU) with EAGA to supply avocados that
ent country. meet EAGA specifications in return for
Findings from the Qualitative embedded services (spraying, sorting,
All of these governance patterns have Research on Governance Patterns grading, and transport). This direct rela-
trade offs and one is not necessarily bet- tionship marks a significant change in the
ter than another for value chain actors. Avocado avocado value chain in Kandara. In this
Before EAGA/Kenya BDS started to work context, a number of governance issues
The tree fruit value chain is characterized in Kandara, brokers played a dominant related to trust, power asymmetries, so-
by a mix of market and network relation- role in the value chain. The relationship cio-cultural biases, and information flow
ships. The horticulture sector has had between brokers and producers was are noteworthy.
limited government involvement and arms-length, with minimal exchange of
private firms have been active and com- information and no assurance of repeat Most farmers expressed trust in EAGA as
petitive – this differs from the pattern of transactions. Producers do not fully trust an organization, but some were mistrust-
hierarchical relationships that character- brokers to be fair, to come at the right ful of some EAGA staff buyers who have
time to pick the fruit, or to come at all. In continuing relationships with brokers
23
general, they feel an imbalance in the re- (who now source avocados from other
Network relationships can be broken lationship. Brokers have a somewhat areas) and who, they say, have enjoyed
down further into modular governance closer relationship with their buyers. kickbacks from the brokers in the past.
(suppliers provide information but not There is a better flow of information re- These buyers tend to reject a lot of the
process technology), relational gov- garding product specifications and price. avocados from the groups – claiming they
ernance (buyers and suppliers rely on Many brokers have long-standing rela- do not meet grading standards; however,
idiosyncratic, face to face interactions, tionships with wholesale and export buy- farmers believe it is an excuse to reject
relationships often based on trust derived ers, and some even are said to pay kick- their fruit so they can continue to buy
from social and/or ethnic ties, spatial backs to assure access to these markets. from the brokers. Some farmers also ex-
proximity, or reputation) and captive gov- Some producers expressed the feeling pressed mistrust of other producer group
ernance (single buyer and provider of that brokers (who are ethnically similar to members, who they say are continuing to
process technology, asymmetric relation- the producers) have a knack of talking to sell avocados “secretly” to brokers when
ship). they need fast cash. This is in violation of
the MOU. Some growers say they do not older members, primarily retired men. In from an area not far from Kandara, is well
fully trust the reliability of EAGA to spray many cases a man is registered as the respected by all despite her age and
on time, pick up on schedule, or use con- member but his wife is the active group gender. At this point, there appears to be
sistent grading standards. In addition, participant. Payments are made in the cautious optimism on the part of EAGA
during the first year, the growers were name of the man. Older members com- and the farmers about the direct linkage.
not all clear as to how the prices for their mented that young women do not parti-
fruit would be determined – they thought cipate because of time constraints and Mango
the prices would fluctuate according to because they consider farming and farm- Relationships in the mango value chain
the market when they were actually fixed er groups as activities for older women. had not significantly changed as a result
for the season. of the project at the time of the baseline.
In terms of information flows, the biggest Some farmer groups had formed in the
Issues of trust also arise in the relation- complaint was from farmers who believe project areas, but they had not yet forged
ship between industry leaders (in both that they did not have good information new linkages to input, extension, or
avocados and mangos) and government. on export prices and that both Kenya BDS product markets. Most growers were
Although they communicate with each and EAGA were holding back price in- selling their mangos individually to a
other, their relationship has been de- formation. Some farmers also felt that in- wide array of unknown brokers and
scribed as fragile. This is due in part to formation on the MOU negotiations re- traders – suggesting a value chain gov-
the widely held belief that private sector garding prices between the groups and ernance pattern of “market
leaders look after themselves, while the EAGA did not flow from producer group relationships.”
government looks after smallholders leaders to members. This was an issue
(Kula, no date). Some producers believe during the first year of the project. The qualitative research found a high de-
that EAGA buyers change grading stand- gree of mistrust in brokers among mango
ards based on the volumes they need at In sum, there is a moderate level of trust farmers. Brokers were described as “out-
a particular time, not a standard protocol. in the avocado value chain between siders”, “little known, and “not to be trus-
Kandara producer groups and the export- ted.” Some farmers felt brokers lied to
In terms of power asymmetries, both the er. The balance of power is still with the them about what they were paying other
avocado producers and the brokers feel buyers, but growers are slowly increasing farm- ers.
that exporters call all the shots in terms their power. Men tend to dominate the Most Mango “… The market pay-
of volumes purchased, prices, grades, leadership of the producers groups, but has just become bad … for ments
timing, and contract terms. women are active members. While there from things sold outside to
was some sense among members that people who are strangers…
In socio-cultural terms, the producer they do not know everything that is going it is like they are mixed up
groups and brokers are ethnically homo- on in their groups, most expressed satis- and one can detect some
geneous but the exporters are primarily faction with the groups and how their fishy business in them as it
Asian. Age seems to be a factor – the pro- leaders managed them. Moreover, the is not straight.”
-Mango Farmer, Masii
ducer groups tend to be dominated by Kenya BDS staff person, a young woman
brokers are made in cash; some farmers political and economic reasons, leaving
expressed concerns that the brokers Brokers and other buyers do not offer in- the farmers with limited alternative mar-
might come back and steal it from them. formation to farmers on standards re- kets for their passion fruit. Around this
Larger farmers tended to have better quired. They do not reveal whom they are same time, a passion fruit processing
communication and more trust with selling to for fear producers will develop plant located in Thika also closed down,
brokers, and even direct links to export- direct links. Producer group leaders compounding the problem. Since then,
ers whom they trusted. Several farmers provide information on mango buyers farmers have sold their passion fruit to
talked about the challenge of trusting and prices to group members and the two main outlets: (1) the HCDA, a govern-
other producer group members to stick to groups provide a good forum for informa- ment sponsored organization promoting
certain agreed upon prices. Because each tion exchange. In general, however, pro- horticulture development and buying up
farmer sells from his/her own homestead ducers have barriers in gaining market in- horticulture products that are sold in ex-
and because there are no central collec- formation (prices brokers are paying) due port markets; and (2) Ugandan brokers
tion points for mangos, prices are negoti- to their geographic isolation. The fact who cross the western border into Kenya
ated one on one, with little communica- that brokers come to individual farms area every week to buy from farmers in
tion between farmers. Farmers expressed with their own pickers to buy mangos re- the El Doret area.
confidence that the leaders of their pro- duces information flows and the negotiat-
ducer groups would represent their in- ing power of farmers. Thus, the governance pattern for at least
terests. part of the value chain (those selling to
In general, there is a low level of trust in Ugandan brokers) is characterized by a
In terms of power asymmetries, mango the value chain and the balance of shift from a somewhat ‘captive’ network
farmers expressed the view that brokers power, from the perspective of farmers, is towards a more open market relationship
dictate prices, when they buy, how much with buyers, mostly brokers. The lack of (moving to the left on the continuum in
they buy, and what they buy. Timing is central collection points for mangos and Figure 1). At the time of the baseline, pro-
especially important for mango farmers, the dispersed settlement patterns of ject activities were just beginning and
as over-ripe or under-ripe fruit commands farmers is a key constraint to the flow of had not affected this pattern.
a lower price. Brokers control the timing. information.
