|
TECHNOLOGY DISSEMINATION AMONG SMALL-SCALE FARMERS IN MERU
CENTRAL DISTRICT OF KENYA: IMPACT OF GROUP PARTICIPATION
By
KRISTIN ELIZABETH DAVIS
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2004
Copyright 2004
by
Kristin Elizabeth Davis
This dissertation is dedicated to my parents, Art and Mary Ellen Davis.
ACKNOWLEDGMENTS
I would like to thank the Ministry of Livestock Development and Fisheries in Meru
Central District for all of their help with my research. Henry Kirimi, the District
Livestock Production Officer, and his staff assisted in many ways. Divisional Livestock
Extension Officers and locational staff from Abothoguchi Central, Abothoguchi East, and
Miriga Mieru East divisions were especially helpful. Patrick Mutia and all of the FARM-
Africa staff are also owed a huge thank you for allowing me to make use of their
personnel, vehicles, and facilities in Meru, in order to facilitate my research. Special
thanks also go to my assistant Franklin Muthamia who worked tireless hours with me,
translated, repaired vehicles, and took care of logistics. Ben and Darlene O'Dell
provided much-needed moral support while in Meru.
Steve Franzel of the World Agroforestry Centre was a brilliant supervisor during
my time in Kenya, and together with Charles Wambugu provided much advice. I would
also like to thank the United States Agency for International Development, especially
Andrew Brubaker and Scott Bode, for funding that enabled me to conduct the field
research. Sarah Workman, the Scientific Liaison Officer between USAID and World
Agroforestry Centre, was instrumental in procuring the funding. I would also like to
thank Charity Blomeley, Betty Finn, and Lisette Staal from International Programs at the
University of Florida, for their assistance in the funding administration. I also must thank
Madhur Gautam, Pierre Rondot, and Willem Zijp from the World Bank; and John
Sanders from Purdue University for resources and feedback.
Special thanks also go to my committee members. Pete Hildebrand was always
available for advice and discussions on small-scale livelihood systems. Christy Gladwin,
during her time on my committee, gave excellent insight into economics in Africa and
ethnographic techniques. Nick Place gracefully and patiently bore with all my questions
and dilemmas as my academic advisor. Tracy Irani provided great help and guidance in
methodology and statistics. Finally, Della McMillan Wilson helped me with critical
materials, with contacts in Kenya, and with meals and moral support.
I would like to thank my parents and family for their support and continued belief
in me. Finally, I acknowledge my Lord and Savior Jesus Christ, who is the one who
guided me to the University of Florida and this particular research study. It has all been
for Him.
TABLE OF CONTENTS
page
A CK N O W LED G M EN TS ................................................................................................ iv
LIST O F TA BLES ............................... .......... ........................................................ x
L IST O F FIG U R E S ........................................................................................................ xiii
A B ST R A C T ..................................................................................................................... xiv
CHAPTER
I IN TR O D U C TIO N ........................................................................ ...................... .......
Introduction ................................................................................................ ..........
Background to the Problem .................................................................................. 2
Smallholder Agriculture ................................................................................ 3
Extension's Approaches to Rural Development.............................. ..............5
Study-A rea B background ...............................................................................................9
The World Agroforestry Centre (WAC) ........................................................ 14
Food and Agricultural Research Management (FARM)-Africa .......................15
Researchable Problem ......................................................... .............................. 16
Purpose and Objectives.............................. ......................................................17
R research Q uestions............................................................................................... 18
A ssum options .................................................. ..................................................... 18
Operational Definition of Terms ............................................................................... 18
Lim stations of the Study ....................................................................................... 20
Significance of the Study...................................................................................... 20
O organization of Thesis........................ ............................................................. 21
2 LITERATURE REVIEW ...............................................................23
Extension History and Models in Kenya..................... .. ..............................23
Introduction ........................................... ..................................................... 23
Diffusion of Innovations Theory and Transfer of Technology .........................25
The 1960s Top-Down Approach: State Transfer of Technology Model ...........27
The 1970s Holistic Approach: Integrated Rural Development Projects and
Farming Systems Philosophy...................................................................28
The 1980s Training and Visit Approach: Expanding the State Extension
Service................................................................ ...................................... 30
The 1990s Pluralistic Approach: Community-Based Farmer-Led Extension....34
Farm er-Led Extension ........................................................................................... 35
S social C apital............................................. ... ......... ...... ............................. .. 39
Farm er G groups ........................... ......... ................................................ 43
Role of Farmer Groups in Extension..................................................44
Factors for G roup Success....................................................... .........................49
C on clu sion ...................................................... .......................... ............................52
3 M ETH O D S ............................................ ............................................................54
Research D design ...................... ........ ........... ..... ...................................... 55
Population and Subjects..................................................................................56
Sam pling Procedure............................................................................................57
Instrum ents ........................................................................................................... 59
D ata C collection ..................................................................................................... 62
D ata A nalysis........................................................................................................ 66
Validity and Reliability of the Results................................................................ 67
V validity ........................................................................................................ 67
R liability ........................................ ............................................................70
4 R E SU L T S ...........................:................................................................................74
The Food and Agricultural Research Management (FARM)-Africa Project and
D airy-G oat G roups........................................................................................... 74
Fodder Training and Dissemination in the Project..............................................80
Description of Area and Subjects .....................................................................85
Objective One: Examine Participation in Groups and Identify What Factors
Affect Participation in Groups ....................................................................................89
Introduction .......................................... ..................................................... 89
Groups in Meru Central District............................... .........................................89
Factors Affecting Participation in Any Groups in Meru Central District ...........93
Issues Regarding Participation in Dairy-Goat Groups.................................... 100
Lack of knowledge/information ............................................................... 102
Wealth and poverty ........................... ............................................102
Social problem s .................................................................................... 112
T im e ............................................ ... ...... 113
G ender ........................... .... ............. .... ........ ............................... 113
Household composition.......................................................................... 114
Other reasons........................................................................................ 115
Participation in leadership................................................................... 117
Withdrawal from groups ............................................................................118
Objective Two: Examine Linkages and Their Outcomes among Farmers and
Other Extension Stakeholders........................................................................... 118
Introduction .............................................. .. ...... ....................... 118
Agricultural Players in Meru Central District .................................................119
Dairy-Goat Group Linkages............................................................................ 124
Linkage Outcomes.....................................................................................126
vii
Objective Three: Identify the Mechanisms by Which Farmer Groups and Their
Members Receive and Disseminate Information and New Technologies .........131
How Dairy-Goat Groups Disseminate Information and Technology.............. 131
How Dairy-Goat Groups and Individuals Receive Information and
Technology ......................................................... ............................... 138
Sources of Information for Dairy-Goat Groups ............................................. 138
Sources of Information for Individual Farmers...................................................140
Objective Four: Identify the Factors Characteristic of Groups Successful in
Disseminating Technology .................................................................142
G group L ocation............................. ... ..... ... ....... .. ......................................145
Group Size and Member Participation .........................................................150
G group A ge ........................................ ........................................................ 154
Formality and Management.................................................................154
L leadership ................................................................................................... 156
A ctivities.................................................................................................... 157
Gender, Poverty and Individuals in the Group................................................ 157
Homogeneity of Members ................................................................................159
G group C cohesiveness .......................................................................................... 160
C capacity .................................................... ................................................ 161
L inkages .....................................................................................................163
Type of Group (Project-Facilitated versus Non-Facilitated)...........................164
G roups' Ow n Indicators ................................................................................ 170
5 CONCLUSIONS AND RECOMMENDATIONS ...................................................174
Objective One: Examine Participation in Groups and Identify What Factors
Affect Participation in Groups.............................................................................175
Are Poorer People Participating in Groups in Meru?........................................ 175
What Affects Participation in Groups of Any Kind? ........................................181
Objective Two: Examine Linkages and Their Outcomes between Farmers and
Other Extension Stakeholders.......................................................................... 183
Objective Three: Identify the Mechanisms by Which Farmer Groups and Their
Members Receive and Disseminate Information and New Technologies...........188
Objective Four: Identify the Factors Characteristic of Groups Successful in
Disseminating Technology .......................................................................... 192
Sum m ary ........................................ ................................................................. 204
Use of Groups in Extension....................................... ............................ 204
How Do Groups Fit in a Pluralistic Extension System?....................................208
Recommendations for Policymakers and Practitioners ..........................................211
Recommendations for Further Research .......................... .............................213
C on clusion ............................................................................... .......................... 2 13
APPENDIX
A ADDITIONAL DATA ON GROUP PARTICIPATION.........................................215
B SEMI-STRUCTURED INTERVIEW TOPIC GUIDE...........................................221
C LETTER OF INVITATION TO INTERVIEW FOR GROUPS.............................. 225
D INTERVIEW SCHEDULE FOR GROUPS......................... ............................226
E INTERVIEW SCHEDULE FOR INDIVIDUALS ...................................... ........... 232
F INFORMED CONSENT.......................... ....... .. .........................238
LIST O F R EFEREN CE S .................................................................................................239
BIO G RA PH ICA L SK ETCH ...........................................................................................250
LIST OF TABLES
Table page
1-1. Contribution of agriculture to gross domestic product (GDP) ................................3
1-2. Rural population in Kenya, 1980-2000 ............................................... ................ 4
2-1. Early extension approaches strengths and weaknesses ...........................................28
2-2. Philosophy of transfer of technology (TOT) and farmer first.................................37
4-1. Source of fodder training for the dairy-goat groups (n = 46)..................................82
4-2. Descriptive data for individual farmers (n = 88) ...............................................85
4-3. Descriptive data for individuals interviewed (n = 88).................................... ...88
4-4. Types of groups to which individual respondents belonged (n = 88) .....................90
4-5. Factors affecting participation in dairy-goat groups (n = 88)..................................94
4-6. Binary logistic regression analysis showing factors associated with membership in
dairy-goat groups................................................................................................96
4-7. Reasons given by individuals for joining most important group (n = 86).................99
4-8. Criteria for being a member of dairy-goat groups (n = 46)...................................100
4-9. Most frequent reasons given for why some farmers are not in any groups (%)......101
4-10. Pearson's product moment correlations between various indicators for wealth for
individual farmers (n = 88)................................................................................... 104
4-11. Frequencies for wealth indicators of dairy-goat group members and non-members
(% ) (n = 88) ........................ ................... ....................................... 106
4-12. Wealth levels of dairy-goat members (divided into type of member) and non-
m em bers .......................... .... ........ .......................... ..........................107
4-13. Number of improved cattle owned by various wealth levels (n = 88) ................109
4-14. Individuals' responses to participation in groups...................................... ........... 111
4-15. Dairy-goat groups' responses to participation in their group (n = 46) ..................11
4-16. Frequencies for group leaders' gender, age and wealth level (%) (n = 138)......... 117
4-17. Agricultural players of Meru Central District .................................................120
4-18. Sizes of chapatis given to various links with the dairy-goat groups (n = 46) .......125
4-19. Frequency of linkage with the dairy-goat groups as shown in chapati diagrams
(n = 46) ....................... ... .. ..... .... ...... ............... .................................... .... 126
4-20. Amount of fodder seedlings given by dairy-goat groups to non-members...........136
4-21. Most important sources of information listed by dairy-goat groups (n = 46) .......138
4-22. Number one source of information for dairy-goat groups by type of group (%) .. 139
4-23. Most important sources of information listed by individual farmers (n = 88)......141
4-24. Pearson's product moment correlations between variables affecting success (n =
4 6) .................................................................................................................. 14 6
4-25. Pearson's product moment correlation coefficients between location and success
indicators (n = 46) ............................................................................................ 148
4-26. Effect of high and low group elevation on success indicators (n = 46)...............150
4-27. Effect of low and high group age on success indicators (n = 46)........................154
4-28. Pearson's product moment correlations between variables affecting success and
wealth indicators for groups (n = 46)................................................................... 159
4-29. Pearson's product moment correlations between cohesiveness and success
indicators ........................................................................................................ 160
4-30. Effect of low and high group training on success indicators (n = 46)................. 162
4-31. Differences in FARM- and extension-facilitated dairy-goat groups (n = 46)....... 165
4-32. Effect of type of group on success indicators (n = 46)........................................167
4-33. Linear regression analysis of variables for prediction of success in dissemination
(adoption index) ..................... .. .................. .. ...........................................169
4-34. Effect of various group factors on success indicators based on t-tests................70
5-1. Studies examining the effect of group factors on success.....................................204
A- Factors affecting participation in church groups..................................................215
A-2. Binary regression analysis for participation in church groups (n = 39) ...............215
A-3. Factors affecting participation in clan groups ................................................216
A-4. Contingency table on gender and participation in clan groups...............................16
A-5. Binary regression analysis for participation in clan groups (n = 23)....................216
A-6. Factors affecting participation in merry-go-rounds..............................................17
A-7. Contingency table on gender and participation in merry-go-rounds....................17
A-8. Binary regression analysis for participation in merry-go-rounds (n = 18).............217
A-9. Factors affecting participation in water groups ..............................................218
A-10. Contingency table on gender and participation in water groups ........................218
A-11. Contingency table on water source and participation in water groups (n = 18)...218
A-12. Contingency table on wealth level and participation in water groups (n = 18)....219
A-13. Binary regression analysis for participation in water groups (n = 18) ................219
A-14. Factors affecting participation in women's groups ............................................219
A-15. Binary regression analysis for participation in women's groups (n = 14)............220
LIST OF FIGURES
Figure page
1-1. Meru Central District in Kenya ...................... ........................................ 11
2-1. Farming systems emerging methodology on technology transfer...........................26
4-1. Dairy-goat group distribution within Meru Central District ...................................79
4-2. Distribution of fodder planted by individual farmers (n = 88)................................83
4-3. Distribution of improved goats owned by individual farmers (n = 88)...................84
4-4. Distribution of years of education by individual farmers (n = 88)..........................89
4-5. Distribution of number of animals owned by individual farmers (n = 88)................89
4-6. Chapati diagram by a dairy-goat group ............................... ....... 125
4-7. B uck station ........................................ .. ............ ........... ...... ........................... 134
4-8. Agroecological zones in Meru Central District.................................................... 147
4-9. Natural resources in M eru ................................................... ............................ 151
4-10. Roads and towns in Meru Central District ..........................................................152
4-11. A main road leading out of Meru town with vehicles stuck in the mud..............152
4-12. Elevation in M eru Central District ...................................................................... 153
4-13. Groups' self-ratings on success in dissemination (%)........................................... 171
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
TECHNOLOGY DISSEMINATION AMONG SMALL-SCALE FARMERS IN MERU
CENTRAL DISTRICT OF KENYA: IMPACT OF GROUP PARTICIPATION
By
Kristin Elizabeth Davis
August 2004
Chair: Nick Place
Cochair: Peter Hildebrand
Major Department: Agricultural Education and Communication
The goal of this study was to examine the role of farmer groups in technology
dissemination, and to assess what factors make groups effective in extending
technologies among small-scale dairy-goat farmers in Meru Central District of Kenya.
The theoretical framework for the study included insight from agricultural extension
theory, farming systems research and extension, social capital, and group theory.
A mixed-methods, multiple-stage approach was used to obtain data. Research
techniques included participant observation, documentary analysis, semi-structured
interviews, social mapping, group timelines, and structured questionnaires. Dairy-goat
farmer groups (n = 46) and individual farmers (n = 88) were interviewed during the
study. Qualitative data provided baseline information, and helped in the formation of
research questions. Quantitative data were entered into the Statistical Package for Social
Scientists (SPSS) and analyzed using contingency tables, descriptive statistics,
correlations, tests of significance, and regression.
Most farmers in Meru Central District belonged to some type of farmer group.
Factors for participation in different types of groups included household size, wealth
level, age, gender, and membership in other groups. The dairy-goat groups were linked
with an average of nine outside entities. Major linkages included government extension,
chief baraza (public meeting), churches, and the non-governmental organization Food
and Agricultural Research (FARM)-Africa. Dairy-goat groups had a variety of
mechanisms for disseminating information and technology, including the baraza, the
buck (breeding) station, through training others, and in other groups that members
belonged to. Certain factors about the dairy-goat groups were associated with success in
dissemination. These included type of group (those facilitated by the dairy-goat project
versus those not), location, age of group, number of training, homogeneity of members,
and number of linkages.
In the pluralistic extension milieu today, farmer groups play an increasingly
important role in the technology-dissemination process. Most farmers in Meru were in
groups, which were linked to other organizations and were disseminating the information
and technologies that they had. Groups are a key way for farmers to receive information
and training, and then to tell or train others. Some ingredients that might increase the
success of such groups include increasing capacity in groups (and especially of key
members), finding ways to link them with other extension providers, strengthening
government administration in the form of baraza, and encouraging groups to form around
common interests at the community level.
CHAPTER 1
INTRODUCTION
A key feature distinguishing extension work from other forms ofprofessional agriculture
is that in the first instance extension deals with people rather than with their crops and
animals.
(Moris, 1991, emphasis added)
Introduction
Government extension in Kenya today is unable to provide many small-scale
farmers with pertinent technologies and information to meet their needs and thus help to
bring about rural development, one of the acknowledged goals of the administration.
Lack of proper extension services is partially to blame for poverty, according to
participatory poverty assessments conducted in ten districts in Kenya in 2000 (Meru
Central District Development Plan, 2002; Republic of Kenya, 2001). This is due to both
reductions in government services in Kenya and ineffective and inappropriate extension
approaches (Eponou, 1996; Gautam, 2000). These issues have led to gaps in extension of
technologies to small-scale farmers, who play a major role in the Kenyan economy.
Technologies to address rural problems have been developed by research,
development organizations and farmers working together in Kenya. A major issue now is
how to extend, or scale up, these technologies to benefit more low-resource farmers
despite the limited government extension resources. Many approaches to technology
dissemination have been developed since the reduction of the state extension service,
such as private extension services and those run by non-governmental organizations
(NGOs). Recently, however, community-based extension has come to the fore as a
means of scaling up these technologies to have a wider impact in the rural economies
(Franzel, Cooper & Denning, 2001; Misiko, 2000; Noordin, Niang, Jama & Nyasimi,
2001). Farmer groups are an important vehicle for such community-based extension.
Background to the Problem
Sub-Saharan Africa is well known for its high poverty levels and other major
obstacles to development. In addition to the limited resources, risk and complexity
inherent in these African livelihood systems, small-scale farmers in Kenya today must
also deal with population pressure, the effects of HIV/AIDS, environmental degradation
and poor rural services and infrastructure. In contrast to other developing regions, the
problems in this particular part of the world do not seem to be improving. Per capital
agricultural production in sub-Saharan Africa has been in decline since the mid 1990s
(Shapouri & Rosen, 2001). As a result, living standards have also declined in the region.