In terms of trust, the research found a
In terms of socio-cultural biases, one of Passion Fruit general sense of trust and cooperation
the biggest issues with mango farmers is The farmers interviewed in the qualitative among producers and between producers
that brokers are “outsiders.” Men appear research were already growing relatively and brokers. While women producers had
to dominate the producer groups while large volumes of passion fruit. In previous a great deal of trust in the Ugandan
women play an active role in mango pro- years they sold it to a large-scale brokers, they did not trust input suppliers
duction and sales on their farms. Farmers farmer/politician who had an export firm to sell them safe and effective chemicals.
in the project areas are from the same selling flowers and passion fruit in
ethnic groups, while exporters are European markets. A few years ago his There are power asymmetries in the
primarily Asians. business closed for a combination of value chain. Women do much of the pro-
duction One social divide in this value chain cept that agro-stockists supplying chem-
and are
Avocado (Persea Americana) ori- is gender. Men dominate the formal icals are not very active or knowledge-
ginated in south-central Mexico,
respons- activities of the passion fruit able about passion fruit). Overall, there
sometime between 7,000 and 5,000
ible for B.C. But it was several millennia be- groups, while women do most of seems to be a fairly good balance of
selling fore this wild variety was cultivated. the production and selling. Men go power within this chain between produ-
passion Archaeologists in Peru have found do- to all the training on passion fruit, cers and buyers.
fruit from mesticated avocado seeds buried grafting, and application of chemic-
their with Incan mummies dating back to als, even through it is women who
farms or 750 B.C. and there is evidence that do all the work related to passion 4. INCENTIVES AND DISINCENTIVES
central avocados were cultivated in Mexico fruit. Women expressed fear and TO UPGRADING
collection as early as 500 B.C. lack of understanding about chem-
points. (from: www.avocados.org) icals to prevent disease – historic- Upgrading can be defined as “innovation
The qual- ally a major problem in that area. that increases value added” (Giuliani,
itative re- They also expressed considerable Pietrobelli, and Rabellotti 2004) and can
search found that women had organized lack of trust in small input suppliers in the take five forms:
informal groups to sell to Ugandan area to direct them to safe and effective
brokers, and had good bargaining power chemicals. 1. Process upgrading: an increase in
on prices during the high demand sea- production efficiency, resulting in
son. In previous years, the farmer/politi- Information flows on price have been fa- either a) greater output for the same
cian played a very powerful role in pas- cilitated by mobile phone contact with level of inputs or b) the same level of
sion exports, but his departure left a va- buyers and by HCDA posting their prices output for fewer inputs. Process up-
cuum. on a weekly basis. This price information grading can involve improved organiz-
flows to farmers through the farmer ation of the production process or im-
In terms of socio-cultural issues, the groups. Information on export certifica- proved technology.
Ugandan brokers are all Muslim men tion standards, approved chemicals, and
while the producers are mainly Christian improved production techniques flow to 2. Product upgrading: a qualitative im-
women, but this cultural difference has men through training but they do not provement in the product that makes
not affected the quality of their relation- transfer it to women, who are the main it more desirable to consumers and
ship, which appears to be excellent. The producers. The research found in some commands a higher unit price;
Ugandan brokers even bring Ghanaian cases information that group leaders had
cloth to the women farmers from Kam- did not get transferred to other group
3. Functional upgrading: the entry of
pala. They are in frequent contact with members.
a firm into a new, higher value-added
each other via mobile phone and meet
function in the value chain; functional
once a week on the farm of the woman In general, there appears to be a high
upgrading moves the firm closer to
who leads the informal group (and offers level of trust and cooperation among act-
the final consumer and positions it to
her farm as the central collection point). ors in the passion fruit value chain (ex-
receive a higher unit price for the Opportunities for upgrading in the avo- grading is that newly planted Hass avo-
product; cado industry can be pursued in several cado trees take 4-5 years before they
ways: by improving quality; promoting fruit. An alternative approach is to graft
4. Inter-chain upgrading: the entry of improved varieties; expanding market Hass avocados onto Fuerte avocado trees
a firm into a new and more lucrative outlets; and complying with certification (most avocado producers in Kandara
marketing channel in the value chain standards. have Fuerte trees that they planted in the
such as moving from the domestic to 1980s and 1990). However, the grafting
the export market for the same Higher quality fruit can be produced by process requires technical expertise and
product; reducing the incidence of antracnose, a diligent care – proper grafting techniques,
disease that leaves cuts and black spots spraying, fertilizer application, pruning,
5. Inter-sectoral upgrading: using on the skin of Fuerte avocados (the vari- picking the right way, and picking at the
knowledge gained in one value chain ety traditionally produced in Kenya). This right time. EAGA and Kenya BDS have fa-
to move into a new value chain or disease is prevented through a spraying cilitated this process quite successfully in
product line, involving a completely regime that must be carried out at appro- Kandara, and many farmers are now suc-
different product or service. priate intervals during the growing sea- cessfully producing Hass avocados.
son. The key incentives are that this will
Both the Kenya BDS and Fintrac HDC pro- increase the yield of high-grade fruits. Avocado upgrading also can involve
jects facilitate activities to promote up- The disincentive is that it takes time, ex- selling to new market outlets. In Kandara,
grading of tree fruits, primarily product, pertise, and money to buy chemicals, through the EAGA sub-project, producers
process, and inter-chain upgrading. The spraying machines, and hired labor. If the link directly to exporters through produ-
aim is to improve the capacity of small- spraying is not done at the right time, the cer groups. The groups sign agreements
holders to respond to changing market investment is lost. Moreover, Fuerte avo- to sell exclusively to the exporter in ex-
demand and increase rural incomes. The cados get a lower price than Hass avoca- change for embedded services, including
baseline research identified specific dos. spraying, assembly, grading, and a cent-
forms of upgrading in the tree fruit value ralized payment system managed by the
chain and the views of producers and Another form of avocado upgrading is to group leaders. Each producer is respons-
other actors in the value chain on the in- grow an improved variety of fruit, spe- ible for picking their own fruit and deliv-
centives and disincentives to upgrade. cifically, Hass avocados. The incentive to ering it to a collection point. Farmers are
These findings are discussed below and upgrade to Hass is that it is disease res- paid through their groups, with deduc-
will be compared to findings in the istant, easy to ship, and in high demand tions made for the costs of spraying at
second round of qualitative research on in the export market. It is the most popu- the time of payment. There are a number
changes in upgrading and how they are lar and common variety of avocado in the of incentives for this direct linkage. Pro-
related to project interventions. world and grows well in Kenya. It is more ducers have a secured market and a ne-
tailored to the export market and com- gotiated contract price. The prices they
Upgrading Avocados mands a higher price than Fuerte or other receive are higher than those paid by
varieties. A constraint to this form of up- brokers. Farmers have access to spraying
on credit, and access to training in im- to pay school fees). Moreover, brokers ments are quite stringent - they require
proved production techniques. The spray- used to provide labor for picking – a func- soil and water testing, exclusive use of
ing service provided through the exporter tion that farmers now have to assume in- approved chemicals, proper application of
saves the farmers labor time, organiza- dividual responsibility for. chemicals, use of protective clothing,
tional responsibility, and hassle. They are proper storage, documentation of chem-
able to access the expertise of EAGA and Another form of inter-chain upgrading in- ical use, storage and handling, and tests
it does not require an up front capital out- volves selling to an avocado oil pro- for chemical residuals. Training in EUREP-
lay. The exporter is assured high quality cessing plant that will be opened in the GAP standards was ongoing in Kenya at
fruit and the use of approved chemicals. Kandara area. This plant was built years the time of the baseline, in anticipation of
ago but closed down when the avocado the introduction of these standards
There are also some constraints to this market collapsed. With the revival of avo- (which are actually voluntary) in 2005.