Kenya perhaps typifies the experience of sub-Saharan Africa. With 70% of
Kenya's workforce involved in agriculture, and smallholders contributing to a significant
portion of agricultural production, Kenya makes a good case for the examination of the
government extension's relationship with the often-ignored smallholder farmers.
Kenya's extension service, much like that of other countries in sub-Saharan Africa, has
gone through many changes since its original inception through the colonial government,
in response to the changing social, environmental, and political settings.
It has long been realized, in Kenya, that a strong agricultural sector is crucial to
growth (Pearson et al., 1995). Agriculture contributes a significant proportion to its gross
domestic product (Table 1-1). Because of this, the Kenyan government has taken a
strong interest and role in agricultural services (Umali & Schwartz, 1994). Since the
early imposition of colonial rule, the government has thus made agricultural extension a
major focus of its administration.
Table 1-1. Contribution of agriculture to gross domestic product (GDP)
Year % Share of GDP
1964 40
1973 34
1978 37
1985 31
1996 25
2001 21
2002 16
Note: From Godfrey, 1986; Nyangito & Okello, 1998; World Bank, 2004
Smallholder Agriculture
In Africa, many of the farms-in fact, most of the rural population-are made up
of smallholdings of less than two hectares (Moris, 1991). These smallholdings are for the
most part diverse and characterized by limited resources. This brings particular problems
to extension and other service providers, who have typically been geared to make blanket
extension recommendations based on technologies designed for larger, more modem and
uniform farms.
Perhaps no one has as adequately described smallholder farmers as Chambers
(1997). Smallholder agriculture, in contrast with large-scale "modem" farming, is
complex, diverse and risk-prone. To these farmers, farming is not a business but rather a
means of achieving a livelihood. Although often ignored by government and other policy
makers, these small-farm livelihood systems are home to perhaps half the world's
population, contribute significantly to agricultural production and are an important source
of rural employment. Furthermore, although the percentage of the rural population is
decreasing in comparison to the overall and urban populations, in absolute numbers the
amount of small-scale farmers is still on the increase (Table 1-2).
Table 1-2. Rural population in Kenya, 1980-2000
1980 1990 2000
Rural population (1000) 13,738 17,914 20,435
Total population (1000) 16,368 23,574 30,669
Percentage 84 76 67
Note: From FAOSTAT Data, 2004
According to Moris (1991, p. 14), smallholder farming characteristics include
* Scattered clientele in remote areas;
* Varied crops in diverse agroecosystems;
* Resource constraints of farmers;
* Highly seasonal and risky agriculture; and
* Low reliability of services.
Of Kenya's population of about 30 million, 80% are located in rural areas and
dependent on agriculture for a living-much of this on smallholdings less than two
hectares (Muturi, 2001). These farmers, however, are key in the government's goal of
reducing poverty. They contribute significantly to the country's agricultural production;
smallholders produce 70% of Kenya's maize, 65% of the coffee, 50% of the tea, 80% of
the milk, 70% of the beef, 100% of pyrethrum, and many of the food crops (Muturi,
2001). However, many (if not most) of these smallholders are not being adequately
reached by extension services (Eponou, 1996; Gautam, 2000). Extension and other
development organizations must find effective means of enabling these small-scale
farmers to reach their full potential and thus contribute to the overall economy of Kenya.
Farming-systems philosophy gives a useful perspective on the smallholder
livelihood systems. A hallmark of farming systems is concern for the small-scale farmer
and the understanding of her or his situation. It also recognizes the diversity among
smallholder farmers, their various systems of farming, and their livelihood strategies; and
it attempts to understand that diversity. A farming-system livelihood strategy can be
defined as the way that a farm family manages its resources to meet household objectives
within social, economic and physical systems (Franzel, 1984). As seen from livelihood
analysis, resources or assets in small farm livelihood systems can be quite complex
(Dorward, Anderson, Clark, Keane & Moguel, 2002). One such resource is known as
social capital or collective action, and involves the networks and relationships that a
farmer may call on to meet her or his objectives (de Haan, 1999; Narayan & Pritchett,
1999). One type of social capital is farmer groups (mentioned in the introduction as a
vehicle for community-based extension).
Extension's Approaches to Rural Development
To meet the challenges faced in the rural sector (mentioned in the previous section),
agricultural research and extension have been used for decades to improve the rural
economy. Agricultural extension, according to the World Bank, is "the process of
helping farmers to become aware of and adopt improved technology from any source to
enhance their production efficiency, income and welfare" (Purcell & Anderson, 1997, p.
55). Agricultural extension has also been defined as the extending of relevant
agricultural information to people (Swanson, Bentz & Sofranko, 1997). Moris (1991, p.
17) calls it "the promotion of agricultural technology to meet farmers' needs." This
process first became known as extension in England in the 1850s (Jones & Garforth,
1997). Extension has long played a role in the development of rural economies. This
"extending" of relevant agricultural information to people (Swanson et al., 1997) has
gone through many evolutions in various countries of the world.
Although the state has typically provided agricultural services, the paradigm for
research and extension has been changing extensively in Africa over the past 10 years.
State extension once played a successful and key role in development in Kenya. The
government service is now going through major changes and disintegration, however
(Kandie, 1997; Omolo, Sanders, McMillan & Georgis, 2001). Although extension has
been used in Africa under many different models, many claim today that it fails to do its
job of adequately reaching the majority of farmers and truly addressing the problems of
rural poverty, environmental sustainability, and food security (Eponou, 1996; Gautam,
2000; Republic of Kenya, 2001). Agricultural research and extension services have not
made the expected impact on small-scale farmers in Africa over the past few decades
(Eponou, 1996). Since the late 1980s, the efficiency of these services in Kenya has
dropped substantially (Government of Kenya, 1999; Kandie, 1997).
A wide range of factors has contributed to the current situation in Kenya. Its
weakening economy combined with poor management and corruption played an
important role. These internal problems coupled with the rapid expansion of the state
extension service created a large inefficient employee base that could only be sustained
with substantial outside support. At the same time, foreign donors (under pressure to
show "results") were presented with a situation in which there was almost no evidence
for successful technology diffusion through these bloated bureaucracies. One result was
a major shift of foreign-donor support to technology research and the search for
alternative non-state mechanisms for delivering the inputs and advice needed. The net
result of these macro-policy shifts, combined with the stagnation of the state extension
services, has been a rapid increase in the number of private sector and NGO actors that
work side by side with state extension services in Kenya. Once totally state-run,
extension in Kenya is now conspicuous by the heavy role and increasing diversity of non-
governmental actors.
The extension service in Kenya has been unable to effectively reach farmers, even
in the high potential areas (Venkatesan & Kampen, 1998). The failure of early extension
models, and more recently the Training and Visit (T&V) model instituted by the World
Bank, has left a shell of an extension structure in Kenya with only a limited ability to
reach farmers in an effective way (Gautam, 2000; Sanders, Shapiro & Ramaswamy,
1996). The structural adjustment programs put in place in developing countries by the
International Monetary Fund and the World Bank helped to contribute to this problem by
reducing investment by the public sector.
The World Bank concluded in their assessment of extension in 2000 that a
decentralized, demand-driven and pluralistic system was needed in Kenya (Gautam,
2000). Pluralism is being promoted worldwide in extension systems (Feder, Willet &
Zijp, 1999). "Extension is not necessarily a government program, but rather a complex
set of institutions whereby rural people obtain new knowledge and information" (Rivera
& Alex, 2004, p. 339). The Kenyan government and donors agree on the need to focus
more on clients and lessen government costs through outsourcing, using farmers' groups,
and cost sharing (Gautam, 2000). The current Kenyan extension model, called National
Agricultural and Livestock Extension Programme (NALEP), funded by the Swedish
government, is focused on pluralism.
Pluralism came about as a result of the inability of state services to provide for
farmers, and led to a search for other potential actors. The private sector emerged as one
important provider of services. Privatization has only recently become an issue in
extension. It entails the turning over of services typically provided by government to
private organizations. A study by Swanson et al. (cited in Umali & Schwartz, 1994)
showed that 81% of extension was provided by the public sector in the 207 organizations
surveyed in 113 countries. The private sector accounted for only 5% (Swanson et al.,
1990 in Umali & Schwartz, 1994). However, budget deficits are forcing both developed
and developing countries' governments to downsize, decentralize, and move toward a
liberalized economy. Private extension is seen as one way of cutting down on the
massive public sector that has for so long characterized countries such as Kenya.
Government extension has also been criticized for many weaknesses including
inefficiencies through bureaucracy and top-down approaches where they are out of touch
with the farmers. Public organizations are seen to be wasteful of resources because they
do not have the same profit motivation as private companies. Umali and Schwartz (1994,
p. xii) encapsulate this perspective:
In view of the changing conditions facing agriculture today, coupled with the
governmental and fiscal constraints faced by many developing countries, a
structural transformation of the agricultural extension system is becoming
increasingly essential. The public monopoly in agricultural extension provision in
many countries is no longer feasible or sustainable, and a shift towards a multi-
organization system consisting of the public, private, non-profit and non-
governmental sectors will be vital for the effective performance of this complex
task. Capitalizing on the comparative advantage of each of the different sectors
will ensure the success of this endeavor.
The private sector's profit motive for services is thought to make it more efficient.
However, this sector tends to ignore areas such as semiarid zones where there is little
chance of profit. The public sector is therefore still needed to advocate and intervene in
areas where the private sector has no interest. The key problem is that with the decrease
in government spending, it is unlikely that public-sector extension will have the means to
fully undertake the necessary support and services in the often-remote semiarid areas.
Therefore there is greater focus now on non-governmental organizations and community-
based organizations as important players in the extension scene.
Today in Kenya, many extension stakeholders and technology-dissemination
approaches exist, with few studies to show their effectiveness. There are still numerous
farmers who must be reached with effective technologies, however. One important need
in the new pluralistic milieu is to determine how community-based mechanisms such as
farmer-to-farmer extension works, and the role that community groups and farmers play
in extending technologies to other farmers. Knowing these mechanisms will contribute
to the effort in bringing about rural development.
With the food and environmental crisis throughout Africa, it is vital that all Kenyan
farmers receive the necessary information and inputs to make a living off their land
(Eponou, 1996). It is therefore crucial to explore all the avenues of rural development.
This includes examining the role that farmer groups play in disseminating technologies,
the mechanisms of farmer-to-farmer extension, what factors affect their success in
extension, who participates in the groups and why, and the implications of farmer-group
performance for extension policy.
If indeed farmer groups in the smallholder sector play an important role in
dissemination of appropriate technologies in Kenya, the question then becomes what
must extension and policy makers do to facilitate the smallholder sector-especially
farmer groups who are organized and already providing services-in scaling up and
therefore increasing production, addressing food security and fighting rural poverty?
This study shows the implications of farmer-group technology dissemination for
extension policy.
Study-Area Background
Meru Central District is an important smallholder agriculture district in Kenya's
Eastern Province, covering 2,982 square kilometers (Meru Central District Development
Plan, 2002) (Figure 1-1). It lies between 003'45" north and 0o2'30" south, and between
370 and 380 east. Administratively, within the district there are 10 divisions, 27
locations, and 75 sublocations.
Meru Central lies on the equator, and is bordered by Mount Kenya on the west and
drier lowlands to the north and east. It ranges in altitude from 300 to 5199 m at the peak
of Mt. Kenya. It has nearly all of the agroecological zones of Kenya (Teel, 1985; Were,
1988).
Farmers in Meru Central District practice mixed cropping methods with maize (Zea
mays) and common beans (Phaseolus vulgarii) as the dominant farming system. Other
food crops include bananas (Musa spp.), yams (Dioscorea spp.), potatoes (Solanum
tuberosum), sweet potatoes (Ipomea batatas), sorghum (Sorghum bicolor), finger millet
(Eleucine coracana), cassava (Manihot esculenta), arrowroot (Maranta anindinacea),
pigeon pea (Cajanus cajun), lablab beans (Dolichos lablab), cowpeas (Vigna sinensis),
groundnuts (Arachis hypogaea), kales ("sukuma wiki") (Brassica spp.), tomatoes
(Lycopersicum esculentum), onions (Allium spp.), cabbage (Brassica oleracea capitata),
pumpkins (Cucurbita spp.), sugar cane (Saccharum officinarum), avocados (Persica
americana), mangos (Mangifera indica), citrus (Citrus spp.) and papaya (Carica
papaya). Coffee (Coffea arabica), tea (Camellia sinensis), tobacco (Nicotiana tabacum),
cotton (Gossypium spp.), sunflower (Helianthus annuus), macadamia (Macadamia
tetraphylla) and pyrethrum (Chrysanthemum cinerariaefolium) are grown for cash.
Catha edulis (also called "miraa" or khatt") is a stimulant used mostly by Somalis,
Swahili people, and in the Arab Gulf. It has been an important cash crop in wetter tea-
growing areas, such as the Nyambeni Hills in the northeast of the greater Meru area. It is
LOCATION OF MERU CENTRAL DISTRICT
4
LEGEND
I Lake
iI District boundary
l National boundary
0 90 180 Kilometers
Figure 1-1. Meru Central District of Kenya
now being grown by many of the farmers in Meru Central District as well. Farmers in
the coffee zones have recently started growing it for cash with the decline of the coffee
industry.
A major feature of the farming landscape, especially in the middle and upper zones,
is the Australian tree Grevillea robusta. Apparently it was promoted for many years for
intercropping with coffee. It is very popular for all farmers, however, as a fast-growing
tree that produces good lumber. Other introduced species include Cassia and Leucaena
species, especially in the lower zones.
Livestock in the area include cattle, goats, sheep, pigs, rabbits, and chickens.
Farmers also keep bees. Livestock goods such as dairy products, meat, and hides are also
produced.
Rainfall is bimodal, falling between March and June (short rains) and October
through December (long rains). The southeastern slopes of Mount Kenya, where many
of the farms lie, receive between 1250 and 2500 mm of rainfall per year (Meru Central
District Development Plan, 2002). The leeward side of the mountain and northern and
eastern lowlands receive between 380 mm and 1000 mm annually.
Population within the district is 521,518. The growth rate is 1.48% (Meru Central
District Development Plan, 2002). Population density is an average of 167 people per
square mile. Farm size averages 1.1 hectares for smallholders. Although people are
moving to urban areas, absolute numbers of farmers in the rural areas are growing,
putting pressure on the natural resources of the district. Over 45% of the population is
classified as poor (Meru Central District Development Plan, 2002).
Some of the causes of poverty in Meru Central are seen as inadequate and
unreliable rainfall, unemployment, poor extension services, lack of land, collapse of the
cotton and coffee sectors, low prices for farm products and poor marketing channels
(Meru Central District Development Plan, 2002). The HIV/AIDS pandemic that is so
rampant in Africa is especially so in this area. Although it has recently dropped, the
Meru Central rate of infection was 38% a few years ago, well above Kenya's national
average of 15%. At the Meru Central District consultative forum for the government's
Poverty Reduction Strategy Paper, stakeholders listed inadequate extension services and
lack of extension services as two of the main problems in the district.
Dairy farming is an important economic activity in Meru, especially with the
decline of the coffee industry due to poor world market prices. Many farmers in the sub-
humid highlands of Kenya own dairy cattle or goats, and keep them in zero-grazing units.
This necessitates the growing of fodder or buying of feed for enhanced milk and meat
production. Zero grazing is a system whereby animals are kept in a stall or enclosure and
fodder is carried to them. Animals may also be managed through grazing and tethering.
However, commercial dairy meal is too expensive for many farmers to purchase, and is
perceived by farmers as unreliable in terms of quality (Daily Nation, Sunday, June 20,
2003).
Mineral fertilizers are also available, but too expensive for many farmers to afford
at the recommended rates. Because Kenya is in the process of liberalizing its markets,
there are now few subsidies on agricultural inputs. Farmers face low prices and poor
marketing channels in most of the rural areas, making farm-generated income difficult to
obtain. Farmers must find ways to increase production without the use of expensive
fertilizers and feeds.
The World Agroforestry Centre (WAC)
Especially for smallholders, agroforestry practices offer useful options and
alternatives to improve their farming systems. Agroforesty is the deliberate use of trees
on farms (in combination with crops, animals, or both) to meet multiple objectives of the
farmer. It is a "dynamic, ecologically based natural resources management system that,
through the integration of trees in farmland and rangeland, diversifies and sustains
production for increased social, economic and environmental benefits for land users at all
levels" (Huxley & van Houten, 1997). Many of the agricultural systems in the tropics are
(by nature) agroforestry, because of the integration of trees, crops, and animals.
However, scientists have recently begun paying attention to the benefits of agroforestry,
and much research has been conducted on the practice since the 1970s.
Smallholders have always used trees on their farms, as noted above. However,
since the introduction of structural adjustment programs in Kenya, intended by donors to
bring about economic recovery of developing countries' economies, inputs such as
fertilizers, chemicals, seed and feeds have been too expensive for many small-scale
farmers to afford. Attention has turned to agroforestry as a means of restoring soil
fertility and providing cash income. Agroforestry has also been shown to produce good
feed for livestock. Low quality and quantity of animal feeds are a further constraint to
production within the livestock sector (Winrock International, 1992).
As a result of current smallholder constraints, various organizations have been
conducting research in agroforestry, as a possible solution to some of the problems faced
by small-scale farmers. The World Agroforestry Centre (formerly known as the
International Centre for Research in Agroforesty, or ICRAF) was established as part of
the Consultive Group on International Agricultural Research (CGIAR) centers in the
1980s. While the headquarters are in Nairobi, WAC also conducts research around the
world in places such as West Africa, Asia, and Latin America. Other organizations
working in agroforestry in Kenya include the International Livestock Research Institute
(ILRI), the National Agroforesty Research Project (NAFRP) of the Kenya Agricultural
Research Institute (KARI), and the Kenya Forestry Research Institute (KEFRI). Much of
this research has been focused on central Kenya, where a high proportion of small-scale
farmers reside.
Because of the need for quality feed for dairy animals, WAC and partner
organizations have been conducting research since the 1990s on Calliandra calothyrsus
Meissner (calliandra) and other agroforestry species for improving milk production on
small farms. Studies have shown that calliandra is effective as a supplement or feed for
dairy cattle (Paterson, Kiruiro & Arimi, 1999; Roothaert & Paterson, 1997), and farmers
have begun to plant and use it on their farms. Over 2600 farmers in 150 groups in central
Kenya are growing calliandra for feed (Wambugu, Franzel, Tuwei & Karanja, 2001).