direct market linkage. Organizing the cados in the area, the original owners For producers, the incentive to comply
farmer groups can be challenging and now have plans to reopen this factory. with these standards relates largely to ac-
time consuming. EAGA is used to This will enable avocado producers to sell cessing export markets; however, these
sourcing crops from small holders with their whole crops, even the lower grade practices also improve productivity and
shorter production cycles (e.g., French fruit that does not meet buyer standards. environmental and occupational safety
beans) and there is ongoing pressure The prices would be low, but the farmer and produce higher quality fruit for do-
from management to see more immedi- would at least get something. The ad- mestic sale. For exporters, smallholder
ate bottom line results than is possible vantage of this for producers is that it di- compliance is challenging to monitor and
with avocados. For farmers, the prices versifies the market; for exporters, farm- document, but the incentive is to main-
are negotiated ahead of time, and may ers are likely to stay with the crop even if tain and/or improve competitiveness in
be lower than the market price. They run they have periodic problems in quality or export markets. The main disincentives
the risk that EAGA may not take all of marketing because they will at least get to producer compliance are lack of capital
their fruit, yet their MOU with EAGA limits something. It is not certain that the plant to finance costs associated with regular
their options for selling to others and de- will open, and farmers are a bit unsure soil and water testing, building and main-
creases access to these buyers. There is how their avocados would be delivered to taining structures for collection and stor-
also the risk that EAGA may miss their the factory (and the cost). age, and buying capital equipment (such
spraying or pick up schedules, with asso- as charcoal coolers) and protective cloth-
ciated risks and costs to farmers. Produ- Upgrading can also include meeting certi- ing. The risk to all is if, after all the effort
cer group leaders subsidize the process, fication standards required in export mar- and investment, the standards are not
and do not get paid. While farmers have kets, specifically those being introduced really enforced in the export markets. In
issues with the brokers, some of them by EUREPGAP. This process involves a this case, other countries that do not
have played a support role that some variety of actors on the value chain – comply may beat them out in the market.
farmers miss (one farmer referred to the farmers, input suppliers, extension work-
fact that a broker used to give him ad- ers, producer groups, exporters, import-
vance payments when he needed money ers, and auditors. The EUREPGAP require-
Upgrading Mangos
Like avocados, mangos can be upgraded
by introducing new varieties, improving
production techniques, improving busi-
ness practices, complying with certifica-
tion standards, and linking producers
more directly to exporters.
a. Scale of enterprise
The average number of trees of the tar-
geted fruit (that is, avocado trees in the
avocado sites, mango trees in the mango
sites, and passion fruit vines25 in the pas-
25
Passion fruit actually grows on vines rather than trees.
Table 6. Number of Trees/Vines per SME by Intervention: Average Fruit production (measured in pieces of
and Distribution fruit harvested in the past year for avo-
EAGA SITE KADI Fintrac HDC Just Juice cado and mango and in kilograms for
passion fruit) also varied widely from
Total
Participant
Total
Total
Participant
Total
Participant
Total
Control
Control
Control
Control
Control
Participant
Participant
farm to farm. In the EAGA avocado area
(see Table 7), smallholders averaged
around 10,000 pieces per year, with pro-
gram participants producing more than
twice as much on average as control
25 25 50 35 34 69 17 14 15 29
BASE: Total Sample 0 0 0 0 9 9 70 0 70 206 3 379 8 1 9
group farmers. Participants at the SITE
mango area averaged about 35,000
pieces per year, again more than twice
14 20
1-9 558 3 2 2 4 0 0 0 0 0 0 0 0 0 the production of control group members.
15 10 At the KADI mango site, average produc-
10-19 82 75 7 50 51 1 3 0 3 0 0 0 1 9 10 tion among participants was much lower
20-29 45 16 61 29 52 81 4 0 4 0 1 1 0 10 10 (11,000 pieces) than in the SITE area.
13 Passion fruit production (see Table 8) was
30-49 48 7 55 55 77 2 14 0 14 2 5 7 1 11 12 very high among Fintrac HDC project par-
17 ticipants and controls, averaging 353,000
50-99 13 2 15 83 91 4 17 0 17 6 31 37 14 34 48 kilograms. In the area covered by the
12 20 13 13 21 Kenya BDS/Just Juice intervention, aver-
100+ 7 2 9 6 76 2 32 0 32 198 5 333 1 87 8 age production was far smaller, 67,000
12 10 10 10 45 37 19 28 kg. per year. It is clear that the Fintrac
Mean 25 12 18 0 82 1 9 0 9 961 3 730 6 2 2 HDC project deals with an entirely differ-
15 15 16 12 12 1,25 81 1,10 87 24 64 ent size class of farms from the Kenya
Standard Deviation 34 16 27 9 9 0 5 0 5 2 1 2 2 4 3 BDS interventions.
Participant
Participant
Participant
that had no production) while fifteen per- ting
Control
Control
Control
Total
Total
Total
cent harvested more than 30,000 man- con-
gos in the previous year. KADI mango sider-
participants clustered in the 3,000- ably
0 0 0 0 7 0 7
10,000 production range. higher
1-1,000 21 35 56 18 25 43 3 0 3
yields,
1,001-2,000 13 43 56 24 56 80 2 0 2
In the Fintrac HDC passion fruit area, the run-
distribution of production was highly un- 2,001-3,000 20 50 70 31 49 80 8 0 8 ning
equal, with 77 farms not yet in production 3,001-5,000 35 54 89 38 43 81 13 0 13 from
and 133 farms producing more than 5,001-7,000 27 14 41 32 34 66 15 0 15 300 to
100,000 kilograms of fruit per annum. Av- 7,001-10,000 28 27 55 39 40 79 10 0 10 more
erage production on participating small- 10,001-15,000 38 10 48 30 36 66 8 0 8 than
holdings was more than twice as great as 15,001-20,000 23 8 31 27 21 48 3 0 3 1,000
on control farms. The Just Juice area also 20,001-30,000 23 4 27 24 20 44 4 0 4 man-
had a significant number of non-produ- 30,001 – 50,000 14 3 17 25 14 39 2 0 2 gos
cing farms but also had fewer large pro- 50,001 – 70,000 4 0 4 19 4 23 1 0 1 per
ducers, hence much less dispersion 70,001 –100,000 2 1 3 13 3 16 0 0 0 tree.
around the mean production level. 100,001+ 2 1 3 23 4 27 1 0 1 Pos-
Total 250 250 500 350 349 699 70 0 70 sible
The productivity of the avocado and Average in 000's 13 6 10 35 14 24 11 0 11 ex-
mango smallholdings in the survey (Table Standard Deviation 17 13 15 147 56 111 16 0 16 plana-
9) is defined as annual production per NOTE: Average production is the average per producing farm. tions
producing tree or vine. In the EAGA avo- for
cado area, modal productivity is in the these
401-800 range for both participants and super-
controls. However, participants have cipants in the SITE area average 300 ior
slightly higher productivity on average mangos per tree, versus 150 for both yields include better tree stocks, better
(600 pieces per tree) than controls (500). control group farmers and participants in cultivation techniques, and greater ma-
the KADI area. The higher average for turity of the trees.