Food and Agricultural Research Management (FARM)-Africa
Another introduced technology in central Kenya for small-scale farmers is the use
of improved dairy-goat breeds. The NGO FARM-Africa has been working in the Meru
area, targeting the poorest farmers in medium- and low-agricultural potential zones, by
working with over 80 self-help dairy-goat groups. The Meru Dairy Goat and Animal
Healthcare Project has been in the Meru area since 1996. The purpose of the project is to
"improve the productivity of local goats through better management and access to
sustainable healthcare and genetic improvement, and of local dairy cattle through better
access to sustainable healthcare systems" (Meru Dairy Goat and Animal Health Care
Phase II April 1999-March 2002 Project Review, p. 8).
Activities of the project to achieve this goal include
* Community-based breeding of local goats with Toggenburg dairy goats;
* Formation and training of autonomous self-help groups to undertake breeding
activities;
* Development of community animal health care workers and a privatized veterinary
and drug supply service;
* Improvement of fodder supplies through community bulking and on-farm planting
of suitable fodder; and
* Development of effective extension support service through the existing Ministry
of Agriculture and Rural Development (MoARD) staff and extension system.
The Project works through both existing extension and the private sector to support
small-scale farmers in the district. Through these linkages, FARM-Africa helps the
dairy-goat groups obtain loans, training, and improved bucks for breeding. The project is
estimated to benefit the welfare and income of 20,000 families in the area.
Researchable Problem
There is much discussion among the government, NGOs, and international research
centers of the increasing role that farmer groups and other community-based extension
mechanisms are playing in the dissemination of technologies in Kenya today. Many are
advocating community-based extension as a means of scaling up (Nyakuni, 2001;
Raussen, Ebong, & Musiime, 2001; Wambugu et al., 2001). However, there is little
research showing what factors make community-based groups effective, if at all, in
disseminating technologies. The researchable problem is the need to examine farmer
groups and the role that they play, and what factors make them effective in extending
technologies. If the role of farmer groups in extension could be more clearly defined, and
evidence found for which factors could or do affect their effectiveness, it would facilitate
technology dissemination to small-scale farmers. This information will be useful for
organizations working with farmer groups, and to the groups themselves. It can provide a
means to strengthen and guide the groups. Finally, it will provide valuable information to
policy makers.
In view of this problem, the goal of this study then was to determine the role of
farmer groups in technology dissemination, and to assess what factors make groups
effective in extending technologies among small-scale dairy-goat farmers in Meru
Central District of Kenya. The FARM-Africa Meru Dairy Goat and Animal Healthcare
Project, working with dairy-goat farmer groups, provides a good case study in which to
examine this research problem.
Purpose and Objectives
The goal of this study was to examine the role of farmer groups in technology
dissemination, and to assess what factors make groups effective in extending
technologies among small-scale dairy-goat farmers in Meru Central District of Kenya.
Specific objectives were as follows:
* Examine participation in groups and identify what factors, if any, affect
participation in groups;
* Examine linkages and their outcomes, if any, between farmer groups and other
extension stakeholders;
* Identify the mechanisms by which farmer groups and their members receive and
disseminate information and new technologies, especially fodder shrubs and
improved dairy-goat breeds;
* Identify the factors characteristic of groups successful in disseminating technology;
and
* Propose policy recommendations to extension and development organizations
regarding farmer groups' roles in extension.
Research Questions
* Who participates in the groups, and what factors affect participation?
* What linkages, if any, exist among farmer groups and other extension players?
* What are the mechanisms, if any, within and outside of the groups for giving and/or
receiving information and technology?
* What factors affect the success of farmer groups in disseminating technology?
* What are the implications of farmer-group dissemination for extension policy?
Assumptions
Several assumptions were made in this study. One was that rural development
(through means such as extension) was something that was desired by all of the
stakeholders involved. Another assumption was that factors that affect the success of
farmer groups in disseminating technology were measurable and valid. One way to assist
with this was to state the operational definition of terms of the study (see below).
Furthermore, it was assumed that respondents were forthright in their responses.
Operational Definition of Terms
Adoption. Use of a new technology or technique by farmers, in any amount, and
for any length of time'.
Agroforestry. Deliberate use of trees on farms (in combination with crops,
animals or both) to meet multiple objectives of the farmer.
Dissemination. The spread of information and technologies through various
means of communication.
' World Agroforestry Centre defines adoption of fodder shrubs as having expanded once and having over
100 trees, for dairy cattle. For this study, farmers, extension agents, and FARM-Africa personnel referred
to "adoption" as the use of a technology without specifying quantity or time. Thus, the definition used in
the study is operationalized for the study purposes and in line with the perspectives of study participants.
Extension. The process of sharing information and technologies among various
development stakeholders.
Group. A local organization of people who have banded together to take
advantage of social capital.
Fodder tree. A tree or shrub that is used to feed dairy animals. The fodder is
often cut from the tree and carried to livestock.
Leadership. Role played by various people (within the group context) to provide
guidance, motivation, and management.
Linkage. An entity (organization, person, group) that has some sort of connection
or relationship with dairy-goat groups for any purpose.
Location. An administrative level below the district level, and above the village
level. Used also as an adjective locationall).
Scaling up. Strategies that lead to an enlargement of program size.
Small-scale farmer. Rural person who makes a livelihood from less than two
hectares of land.
Social capital. Norms and networks that enable collective action (the management
of resources by groups).
Success. The determination by a group, individual or organization that a group has
effectively disseminated information and/or technology. For this study, success was
determined through variables such as self-ratings of the groups themselves and outside
entities on perceived success in dissemination, number of neighbors adopting
technologies, number of farmers and groups trained, and number of buck services.
Technology. An idea, practice, or object used and/or promoted to improve
agricultural production (adapted from Rogers, 1995).
Limitations of the Study
Several constraints may have limited the study. Factors such as geography, tribal
identity, history, and gender may have affected both group performance and the role that
groups play in disseminating technology.
Language, culture, and gender may have also biased or complicated the findings.
This may have occurred not only with the North American researcher, but also with
assistants and extension observers who were possibly different from the farmers through
gender, culture, economic situation or education. Personal bias may have affected
interviews. Finally, many of the questions on the questionnaires dealt with perceptions of
farmers, and so have the possibility of bias on behalf of the respondents. The researcher
attempted to avoid these issues through
* Awareness of and attentiveness to potential bias;
* Use of trained assistants to help with the interviews and to provide input and
interpretations;
* Triangulation through research design and data sources, such as interviewing
various extension players, including individual farmers, groups, and government
and non-government extension personnel;
* Establishing of a record trail of data obtained;
* Use of local languages; and
* Attention to both who is being interviewed and who is not.
Significance of the Study
Agriculture is the backbone of many African economies, yet many development
obstacles prevent improved agricultural production from increasing the standards of
living and decreasing poverty in rural areas. Key to sustainable livelihoods is the
opportunity for farmers to obtain and share useful information and technologies for their
farming systems. For decades now, extension has been attempting to increase production
through the provision of such information and technology. Only recently, however, has
the focus been on the needs of the farmers themselves, and their empowerment through
participatory methods of technology development and dissemination. If "farmers are the
owners of development" (Barkland, 2001), then only by facilitating their methods and
priorities can development organizations truly make a dent in the obstacles to their goals.
This study is important because it recognizes the crucial role that smallholders play
in the technology development process, and attempts to portray the role that farmer
groups play in disseminating technologies, what factors affect their success in doing so,
who participates in the groups, reasons for participation and the mechanisms of farmer-
led extension. Providing evidence on what role groups play and what factors affect their
performance can help to strengthen groups and to guide outside organizations in
facilitating and collaborating with groups (Place et al., 2002).
This study will be of significance to the many organizations working with small-
scale farmers in Kenya. It can also assist groups in reaching their goals, and in becoming
more effective. Finally, the study will have implications for extension systems in similar
regions.
Organization of Thesis
This chapter has given a brief introduction to the particular problems faced by
smallholder farmers in Kenya, and to projects working in Meru with these farmers.
Chapter 2 is a literature review of theories and studies related to agricultural extension,
social capital and farmer groups. The third chapter presents the research design, methods
22
used and procedures followed to collect data. Chapter 4 presents the results of the study
with regard to the first four objectives, and Chapter 5 contains the conclusions and
recommendations of the study. Instruments used are included in the appendices.
CHAPTER 2
LITERATURE REVIEW
Coming together is a beginning
Keeping together is progress
Working together is success
-Henry Ford
It was shown in Chapter 1 that agricultural extension is an important component of
rural development. A brief description of extension was given, along with an overview
of extension's role in Kenya and its status today. Extension has evolved over the years to
meet the changing needs of its clients, to become more effective, and in response to
economic and environmental realities present in various countries. A description of this
evolution in Kenya is presented in this chapter, detailing the move from the top-down,
transfer of technology model to the so-called "farmer first" methods. The chapter goes
on to discuss social capital and group theory, which are important factors in extension in
Kenya today.
Extension History and Models in Kenya
Introduction
To meet the challenges of development, bring about rural improvement and address
farm constraints such as those faced by farmers in Kenya, agricultural research and
extension have been used for decades to advance the rural sector in nearly all countries.
In the previous chapter, the World Bank defined extension as "the process of helping
farmers to become aware of and adopt improved technology from any source to enhance
their production efficiency, income and welfare" (Purcell & Anderson, 1997, p. 55). It is
important how extension is defined because that in turn affects how it is conducted.
Extension has traditionally been defined as the delivery of information and
technologies to farmers (Moris, 1991). This leads to the technology transfer model of
extension, seen by many as the main purpose of agricultural extension (Moris, 1991).
This is based on the idea that "modem" knowledge and information is transferred through
extension agents to recipient farmers.
The conventional provider of extension, the state, has typically used top-down,
"transfer of technology" (TOT) methods for extending new technologies. Top down
methods characterized the United States extension model, which was instituted by many
colonial governments in Africa. In the TOT approach, technologies are generated at
research stations and diffused to farmers using the extension service (Put, 1998). Not
only technologies but also intangibles such as power, prestige and skills are located at
these centralized stations (Put, 1998). Technologies are spread vertically in this top-
down approach. The TOT approach is often biased toward better-endowed farmers
whose fields and infrastructure are more like those of the research stations (Chambers &
Ghildyal, 1984).
Early extension models in Kenya therefore followed a "cookbook" approach to new
technology through state extension services (McMillan, Hussein & Sanders, 2001, p. 1).
Technologies were developed at the Ministry of Agriculture and run through the
extension pipeline via extension agents to farmers, with agricultural development being
the desired product. Farmers were not much involved in the development or
dissemination of technology. Research and extension were focused mainly on large-scale
farms or those smallholders living in high and medium-potential areas, and trials and
demonstrations were mostly on research stations. This approach began during the
colonial era and continued into the 1980s.
Diffusion of Innovations Theory and Transfer of Technology
Transfer of technology approaches are strongly linked to the diffusion of
innovations philosophy. Diffusion of innovations theory says that technologies are
communicated over time among the members of a social system, and adopted according
to various characteristics of both the technology and the user (Rogers, 1995). The
diffusion of innovations model was focused on a very linear process of technology
development. Rogers' model has been critiqued for this and for other shortcomings, such
as the pro-innovation bias, blame of farmers for "non-adoption" of technologies, lack of
recognition of farmer innovations, and focus on the change agency/change agent instead
of the ultimate end users of technology (the farmers).
More recent thinking has developed models that are more iterative, dynamic, and
cyclical in nature (Figure 2-1). Rogers himself moves away from linear technology
transfer with the convergent model in the latest version of his theory on the diffusion of
innovations (Rogers, 1995).
The theory of innovations and related transfer-of-technology model has tended to
work better in developed rather than developing nations, but even within developed
nations, the perceived process has evolved into the more iterative model. The linear
model originally proposed by Rogers works better when there are limited
recommendation domains for the technology. Technologies can then be recommended in
"blanket" form.
Figure 2-1. Farming systems emerging methodology on technology transfer
Note: Adapted from Bastidas, 2001
Researchers in developing nations first recognized the need to apply new thinking
to the "problem" of slow or non-adoption (Dunn, Humphreys, Muirhead, Plunkett,
Croker et al., 1996). Small-scale farmers living in risk-prone, complex environments are
often unable to take advantage of many of the technologies developed on research
stations for large-scale farms. Researchers working around the world noticed the unique
problems of the small-scale farmer livelihood system, and developed strategies to solve
these that are now known as the farming systems approach (Collinson, 2000; Escobar,
2000; Harwood, 2000; Hildebrand, 2000; Norman, 2000).
In the iterative model, much more focus is on the endogenous nature of
innovations. Starting in 1982, development practitioners began emphasizing the notion
that research activities should begin and end with farmers. Rhoades and Booth (1982)
coined the term "farmer-back-to-farmer." Chambers developed this into the "farmer-
first" philosophy (Chambers, 1990, as cited in Dunn et al., 1996). Along with these were
the "putting people first" (Cernea, 1985) and "farmer participatory research" models
(Farrington & Martin, 1988, as cited in Dunn et al., 1996). The linear model does not
show the many innovations that come from sources other than formal research. Roland
Bunch (1985) and many others (described in Haverkort, Van de Kamp & Waters-Bayer,
1991) have shown that farmers are experimenters.
Current dissemination thinking takes a much more participatory, farmer-centered
approach than the diffusion of innovations theory. Farmers are involved in every aspect
of technology generation, from generation to testing to dissemination. However, it has
not always been this way. Much of the history of extension in Kenya is beset with
examples of top-down, transfer-of-technology models of technology dissemination, many
following the theory of diffusion of innovations.
The 1960s Top-Down Approach: State Transfer of Technology Model
During the colonial period the state extension service developed into a major
service provider for large-scale colonial farmers. After independence in Kenya, the state
continued be the major actor in agricultural extension for the first twenty years (Schwartz
& Kampen, 1992) (see Table. 2-1). The colonial extension had used a regulatory and
commodity approach (Kandie, 1997). Following independence, the government
instituted a more general approach, based on the U.S. extension system and with funding
from the U.S. Agency for International Development (USAID). This approach assumes
that agricultural ministries have useful information for farmers, and extension's job is to
transfer this to farmers (Schwartz & Kampen, 1992). USAID promoted new technologies
through demonstrations, technical packages and information.
At the same time, the commodity extension approach was still used for both small
and large-scale farmers. The commodity approach has been one of the most enduring
approaches used for extension in Kenya. It was successful in disseminating hybrid maize
technology, which was developed in Kenya around 1955.
In the late 1960s, the Ministry of Agriculture adopted a farm management approach
to extension (Gautam, 2000). This was initiated as part of a credit program for farmers.
The Ministry started the Farm Management Division at the same time to run the credit
program.
Table 2-1. Early extension approaches strengths and weaknesses
Type of
extension When adopted Strengths Weaknesses
Regulatory, Colonial period, Good management, Top-down, ignores
commodity 1945-1963 effective for smallholders,
resource-rich coercive
General Early Focus on whole family, Top-down, poor
independence, increased management and
1963 participation linkages
Farm Late 1960s Provides inputs, Unsustainable, focused
management management skills on credit
Integrated 1976 Provides inputs, holistic Lack of training and
linkages; top-down
The 1970s Holistic Approach: Integrated Rural Development Projects and Farming
Systems Philosophy
During the 1970s, donors began to place an increasing emphasis on the poverty of
rural people, and the Integrated Rural Development Project (IRDP) was started in Kenya
in 1976 with World Bank support, using an integrated extension approach. The IRDP's
goal was to build institutional infrastructure and to provide inputs to farmers to increase
production (Moris, 1991). These inputs included extension, research, irrigation, credit,
roads, water, electricity, and sometimes schools and health centers (Venkatesan &
Kampen, 1998). The focus was mainly on technical aspects, however, and left out crucial
issues such as training, linkages with research, and management. Integrated Rural
Development Projects are seen to have failed mainly due to lack of sustainability,
administrative problems, a top-down approach, and failure to build local capacity
(Venkatesan & Kampen, 1998).
There was concern by donors during this period that developing countries were at
risk of famine due to shortages of major staples. Many thought that agricultural research
would help address this issue (Hansen & McMillan, 1986). The same time period
coincided with the first activities of the Rockefeller and Ford-funded international centers
in Africa, and the World Bank consultive groups on agriculture in individual countries.
The international community decided to create an organization for international
agricultural research, and the Consultative Group on International Agricultural Research
(CGIAR) was launched in 1971 (CGIAR web site, available at
http://www.cgiar.org/who/wwahistory.html). It established International Agricultural
Research Centers (IARCs) around the globe, many of which provided research on food
crops.
One of these CGIAR centers was the International Maize and Wheat Improvement
Center, CIMMYT, which established a branch in Kenya. This was one of the
organizations that, during the 1960s and 1970s, gave greater emphasis on smallholder
farmers and their livelihood systems, as researchers realized that such people were not
being reached effectively with the traditional extension approaches. They thus began to
use what is known as the farming systems approach to research and extension. In Africa,
this was initiated through the work of Michael Collinson with CIMMYT (Collinson,
2000). Farming systems is a holistic type of approach that looks at the entire farm as a
system with various subsystems. It provides for greater dialog with and input from
farmers, and for enhanced linkages between research, extension and farmers. This model
was marked by participation at the farm level (through farmer input on research and on-
farm trials) and by interdisciplinary linkages and a systems approach to extension.
The farming systems approach (Norman, 2002) was characterized by
* A holistic approach viewing the farm as a whole;
* Involvement of farmers and their priorities;
* Research reflecting the various subsystems' interactions and linkages and
* Reliance on informal surveys or "rapid rural appraisal" (RRA).
The 1980s Training and Visit Approach: Expanding the State Extension Service
During the 1980s, state extension systems in developing countries were further
altered, due to recognition of the need to reach more farmers and to better train extension
agents. The Training and Visit (T&V) system was a model that was instituted by the
World Bank in Kenya in the early 1980s (Gautam, 2000). The National Extension
Project (NEP I) of Kenya was the World Bank-funded T&V extension approach.
The T&V system was designed to address some of the weaknesses in the previous
extension approaches, such as weak linkages with research and low training of field
extension workers. It was introduced as a pilot project in two districts in Western Kenya
in 1982, and expanded to cover 30 districts by 1983. The objective of NEP I was to have
sustained increases in agricultural production in 30 of Kenya's 41 districts, all medium-
and high-potential areas. National Extension Project I was also to promote institutional
development within the extension service (Gautam, 2000).
As its name suggests, the basic premise of T&V was training (instilling
professionalism in extension agents) and regular visitation of farmers by the agents.
Extension agents were to be trained in new technologies every two weeks. Under the
T&V system, subject matter specialists advised extension agents, and also provided a link
between extension and research. They provided monthly workshops where field
extension workers are trained. The agents then took these techniques to the farmers for
two weeks before returning for further training. Farm families were divided into groups,
with five to ten contact farmers per group. These contact farmers were to provide a
multiplier effect.