The SITE mango area is more productive SITE area participants is accounted for by
on average than the KADI area. Parti- the presence of a significant number of
Fintrac HDC passion fruit participants mango sales. In the KADI area, earnings
(see Table 10) attain productivity levels were much lower: average mango sales
(500 kg. of fruit per vine on average) that were only Ksh 6,000 (US$76) and no
are two and a half times as high as Just farmer earned more than Ksh 20,000
Juice participants (200 kg.). Fintrac HDC from this source.
controls also have relatively high pro-
ductivity, while Just Juice controls have Significant numbers of passion fruit grow-
the same average productivity as Just ers did not sell any fruit because their
Juice participants. vines were still immature. Fintrac HDC
project participants averaged Ksh 11,000
c. Participation in trade in sales; 15 project participants earned in
Many Kenyan farmers grow tree fruit for the Ksh 20,000-90,000 range and three
use at home and limited sales to the local made more than Ksh 100,000 selling pas-
market, but few participate in supplying sion fruit. Control group members in the
the export trade. In our sample, 97-98 Fintrac HDC project areas did consider-
percent of avocado and mango growers ably less well, averaging Ksh 5,000 in
earned some income from tree fruit sales, passion fruit sales. Passion fruit growers
but nearly all of them earned only very in the Just Juice intervention area also av-
small amounts (see Table 11). In the eraged just Ksh 5,000 in sales, with little
EAGA intervention area, 84 percent of difference between participants and con-
farmers earned less than Ksh 10,000 trols. No one in this area earned over Ksh
(US$127) per year from avocado sales 20,000 from passion fruit sales.
and the average for all growers was only
Ksh 8,000 (US$101). Just two farmers,
both program participants, earned more
than Ksh 100,000 (US$1,266) from avo-
cado sales in the year preceding the sur-
vey.
Table 10. Kilograms of Passion Fruit Harvested in Past Year per Produ-
cing Vine
Fintrac Passion Just Juice Passion
(Kgs) (Kgs)
Parti- Parti-
cipant Control Total cipant Control Total
0 27 50 77 37 20 57
1-100 23 6 29 68 66 134
101-200 41 36 77 11 33 44
201-300 32 41 73 15 11 26
301-400 28 28 56 6 10 16
401-800 44 10 54 9 11 20
801-1000 1 0 1 1 0 1
1000+ 10 2 12 1 0 1
Total 206 173 379 148 151 299
Average 500 300 400 200 200 200
Table 11. Earnings from Sales of Fruit in Past Year (Kenya shil-
lings)
Fintrac
EAGA SITE KADI Just Juice
HDC
Participant
Participant
Participant
Participant
Participant
Control
Control
Control
Control
Control
Total
Total
Total
Total
Total
0 6 4 10 11 7 18 0 0 0 24 52 76 18 12 30
1-10,000 197 213 410 239 271 510 65 0 65 152 111 263 129 138 267
10,001-20,000 30 24 54 53 48 101 5 0 5 12 7 19 1 1 2
20,001-30,000 7 4 11 13 7 20 0 0 0 2 1 3 0 0 0
30,001-40,000 5 3 8 10 4 14 0 0 0 2 0 2 0 0 0
40,001-50,000 0 2 2 3 5 8 0 0 0 2 0 2 0 0 0
50,001-60,000 2 0 2 7 3 10 0 0 0 5 1 6 0 0 0
60,001-70,000 1 0 1 7 1 8 0 0 0 2 0 2 0 0 0
70,001-80,000 0 0 0 3 1 4 0 0 0 1 1 2 0 0 0
80,001-90,000 0 0 0 2 1 3 0 0 0 1 0 1 0 0 0
90,001-100,00 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0
100,001-150,000 1 0 1 1 1 2 0 0 0 1 0 1 0 0 0
150,001+ 1 0 1 0 0 0 0 0 0 2 0 2 0 0 0
Total 250 250 500 350 349 699 70 0 70 206 173 379 148 151 299
Average (000's) 9 6 8 12 9 11 6 0 6 11 5 8 4 5 5
NOTE: Average sales the average for all farms, regardless of whether they sold
fruit.
Table 12. Sales by Type of Customer and Interventions (percent of total
sales)
Fintrac
EAGA SITE KADI HDC Just Juice
Avocados Mango Mango Passion Passion
Participant
Participant
Participant
Participant
Participant
Control
Control
Control
Control
Control
Total
Total
Total
Total
Total
HCDA 3 n 2 1 - n - - - 13 3 9 1 - n
Exporters (direct) 4 1 2 3- 1- - - 3- 2 5 3 4
Subtotal direct sales
to exporters 7 1 4 4 0 1 0 0 0 16 3 11 6 3 4
Broker selling to exporter 7 20 13 12 1 7 1 - 1 12 6 9 18 21 20
Brokers selling to Uganda
/Rwanda/Other regional mar-
ket 1 n 1 1 n n - - - 28 23 27 1 1 1
Broker selling in Kenya 17 24 20 16 11 14 23 - 23 9 13 10 8 10 9
Broker (Do not know
whom selling to) 18 20 19 14 12 13 6 - 6 12 19 15 9 14 11
Local traders selling
in Kenya (from nearby towns) 18 26 22 22 22 22 21 - 21 10 13 11 22 31 26
SubTable
total13: Percentage
sales to of 61
Contract Sales
90 75 65 46 56 50 0 50 71 74 71 58 77 67
brokers & traders PARTICIPANT CONTROL TOTAL
Consumers 5 5 5 17 29
(%) 23 (%)
21 - 21 10 17 12(%)
21 19 20
Wholesale markets/ whole- 2 3 2 4 5 5 6 - 6 1 1 1 2 - 1
EAGA Avocado
saler 80 12 46
Shop / Supermarket n - n 4 11 8 9 - 9 1 2 2 1 n 1
SITE Mango
Masii mango 47
24 - 13 1 - n 13
- - - n 1 30
1 7 - 4
Fruit processor
KADI Mango factory 1 1 1 - - - 3 - 3 n - n 1 1 1
Another farmer 23 - n
n 3 7 5 -0-
10 - 10 1 2 123 3 - 2
Top notch
Fintrac Passion n - n 1 n 1 - - - - - - - - -
61 74 64
Schools - - - 1 2 1 - - - - - - 1 - n
Just Juice Passion
Hospitals -6 - - n - n 19- - - - - - 38- - -
Subtotal direct sales 32 9 21 31 54 43 49 0 49 13 23 17 36 20 29
to other domestic buyers
10 10 10 10 10 10 10 10 10 10 10 10 10 10
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Table 14: Distribution of Respondents by Amount of Hired Labor
f. Use of inputs Used
The survey included questions about the EAGA SITE KADI Fintrac HDC Just Juice
use of water, fertilizer, and insecticide or
Participant
Total
Total
Participant
Total
Participant
Total
Total
Control
Control
Control
Control
Control
Participant
Participant
fungicide sprays. Water is necessary for
fruit tree cultivation, but very few of the
smallholders surveyed had access to ir-
rigation. Of the 1,947 farmers in the
sample, only 40 reported using drip irrig-
17 18 10 15 26 13
ation and 135 had sprinkler systems. The
None 11 2 3 9 8 7 1 0 1 50 78 128 63 70 3
remainder relied on hand watering or
10 16
rainfall.
1-5 30 34 64 56 4 0 2 0 2
5 11
16 15 15 30
6-10 39 16 55 48 29 77 5 0 5
14 22 85 11 16
A total of 798 farmers, 41 percent of the
11-15 30 6 36 42 17 59 4 0 4
2 6 45 15 20
total sample, reported spending money
16-20 43 3 46 29 13 42 5 0 5
3 8 59 7 16
to purchase fertilizer for use on their fruit
21-25 23 3 26 19 12 31 11 0 11
2 3 16 3 9
trees. Expenditure by these farmers aver-
26-40 33 7 40 15 13 28 18 0 18
5 14 95 14 19
aged Ksh 3,400 (US$43). On average,
41+ 41 9 50 32 3 35 24 0 24 118
64 182 40 16 56
program participants spent more than
Total 25 25 50 35 34 69 15 29
twice as much on fertilizer as control
0 0 0 0 9 9 70 0 70 206 173 379 148 1 9
group members (see Table 16).
Average days used 34 6 20 16 5 10 68 0 68 198 375 278 53 17 35
A larger number of respondents (1,388 or
71 percent of the total) said they had pur- nearly all program participants are mem-
chased sprays for use on the targeted bers of such groups. In sharp contrast,
fruit trees during the past year. The aver- hardly any control group members belong
age expenditure reported was the same to producer groups. This dichotomy is
as for fertilizer, Ksh 3,400. Again, pro- shown in Table 18.
gram participants spent more than twice
as much as control group members (see
Table 17).
Average #
income
sources 3.6 3.6 3.6
Table
Table
25:22: Sources
Respondent
of household
Households by have five or more sources and sale of vegetables (29 percent).
b. income
Members (Totalwith
sample)
Salaried Employment of income. Mango farmers Within the sample, 27 percent of house-
Sources Source and of Intervention
income % house- have a slightly higher aver- holds had income from business activit-
of holds
% HH with salar- withAverage age number of income ies, 12 percent from farm labor, and five
House- ied workersincome number sources, but otherwise percent from non-farm labor. Only five
hold In- from salaried there are few differences percent of respondents had income from
sourceworkers
come across interventions, fruits, remittances (Table 25).
(%)
The sur- (%)Participant or between participants
Participant
(%)Control
Tree fruit production 96
(%)Total
vey gen-and sale and controls (Tables 23 and While earnings from tree fruit sales are
Control
Total
erated Livestock production 56 24). not high in absolute terms (see Enterprise
data fromand sale section B.1.c), this income plays an im-
respond- Cereals and tubers pro- 42 The most frequently repor- portant role as a source of household in-
ents onduction and sale % % % ted sources of income in- come. Almost half of all households
their OtherEAGA
fruits production
40 27 40 .5 .3 .4
33 clude production and sale ranked tree fruit as the number one
sources and sale 1 6 4 of tree fruits (96 percent)31; source of income (Table 26). Tree fruit in-
of house-SalariedSITElabor 49 67 30 .6 .9 .8
58 production and sale of live- come is more important for participants
hold in-Vegetables production 29 6 9 2 stock (56 percent); produc- (61 percent ranked it number one) than
come, and sale
KADI 50 50 .7 -0 -0 tion and sale of cereals for controls (32 percent ranked it number
the most Business activities 27 0 - - and tubers (42 percent); one). Tree fruit income is especially im-
import- Farm labor
Fin- 42 42 12
42 production and sale of oth- portant for Fintrac HDC participants,
.5 .4 .5
ant Non-farm
tractlabor 5 er fruits (40 percent); and three-fourths of whom ranked passion
2 7 0
Remittance
HDC 5 fruit income as their number
sources production
Investment/interest
JUST 33 22 5 .4 .3 .4
27 Table 24: Average
of house- one source. The im-
JUICE 9 7 3 Number of Household
hold in- Table 26: Percent of Households Ranking Tree Fruit In- portance of tree
come, come Income Sources by fruits as a source of
TOTAL 43 44 44 .5 .6 #1.6 or #2
and the PARTI- InterventionTOTAL
CONTROL income for farmers
7 2 0
relative CIPANT (%) (%) is further supported
PARTICIPANT
(%) by the number of re-
CONTROL
position
TOTAL
of tree fruit income in the households. % Household ranking tree fruit income #1 spondent house-
EAGA 58 37 48 holds ranking it
The findings show that households in SITE 62 27 44 number two in im-
the sample are quite diversified in their KADI 50 N/a N/a portance – 22 per-
Fintrac HDC 77 46 63 cent of participants,
sources of income. On average, they 3. 3. 33 3.
JUST JUICE 45 21
EAGA
have 3.6 sources. Very few households TOTAL and 30 percent of
31
The survey supervisor 61
believes that this 32 may7be 6 476
number
have only one source of income (one % Households ranking tree fruit income #2 3. 4. 3. controls (Table 26).
overstated because the subject of the surveySITE
was tree fruit Overall, 35 percent
percent) and 21 percent of households EAGA 25 29 6 0 8
27
production. 3. -0 3.
SITE 20 28
KADI 24
KADI 24 -0- 8 - 248
Fintrac HDC 15 Fintrac
28 3. 3. 3.
21
HDC 5 4 4
JUST JUICE 30 36 33
JUST 3. 3. 3.
TOTAL 22 30 26
JUICE 8 2 5
of households estimated that tree fruits than Ksh. 3,000 (US$40) and 40 percent sets, including productive assets. Based
comprised more than half their household have average monthly per capita ex- on these scores, the households were
incomes – 48 percent of participant penditure less than Ksh. 2,000 (US$27) classified into eight wealth categories
households and 21 percent of control (Table 30). This suggests that a signific- providing a socioeconomic profile of parti-
households (Table 27). Income from pas- ant number of households are under the cipants and non-participants. 32
sion fruit was particularly important for $1/day poverty line (39 percent of parti-
Fintrac HDC participants (Table 28). cipants and 41 percent of controls). By The findings show that there are more
comparison, 26.5 percent Table 27: Proportion of Household controls than
c. Household Consumption of the Kenyan population Income from Tree Fruit (Estim- participants in
Expenditure was under the $1/day ated) the lowest asset
Consumption expenditure as defined in poverty line in 1999 PARTI- CON- TOTA categories (1
the study includes the estimated value of (UNDP, 2001). At the same CIPANT TROL L and 2) and
food grown at home, education expendit- time, many of the respond- (%) (%) (%) more parti-
ures, and all other cash expenditures ent households have relat- 1- cipants than
21 43 32
over the past month. From these ques- ively high levels 24% con-
tions, total monthly consumption ex- of consumption trols
Table 28:25-
Proportion
30 of Household
36 Income
33 from Tree
penditures and per capita monthly con- (15 percent have 49% in the
Fruit (Estimated) by Intervention
sumption expenditures were calculated average monthly 50-
EAGA SITE38 KADI 17Fintrac28Just Juice
for each respondent household. House- per capita ex- 74%
HDC
hold consumption expenditure generally penditures over 75-
9 3 6
Control
Total
Total
Control
Participant
Participant
Participant
Total
Total
Participant
Control
Control
Control
Participant
is preferred over income as a measure of Ksh.10,000. 99%
Total
household poverty. (US$138)). 100% 1 1 1
Total 100 100 100
Monthly per capita expenditure averages d. Household
Ksh. 5,800 for the overall sample (Table Assets 0-
29). Consumption is higher for parti- Assets can be a 24% 20 31 26 19 50 35 30 0 30 14 17 25 36 57 46
cipants (Ksh 6,500) than controls (Ksh. proxy measure 25-
5,200). The table further shows much for household 49% 39 46 42 31 36 34 34 0 34 15 25 19 32 29 31
lower expenditures in the EAGA sample, wealth or eco- 50-
especially among controls (Ksh. 2,900), nomic status. 74% 37 20 29 43 13 28 33 0 33 42 26 35 23 13 18
and much higher expenditure among Respondents in 75-
the study were 99% 2 3 3 5 1 3 3 0 3 27 8 18 9 2 5
KADI participants (Ksh. 16,000).
asked a set of 1
00% 1
0 1 1 0 n 0 0 0 2 3 3 0 0 0
yes/no questions
Half of the sample households have aver- on a long list of 10 10 10 10 10 10 10 10 10 10 10 10 10
100 0 0 0 0 0 32 A 0 description
0 0 0of the 0 methodology
0 0 100 used 0 to develop the
age monthly per capita expenditures less household as-
asset scores in included in Annex C.
highest two asset categories (Table 31). Table 29: Average Monthly Consumption Expenditure per Capita e.