The second National Extension Project, NEP II, a continuation of NEP I, was
started in 1991. The second project's objective was to increase smallholders' incomes. It
sought to reach lower-potential areas and more marginalized farmers, and to further
improve links with research. NEP II expanded extension coverage to 40 of the now 45
districts, including some of the previously ignored semi-arid zones.
There are conflicting reports on the effectiveness of T&V in Kenya, which give
insight into the difficulties of measuring the impacts of extension. Bindlish and Evenson
(1997) performed an econometric review of T&V extension in Kenya and Burkina Faso.
Their study claimed high returns to extension through T&V. On the other hand, Gautam
(2000) reported limited impact of T&V extension in Kenya in a review of NEP
commissioned by the World Bank Operations Evaluation Department (OED). The OED
concluded that NEP I had some beneficial aspects but several operational deficiencies,
and was not financially sustainable (Gautam, 2000). Following discussions with the
Africa Region of the World Bank, the final rating of the NEP outcome was "marginally
satisfactory" (Gautam, 2000).
Whatever the arguments, there are indications that T&V had many shortfalls.
Some feel that T&V focused so much on training that the system lost sight of the goals of
meeting farmers' needs and improving their livelihood. It was essentially a supply-
driven and top-down system, promoting agricultural messages that had been designed and
developed by research scientists, with limited input by the ultimate users of the
technologies (the farmers). The delivery method was perhaps efficient, but the messages
often irrelevant, according to farmers surveyed (Gautam, 2000). At the end of NEP II in
1998, the extension service was characterized by weak management, a lack of strategy
for the service, and general ineffectiveness (Gautam, 2000).
T&V, like the general extension approach, was characterized by limited feedback
from farmers. The packages were somewhat mechanistic, and not flexible enough to
meet the needs of Kenya's variety of farming systems. Training and Visit relied on
contact farmers, and tended to neglect the larger rural population (Moris, 1991). In NEP
I and NEP II, there were no real mechanisms for choosing contact farmers who truly
represented many of the farming systems in the areas. The hierarchical structure set up
by the Bank prevented innovation, partnering, and efficiency. Despite a supposedly
improved system, farmers before and after NEP said they were not receiving advice from
extension, or else not the advice that they needed (Gautam, 2000).
An important factor in agricultural technology development and dissemination
during the 1980s and 1990s was structural adjustment programs (SAPs). The World
Bank and the International Monetary Fund introduced SAPs to help address some of the
economic crises that were facing Kenya and other developing nations. The 1980s had
brought about economic hard times to many developing countries, when the high price of
oil coupled with drought led to growing foreign debt. Also contributing were the
excessive growth of parastatals in African nations and declining prices of primary
products (Sanders, Shapiro & Ramaswamy, 1996).
Constraints to development in African countries were seen to be in four main areas:
budget deficits, over-centralized governments, recurrent personnel costs in bloated
bureaucracies, and declining administrative capacity (Cohen, 1993). The highly
centralized government led to much inefficiency and corruption. Over half of Kenya's
employed population worked in the public sector. Government expenditures escalated
mainly due to the large growth in the public sector and debt service. One of the main
ways to cut down on expenditures was therefore to reduce government employment.
Donors instituted structural adjustment programs in Kenya to address budget
deficits and kick start the ailing economy. These programs included reduction of the civil
service, liberalization of markets and pricing policies, reforms of parastatal organizations
and removal of foreign exchange controls (Cohen, 1993; Ikiara, Jama & Amadi, 1992;
Sanders et al., 1996). The bloated civil service was to be reduced while privatization was
encouraged. The hope was that these steps would bring about more economic growth by
liberalizing marketing and pricing policies and reducing state control. The free market
was seen by pundits to be the best way to efficiently allocate resources (Ikiara, et al.,
1992).
Structural adjustment programs played a large role in technology dissemination and
growth during that period. Although SAPs were expected to adversely affect the urban
population, who would now have to pay more for food, the smallholders among the rural
areas were affected negatively as well. In the long run, SAPs were meant to increase
incentives to farmers, expand private investment, improve economic efficiency, improve
trade balances and develop appropriate energy sources (Sanders et al., 1996). They
aimed to encourage the private sector by reducing the size of the civil service.
However, although product prices were increased, privatization of the input
suppliers meant that smallholders no longer obtained subsidies on fertilizers and other
inputs. With the cost of inputs dramatically increased, SAPs reduced the smallholder's
ability to purchase inputs such as chemical fertilizers when the government stopped
subsidizing them. Growth in the private sector has not made up for the SAP-induced
reductions in the public sector (Sanders et al., 1996).
Extension funding was reduced more and more during the 1980s and into the 1990s
due to the continuing economic crisis and the structural adjustment programs which
encouraged state downsizing. As the Kenyan government failed to recover from the
budget deficits so prevalent in the 1980s, it became obvious that they could no longer
fund T&V and other expensive extension models. There were limited funds for
operational costs, and most of the funding (80%) tended to go toward personnel. It was
therefore difficult to reach their clientele since there were no funds to fix vehicles or pay
for fuel to get to the field.
The 1990s Pluralistic Approach: Community-Based Farmer-Led Extension
Although state extension once played a major role in Kenya, the paradigm for
research and extension has been changing extensively over the past decade. The state
extension service has recently gone through major changes and disintegration (Kandie,
1997; Omolo, Sanders, McMillan & Georgis, 2001). Once totally state-run, extension in
Kenya is now conspicuous by the heavy role and increasing diversity of non-
governmental actors. The outcome of the economic crises and structural adjustment
programs in the 1980s was a search for other potential extension actors within the
extension domain during the 1990s. Many stakeholders began calling for a pluralistic
(multi-provider) extension model, in which the state takes on the role of facilitator for the
many other actors involved in extension such as non-governmental organizations,
farmers' groups and private extension (Gautam, 2000; McMillan, Hussain & Sanders,
2001; van den Ban, 2000).
Because farmers are already receiving information and technology from a range of
sources from other farmers to private agro-business to the public government extension
system, Zijp (2002) calls for the promotion of pluralistic extension approaches. The
World Bank is now also promoting pluralism in extension in Kenya (Gautam, 2000).
This pluralistic type of system is meant to contribute to flexibility and complementarity
of extension systems, and meet the diverse needs of a wide range of farmers (Crowder,
1996).
Farmer-Led Extension
Not only in Kenya, but in many countries today, extension is marked by
partnerships between various agencies such as the state, private companies, non-
governmental organizations, and farmers' groups. These partnerships and linkages are
seen as necessary to both cut costs and to involve all of the stakeholders in the extension
process. In such "coalition systems" of extension, the various stakeholders acknowledge
the role and skills of other partners, and are strengthened by alternate perspectives and
expertise (Anderson & Crowder, 2000). Multi-partner interaction promotes mutual
learning and innovation. Coalition systems can be flexible and provide a system of
checks and balances (Anderson & Crowder, 2000).
Extension systems today are characterized by approaches placing greater emphasis
on farmers playing a central role in the technology development process. The new
approaches collectively might be called farmer-led extension. The farmer-led model for
extension in developing countries is designed to be participatory, demand-driven, and
client-centered. This approach essentially evolved from the farming systems and similar
approaches, with a greater emphasis on the needs of resource-poor farmers, gender, and
the value of indigenous knowledge systems. Many researchers have described such
farmer-centered extension systems, including Chambers (1997), Esman & Uphoff (1984)
and Payson, Ganpat, Hartmann, Peters and Place (2004). Scarborough, Killough,
Johnson and Farrington (1997) describe farmer-led extension as
a multidirectional communication process between and among extension staff and
farmers involving the sharing, sourcing and development of knowledge and skills
in order to meet farming needs and develop innovative capacity among all actors, in
which all farmers have a controlling interest; are "centre stage", are the
protagonists and play a key role in technology development and delivery; and
involving farmers in training other farmers and trainers, and in sharing, sourcing
and transferring knowledge and skills. (p. 4)
The philosophy in farmer-led extension is very different in its view of farmers and
scientists, when compared with the TOT model (Table 2-2). It is a "bottom-up"
approach. It sees farmers as part of the entire process of technology generation,
providing essential input and assisting in the design and evaluation of new technologies.
These models both begin and end with the farming family and the livelihood system. The
farm, not the research center, is the central location to the model. Scientists work closely
with farmers in this type of extension. It is believed that a farmer-led approach will
generate more appropriate technologies to farmers in low-resource areas.
Farmer-led extension models began showing up among non-governmental
development agencies that sprang into action at the reduction of many government
services in the 1980s and 1990s. Non-governmental organizations and community-based
organizations (farmers' groups) are playing a key role in extension around the world
today. "Farmer-led approaches to extension are spreading ... and farmers' associations .
.. are contributing handsomely to the diffusion of modem technology" (Picciotto &
Anderson, 1997, p. 6).
Table 2-2. Philosophy of transfer of technology (TOT) and farmer first
TOT Farmer First
Diffusion of technology Top down Bottom up
Farmer's role Beneficiary Client; colleague
Scientist's role Technology generator Consultant; collaborator
Extensionist's role Deliver technology and Facilitate and network
demonstrate
Determination of research Perceptions of scientists Perceptions and needs of
priorities farmers
Main research location Research station Farmers' fields
Explanation of non- Failure of farmer to learn, Failure of technology and
adoption farmer's constraints of scientists
Note: Adapted from Chambers & Ghildyal, 1984; Scoones, 1996; Scoones & Cousins,
1996
Extension is being called upon to engage communities more and to work with
them. Extension then is given a wider mandate than simply transfer of technology, one
that includes farmer mobilization, education, and organization (Zijp, 2002). Involving
the community in extension is one way for such engagement to occur.