Comparing interventions, EAGA respond- by Intervention (Ksh) Dif-
ents generally have lower asset scores Intervention PARTI- CONTROL TOTAL fer-
than other groups; Fintrac HDC and Just CIPANT (Ksh) (Ksh) ences
Juice respondents have higher asset (Ksh) by
scores (Table 32). The data shows a lar- EAGA 4,900 2,900 3,900
ger proportion of EAGA respondents in SITE 5.200 4,800 5,000
the lowest asset category and a smaller KADI 16,000 -0- 16,000
FINTRAC 7,100 7,500 7,300
proportion of respondents in the highest
JUST JUICE 6,600 6,900 6,800
asset category compared to other inter-
TOTAL SAMPLE 6,500 5,200 5,800
vention groups. The data further shows a
larger proportion of Fintrac HDC and Just There are slightly more participants than gender of household head
Juice respondents in the highest asset controls in the lower size landholding Years of research in Kenya have shown that wo-
category compared to other respondent groups (two acres and below) and more
groups; and fewer respondents in the men headed households are more vulnerable
controls than participants in the higher than those headed by men. They tend to have
lowest asset categories. KADI and SITE size landholding groups (over six acres)
are in between. more limited access to productive resources
(Table 33). Comparing landholding across (land, labor, credit, technology, extension ser-
interventions, a higher proportion of re-
Landholding is another important house- vices, and other productive resources), and less
spondents in the EAGA and Just Juice
hold asset. It is important to note, samples – which are located in Central
extensive social networks. Many Kenyan wo-
however, that the sample covers a num- province -- are in the smaller landholding men who head households have both time and
ber of ecological zones, affecting the type size groups (Table 34). Given smaller av- spatial constraints, which combine to limit their
of land, its productivity, and average erage landholding size in these areas, mobility and participation in market activities
landholding size. For example, landhold- this is not unexpected. A higher propor- beyond a certain range of their homes. Women
ings in the fertile and populous areas of tion of the SITE, KADI, and Fintrac HDC who head households in geographically isolated
Central province (EAGA, Just Juice) typic- participants are in the larger landholding areas are particularly vulnerable. In general,
ally are smaller than landholdings in the size groups. This also is likely to be re- woman-headed households have fewer business
more arid areas of Eastern or Rift Valley lated to their geographic location (East-
provinces (SITE, KADI, Fintrac). Given this resources to draw upon than men headed house-
ern and Rift Valley provinces). holds, but are more dependent on these activit-
variation, the figures presented are not a
good comparable measure of wealth ies to meet their household needs (Ruth-Aspaas,
across interventions in the sample, but 2003). In light of these constraints, increased
more useful in comparing respondents participation in on-farm cash crop activities
within intervention samples. such as tree fruit production and sales has po-
tential to reduce the vulnerability of households
headed by women.
Table 30: Distribution of Re-
spondent
Households by Monthly Con-
sumption producers in Central province (Table sorb the risks associated with participat-
Women Expenditure per Capita 35). ing in the projects.
comprise Parti- Con- T
24 per- cipant trol otal As a point of comparison, Kenya’s The baseline research reveals some dif-
cent of 1 1997 Welfare Monitoring Survey ferences between men- and women-
the farm-BASE: Total ,94 (Government of Kenya, 2000) found headed households (Table 36). On aver-
ers in theSample 1,024 923 7 a significantly higher proportion of age, women-headed households have
sample None 1 1 1 women-headed households in rural fewer members, but earner- dependent
but head 1,000 and be- areas of Kenya – overall, 31.2 per- ratios are about the same as for men-
only tenlow 23 21 22 cent of poor rural households and headed households. A smaller proportion
percent 1,001 and 25.5 percent of non-poor rural of women-headed households have salar-
of the2,000 15 19 17 households included in this survey ied members and children in school than
house- 2,001 and were headed by women. The pro- men-headed households. In terms of pro-
holds –3,000 10 10 10 portion of poor rural households ductive resources, the average landhold-
12 per- 3,001 and headed by women in Eastern ing size for women- headed households is
cent in4,000 7 10 9 province was 33.6 percent, in Cent- almost half that of men, 6.3 compared to
the test 4,001 and ral province 32.3 percent, and in 11.5 acres. Average consumption ex-
sample 5,000 7 8 7 Rift Valley province 22.5 percent. penditure per capita per month is lower
and eight 5,001 and for participant women-headed house-
percent 6,000 7 5 6 This finding suggests that the tree holds than similar men headed house-
in the 6,001 and fruit projects may have reached a holds (Ksh. 4,700 compared to Ksh.
control 7,000 4 4 4 lower proportion of women headed 6,700). However, among controls con-
sample. 7,001 and households than in the general pop- sumption is higher for women- than for
The 8,000 3 4 3 ulation. One reason may be that men-headed households (Ksh 7,800 com-
largest 8,001 and project outreach strategies target pared to Ksh 4,900). This suggests that
share of 9,000 3 3 3 women farmers, but not explicitly the woman-headed households reached
women- 9,001 and women household heads, who are by the projects are poorer. A higher pro-
1 2 2
headed 10,000 harder to reach and recruit given portion of women-headed households is
house- 10,001 and their geographic and social isola- below the Ksh 2,000/day consumption
15,000 9 7 8 tion. Another reason may be that
holds, 19 figure, but the difference is slight (42 per-
15,001 and women who head households have
percent, cent compared to 40 percent) and may
20,000 4 2 3
is among less time available to participate in not be significant. A larger proportion of
20,001 and
avocado 50,000 producer groups and fewer labor re- women- compared to men-headed house-
5 2 4
sources to upgrade tree fruit pro- holds is in the lowest asset group (13 per-
50,000+ 1 1 1 duction. Given their vulnerability, cent of woman-headed households com-
Total 100 100 100 they may have less capacity to ab- pared to six percent of man-headed
house- Table 31: Distribution of Respondent less than Ksh.2,000 per
holds), Households by Asset Score month is highest for the
indicat- PARTICIPANT CONTROL TOTAL bottom two wealth groups
ing theirBASE: Total (68 percent and 53 percent
1,024 923 1,947
greater Sample respectively).
vulnerab- 1 (Low-
5 10 8
ility. This est)
gender 2 13 17 15
gap is 3 18 14 16
larger 4 13 14 14
among 5 10 10 10
controls 6 13 14 13
than par- 7 14 12 13
ticipants. 8
14 9 12
There is(Highest)
little dif-Total 100 100 100
ference Mean 4.7 4.3 4.5
between
women- and men-headed households in
the proportion of households in the top
asset category.
f. Differences by Socio-economic
Status
Earner-dependent ratios, average number
of income sources, and the importance of
tree fruit income within households are
similar across wealth levels (Table 37).
However, as expected, consumption ex-
penditures increase with wealth level. Av-
erage monthly per capita consumption
expenditure rises steadily from Ksh.3,000
for households in the bottom asset cat-
egory, to Ksh.11,200 for households in
the top asset category. The proportion of
households with per capita expenditures
g. Intra-household Issues
A number of intra-household factors can
affect the supply response of farmers in
the tree fruit value chain and provide a
context for understanding impacts. With-
in households, who assumes responsibil-
ity for doing the work, who makes man-
agement decisions, and who controls the
earnings can all influence decisions re-
garding the investment of time and re-
sources in tree fruits. The tree fruit value
chain is characterized by a gender differ-
ences
Table 32: Distribution of Respondent Households by Asset Score in all
and Intervention these
areas.
Fintrac
EAGA SITE KADI Just Juice
HDC
Participant
Participant
Participant
Participant
Participant
Control
Control
Control
Control
Control
Total
Total
Total
Total
Total
BASE: Total 35 20 14 15 29
Sample 250 250 500 0 349 699 70 0 706 173 379 8 1 9
1 9 28 19 6 6 6 11 0 110 2 1 1 1 1
2 23 28 25 15 15 15 14 0 144 9 6 5 11 8
3 17 17 17 26 17 21 11 0 11
10 8 9 15 11 13
4 12 8 10 14 17 16 20 0 20
13 13 13 13 15 14
5 12 6 9 9 13 11 4 0 4
11 9 10 14 11 12
6 10 7 8 10 14 12 13 0 13
18 16 17 16 21 18
7 8 4 6 11 12 11 10 0 10
22 20 21 19 19 19
8 10 2 6 9 7 8 16 0 16
21 23 22 19 12 15
10 10 10 10 10 10
Total 100 100 100 0 100 100 0 0 100 0 100 100 0 0 0
Mean 4 2.8 3.4 4.2 4.4 4.3 4.5 0 4.5 5.8 5.6 5.7 5.5 5.2 5.3
Passion fruit is mostly a woman’s crop vest season for other crops. Women’s income from the passion goes to ‘obvious
and women make most of the decisions “ownership” of passion fruit is reflected in things’ like school fees and paying off
regarding production and sale. Men de- the comment of a farmer in Eldoret who loans. If there is a surplus after covering
cide on the initial investment; but women accused a man of “stealing his wife’s fruit these expenses, husbands and wives
make most other decisions. They also do and selling it for drink”. Women are the jointly negotiate the use of the surplus.
most of the work. Men dig holes for the first to receive income from sales to
plants and install the poles and wire (one- brokers, and generally control this in- For avocados, men pick and prune the
off tasks) while women maintain the come. According to one woman farmer, trees and women do most other mainten-
vines, ance work. If there is a shortage of
harvest Table 34: Distribution of Respondent Households by Size of Landholding male labor within the family, wo-
the fruit,by Intervention men will hire male labor for picking
and play EAGA SITE KADI Fintrac HDC Just Juice fruit, but they also face labor short-
a key role ages. Although women do most of
Participant
Participant
Participant
Participant
Participant
Control
Control
Control
Control
Control
in sales the work cultivating avocados,
Total
Total
Total
Total
Total
to group payments are in the name of
brokers. their husbands or sons. Women tra-
Men play ditionally are responsible for nego-
BASE: Total
a role inSample 250 250 500 350 349 699 70 0 70 206 173 379 148 151 299 tiating with brokers, so they may
selling to Upto 0.5 acre 1 2 1 1 - n 1 - 1 -1 - 1 5 15 11 lose some control over avocado in-
export- 0.51 to 1.0 come in working through groups.
ers. Wo-acre 18 7 13 5 n 3 3 - 3 6 1 4 26 22 24 However, when asked about de-
men 1.1 to 2.0 cision-making, both men and wo-
some- acres 38 36 37 10 2 6 9 - 9 13 12 12 18 25 21 men group members said that hus-
times 2.1 to 4.0 bands and wives make decisions
hire maleacres 29 36 33 25 15 20 19 - 19 17 13 15 30 19 24 together about money earned from
labor for 4.1 to 6.0 avocados.
picking, acres 8 12 10 20 13 17 34 - 34 21 17 20 13 11 12
although 6.1 to 10.0
they of-acres 5 6 6 17 27 22 14 - 14 13 23 18 7 4 6
ten face 10.1 to 20.0
labor acres 2 n 1 12 19 16 14 - 14 14 15 14 1 2 1
short- 20.1 + acres - - - 11 23 17 6 - 6 16 17 16 - 1 n
ages, es-Total 100 100 100 100 100 100 99 - 99 99 99 99 99 99 99
pecially Average (acres) 2.6 2.8 2.7 7.3 10.8 9 6.9 0 6.9 8.1 9.1 8.6 2.7 2.5 2.6
during Standard Devi-
the har-ation 2.4 2 2.2 6 6.4 6.4 5.3 0 5.3 6.7 6.5 6.6 2.3 3 2.7
Table 33: Distribution of Respondent In
Households by Size of Landholding
PARTICIPANT CONTROL TOTAL
BASE: Total
1,024 923 1,947
Sample
Upto 0.5
1 4 2
acre
0.51 to
11 6 9
1.0 acre
1.1 to 2.0
18 17 18
acres
2.1 to 4.0
25 21 23
acres
4.1 to 6.0
17 13 15
acres
6.1 to
12 17 14
10.0 acres
10.1 to
8 11 9
20.0 acres
20.1 +
7 12 9
acres
Total 100 100 100
Average
2.9 2.8 2.9
(acres)
Standard
1.6 1.6 1.6
Deviation
mango farming, men appear to play a
more central role in crop management
and upgrading activities. Brokers are re-
sponsible for picking mangos (the trees
are tall) and factor this into their prices.
Households sometimes hire outside labor
if there is excess production or they are
involved in non-broker sales. Men and
women make joint decisions at the farm
level, and farmers say that payments
[from brokers] may be received by any-
one in the household, even children.
Household members generally discuss
ahead of time a ‘floor’ price they will not
sell below. Once income is in household,
men and women decide jointly on ‘big’
expenditures, while other expenditures
are routine and don’t require discussion.
Although women play a secondary role in
mango farming in some households, they
said they have a fairly good idea of how
much money is coming in.
33
The survey teams were careful NOT to tell those inter-
viewed that the study concerned the impact of the tree
fruit projects. However, given the number of questions
about tree fruit, it would be natural for respondents to as-
sociate the study with the USAID projects, which are rel-
atively high profile.
IV. CONCLUSIONS OF THE BASELINE STUDY AND IMPLICATIONS FOR ROUND TWO OF THE
IMPACT ASSESSMENT
This impact assessment is designed to Broadly speaking, the baseline research Vertical links to higher-value markets
test the hypotheses enunciated on pages shows that smallholders are part of the provide critical incentives for tree fruit
16-18, above. The general hypothesis is tree fruit value chain, but they occupy a producers to upgrade. So far, only one of
that the activities of the Kenya BDS and low position within that chain. They are the interventions studied – the EAGA avo-
Fintrac HDC projects can be effective in numerous and active producers, but their cado intervention – has begun to realize
opening up opportunities for smallholder productivity is low and they sell much of this potential by forging a direct link from
MSEs in local, regional, and global mar- their produce under unfavorable condi- farmers to the European market. This has
kets and in improving the competitive- tions. Income from tree fruits plays an im- required considerable ‘hand holding’ by
ness of Kenya’s tree fruit value chains. portant role as a source of household in- Kenya BDS and has involved an exporter
More specifically, the impact assessment come, especially for the poorer farmers, that has received other support from
hopes to identify and measure improve- but income from tree fruit and total USAID to help prepare smallholders to
ments that lead to greater integration of household income are both very low in meet EUREPGAP standards.
smallholder MSEs into the value chain, to most cases.
improved enterprise performance and Brokers remain alive and well in all three
household well-being, and to improved To varying degrees, the five interventions fruit value chains and continue to be im-
competitiveness and growth of the tar- included in the baseline study succeed in portant marketing channels for many
geted value chains. reaching low-income farmers. This means farmers.
that there is potential for direct impact by
The study is longitudinal and has two raising their incomes through the pro- It is too soon to tell whether the project
stages. This report covers the first, or jects. activities will result in "sustainable solu-
baseline, stage. It examines the condition tions" to the recurrent needs of tree fruit
of the value chain and a sample of small- Producer groups are an important part of producers. This includes both embedded
holders and others involved in the pro- this potential because they provide hori- and stand-alone solutions/services that
duction and marketing of tree fruit at an zontal linkages that give poor farmers a provide inputs, TA, or market access. In
early stage of implementation of the two chance to link to export markets – some cases changes might take place
projects. The main purpose of the something they have very little opportun- due to direct provision by the projects but
baseline study has been to establish a ity to do by other means. The projects it remains to be seen if embedded ser-
standard against which change can be have been instrumental in organizing and vice arrangements, the commercializa-
measured two years later. strengthening tree fruit producer groups. tion of nursery and extension services, or
the “network broker” concept of EAGA
and Kenya BDS will last once the project at the specific services/solutions under-
activities end. taken during the course of the projects.
Time for these activities should be in-
Annex B (below) provides a checklist of cluded in the budget for the second-
issues that should be addressed in the round assessment. Provision should also
follow-up survey. A general point is that be made for reviewing the monitoring
while scrupulous efforts were made to se- data that are being collected by both pro-
lect control group samples for the jects.
baseline survey that were comparable to
the participant samples, the findings
cited in this report show that at the time
of the survey the participants as a group
were significantly better-off and more
productive -- in many cases and in sever-
al ways -- than the controls. When each
group is resurveyed two years hence,
care will need to be taken in analyzing
the results to ensure that differences in
household wealth and other mediating
variables are taken into account in de-
termining the impact of the programs.
Dunn, Elizabeth and Lillian Villeda. 2005. “Weaving Micro and Small Enterprises
into Global Value Chains: The Case of Guatemalan Textile Handicrafts.” AMAP pa-
per.
FKAB Feldt Consulting. 2001. “Sector Study of the Horticultural Export Sector of
Kenya. A Study Made on Behalf of USAID, Kenya.”
Government of Kenya. n.d. Kenya Interim Poverty Reduction Strategy Paper 2000
– 2003. Prepared by the Government of Kenya. World Bank (no Date)
Giuliani, Elisa, Carlo Pietrobelli and Roberta Rabellotti, 2004. "Upgrading in global
value chains: lessons from Latin American clusters," Working Paper 72, SEMEQ
Department, Faculty of Economics, University of Eastern Piedmont.
HCDA. 2003. Export Statistics for Fruits, Flowers, and Vegetables for 2003.
Humphrey, John and Hubert Schmitz. 2003. “Chain Governance and Upgrading:
Taking Stock.” In Hubert Schmitz, ed. Local Enterprises in the Global Economy:
Issues of Governance and Upgrading.
Jaffee, Steven, and Spencer Henson. 2004. “Standards and Agro-Food Exports
from Developing Countries: Rebalancing the Debate.” World Bank Policy Re-
search Working Paper 3348.
Kula, Olaf, “Activity Status Report: Holding Hands with Folded Arms: Upgrading
Kenya
Tree Fruit Value Chains”. Deloitte, Touche, Tomatsu, No date.
Lipton, Michael. 1994. “Growing Points in Poverty Research: Labour Issues.” Insti-
tute of Development Studies, Sussex University.
McCulloch, Neil and Masako Ota. N.d. “Export Horticulture and Poverty in Kenya.”
Institute of Development Studies, University of Sussex.
Minot, Nicholas and Margaret Ngigi. 2003. “Are Horticultural Exports a Replicable
Success Story? Evidence from Kenya and Cote d’Ivoire.” Paper presented at the
InWEnt, IFPRI, NEPAD, CTA Conference, “Successes in African Agriculture,” Pre-
toria, December 1-3.
86
Morton, Julia. 1987. Fruits of Warm Climates. (chapter on passion fruit. p. 320–
328). Miami, FL
Muendo, Kavoi Mutuku, David Tschirley, and Michael T. Weber. 2004. “Improving
Kenya's Domestic Horticultural Production and Marketing System: Current Com-
petitiveness, Forces of Change, and Challenges for the Future. Volume II: Horti-
cultural Marketing.” Nairobi: Tegemeo Institute of Agricultural Policy and Devel-
opment, Egerton University.
Tschirley, David and Mary Mathenge. 2003. “Developing Income Proxy Models for
Use by the USAID Mission in Kenya: A Technical Report.” Tegemeo Working Paper
7. Nairobi: Tegemeo Institute of Agricultural Policy and Development.
Wallis, William. 2004. “African Farmers Dig in to Comply with EU Food Rules.” Fin-
ancial Times, April 7.
87
ANNEX A. DESCRIPTION OF KENYA BDS AND FINTRAC
HDC PROJECT ACTIVITIES
KENYA BDS
Facilitate improved
transport of goods to
Nairobi
88
KENYA BDS
Information campaign
to sensitize producer
groups to value of ex-
tension services
89
KENYA BDS
90
KENYA BDS
91
ANNEX B: ISSUES TO FOLLOW UP IN ROUND TWO
Avocado
Mango
92
- The extent to which cell phones are used to improve information flows within
the value chain.
Passion fruit
- Assessment of the types of activities in which they are involved and the
extent/degree of management responsibility that the group assumes for particu-
lar activities vis-à-vis other value chain actors.
93
ANNEX C: CALCULATION OF THE ASSET SCORE GROUP-
INGS
1. The following attributes were selected to indicate a household’s standard of
living:
The philosophy guiding this procedure is that no single question or answer on its
own can adequately summarize a reality as complex as standard of living.
Hence, many variables were used in conjunction with each other to contribute to
the discovery of a single measure of standard of living.
2. The attributes listed above were than factorized into a single factor using
SPSS to find a single unifying theme. The output of this factorization was that
each respondent received a single factor score based on his/her responses to
questions about ALL the listed attributes. Each respondent’s factor score de-
scribes the respondent’s relationship to the extracted theme. Thus the respond-
ents’ factor scores form a continuum from the lowest value to the highest value.
The continuum of values was then cut into eight roughly equal groups, one being
the group with the lowest values and eight being the group with the highest val-
ues. The eight groups are not exactly equal in size because we wanted to
achieve a good spread of each farmer group across several asset score groups.
94
The factor was then cross-tabulated against attributes that were outside those
listed above to determine whether the theme described by the single factor
really was a measure of living standard. The theme described by the single
factor did correlate with the distributions of these other, unused validation vari-
ables (and this is reflected wherever we see the asset score cited in the report)
and thus the factor was deemed suitable to describe the standard of living.
4. Because the predictor scores do not match the factor scores exactly (al-
though they are extremely close), it follows that the “predictor” groups cannot
be exactly the same as the original “factor” groups. Thus, the data were run
again using the predictor attributes, and this is what is then used in analysis for
this round of the survey and will be used again in the next round to ensure com-
parability between the two rounds. These are the asset score groupings that are
shown in this report.
95