By working through community groups, development agencies are more free to
focus on training and provision of materials, while community groups can be more
involved with planning, mobilization, and facilitation (Raussen, Ebong & Musiime,
2001). Such empowerment leads to a sense of ownership and contributes to more
effective programs. Community groups give development agencies entry points for
understanding village problems (Noordin, Niang, Jama & Nyasimi, 2001). Such groups
allow development agencies to build on existing social capital, therefore accelerating and
enhancing impact. It is a way to reach many in the community, including the poor
(Noordin et al., 2001).
Community-based extension mechanisms have both advantages and disadvantages.
One obvious advantage of state extension, of course, is a source of funding for such
outreach. Private extension players also have the funds for outreach. What options, if
any, do farmer-led extension mechanisms have? One option is contracting out, where the
public sector uses funds to contract NGOs, community-based or private extension
providers (Anderson & Crowder, 2000). Also, farmers and their groups are willing to
help shoulder the cost of extension, according to some studies. Gautam (2000) found in
his survey of T&V extension in Kenya that farmers were willing to pay a certain amount
for extension advice. In Nigeria, Apantaku, Fakoya and Sodiya (2002) found that many
stakeholders, including farmer groups, were willing to help fund extension. This
included community-based associations, religious organizations, non-governmental
organizations and farmer groups. Research conducted by Bebbington, Merrill-Sands and
Farrington (1994) and Ashby and Sperling (1994) corroborates this view.
Another concern with farmer-led extension is sources of information and new
technologies. The sources of new information and technologies remain diverse in
farmer-led extension. Previous TOT extension methods focused on the research station
and scientists as the key source of knowledge. Farmer-led extension is better adapted to
the newer thinking that embraces all sources of knowledge including farmers themselves.
The international research centers in Kenya today are focusing on participatory research
where farmers are heavily involved in research, almost all of which is on farmers' fields.
Community-based extension does not stand on its own, but is yet one more
opportunity in the basket of options for development practitioners and policy makers to
use in bringing about rural development. Utilizing farmer groups or any other alternative
vehicle for extension of technologies does not negate the need for the state to remain
involved. Government extension is still needed (Rivera & Alex, 2004). However, with
the new pluralistic approach, other actors such as community-based mechanisms are
playing a larger role.
In conclusion, based upon both the farmer first approaches and the current
government budget deficits, community-based extension mechanisms are one of the most
promising means of scaling up, or extending, technologies. In order for effective
dissemination to take place, development players now view approaches using extension
by farmers and farmer groups as most appropriate to effectively and sustainably transfer
technologies to smallholders (Cooper & Denning, 1999). This type of farmer-led
extension looks to the communities to mobilize themselves to determine their own
problems and priorities in development, and to form groups to address their community
concerns. Because of the millions of small-scale farmers in the area, every farmer cannot
be reached by formal extension services. Therefore one of the major ways to bring about
development is to build capacity and empower communities to extend technologies.
Social Capital
To better understand community-based extension mechanisms such as farmer-led
approaches, it is necessary to examine social capital. Social capital is a construct that has
been viewed recently by many development players as an overlooked yet important factor
in rural growth (Grootaert, 2001; Pretty & Ward, 2001). Itjoins other forms of capital
that also play a role in development-natural, physical and human capital. Some claim
that social capital has been a missing link in development, and that by paying attention to
it development can better be achieved (Pretty & Ward, 2001; Robinson, Siles & Schmid,
2002).
There are many definitions of and theories about social capital. Grootaert (2001)
defines social capital as the internal social and cultural coherence of society, or the norms
and values that govern interactions among people and institutions. Robinson et al. (2002,
p. 5) define it as "a person or group's sympathetic feelings for another person or group.
Sympathetic feelings may include admiration, caring, concern, empathy, regard, respect,
sense of obligation or trust for another person or group."
The World Bank has a program called the Social Capital Initiative. It defines social
capital as "the norms and networks that enable collective action" (the management of
resources by groups). According to the World Bank, "increasing evidence shows that
social cohesion social capital is critical for poverty alleviation and sustainable
human and economic development" (World Bank, 2003).
Both narrow and broad conceptions of social capital exist. The most narrow, put
forth by Putnam (1993, in World Bank, 2003) sees it as the horizontal networks among
people. His view of social capital is at the micro level. Coleman, on the other hand
conceived of social capital as having both horizontal and vertical associations (1990, in
World Bank, 2003). Vertical links are hierarchical and have power differentials. Social
capital, in Coleman's view, is more of a meso-level view. The even more encompassing
macro level involves political aspects and formal institutions such as the court system in
its view of social capital.
According to social capital theorists, there are two basic types of social capital:
structural and cognitive. The structural type involves information sharing, and collective
action through established roles and social networks (Uphoff & Wijayaratna, 2000).
Structural capital is easier to observe and quantify. Cognitive social capital, on the other
hand, is more related to shared norms, values, trust, attitudes, and beliefs (Uphoff &
Wijayaratna, 2000). Cognitive social capital is more of a subjective concept and is harder
to measure and quantify.
Although some argue that social capital has changed in Africa as a result of
capitalism and structural adjustment, collective action is still important among farmers.
Hoon2 (2002) describes the transformation of labor arrangements in the Eastern Zambia
from more egalitarian "horizontal" relationships such as collective labor arrangements to
more vertical relationships determined by the labor market or patron-client relations.
Traditional communal labor arrangements (kalimalima) have declined, while more
individualized labor agreements (ganyu) have increased. This is a result of the
increasingly monetized and individualistic relationships brought about through both the
growth of capitalism and structural adjustment programs. In response, collective action
among farmers has changed and adapted components of both the more traditional labor
practices and more modern capitalist practices. Three new types of collective labor
arrangement since the 1990s are known as "group ganyu," "Kalimalima-in-clubs," and
2 Hoon, P. (2002). Balancing labor market demands with solidarity networks: Changes in labor
mobilization in eastern Zambia. Paper presented for the Department of Political Science and Center for
African Studies, University of Florida, Gainesville
"neo-Kalimalima." These types of institutions are known as syncretic, and "blend rules,
habits, or norms of an earlier time with modem institutions" (Hoon, 2002, p. 22).
Therefore social capital may have changed, but still exists strongly in African rural life.
There are many studies on social capital and its effects on people in rural areas.
Does it really contribute to rural development? How can development agencies capitalize
on social capital? Following are the findings of various studies on social capital.
One important finding on social capital was Krishna's 2001 study in India. He
found that having a high level of social capital was not necessarily correlated with high
development performance. However, once a capable agency was added to high social
capital, high performance was achieved. This points to the value of linkages between
local social capital such as community groups with outside agencies such as extension,
NGOs and local research organizations.
Other researchers found a positive relationship between income and social capital.
Narayan and Pritchett (1999) found that social capital was positively correlated with
increased incomes. Because they were studying farmers who did not necessarily make a
measurable income, they used expenditures as a proxy for income, and found that
increased social capital led to increased expenditures. In fact, a one-standard deviation
increase in social capital led to increases in incomes of all households by about 50%
(Narayan & Pritchett, 1999). Furthermore, they found that adoption of improved
practices was positively correlated with social capital. Narayan and Pritchett argue that
social capital is just as important to households as many other factors such as schooling,
distance to market and gender of household head. They believe that social capital is an
important missing dimension of poverty analysis.
The World Bank Local Level Institutions (LLI) study on social capital was
conducted in Bolivia, Burkina Faso and Indonesia (Grootaert, 2001). In line with
Narayan and Pritchett's 2001 study, Grootaert found that social capital contributes to
welfare. Because poor people do not necessarily have an income, both these studies
examined expenditure as a proxy for income. They also examined asset accumulation.
Increases in social capital led to increases in expenditure and other welfare indicators.
Gugerty and Kremer (2002) examined the question of whether development
assistance helped to build social capital. They found that outside funding had no effect
on social capital formation in a study of three development projects in western Kenya.
Their paper suggests that social capital is not easily created, and that many organizations'
goals of increasing social capital through funding may be limited. The paper does not
discuss the historically strong role of social capital in rural Kenya.
There are also studies on the role of heterogeneity in social capital. Groups may be
heterogeneous on many variables, such as age, gender, tribe and wealth level. Varughese
and Ostrom (2001) found that heterogeneity was not an important predictor of outcomes
in collective action such as farmers' groups.
Farmer Groups
One of the most promising means of scaling up technologies in the new pluralistic
extension environment is through social capital in the form of community-based
extension mechanisms. Social capital in the form of groups is used in communities
worldwide, especially in rural areas, as safety nets to cope with risks and for mutual
assistance. Groups are valuable as a form of collective action to farmers, providing
resources such as credit, labor and information. Groups allow farmers to obtain new
technologies, benefit from economies of scale, enter into stable relationships with
suppliers, and set rules for natural resource management (Place et al., 2002; Stringfellow,
Coulter, Lucey, McKone & Hussain, 1997). Donors are seeing the value of farmer
groups, such that they are sometimes a prerequisite for various agricultural projects
(Stringfellow et al., 1997).
Role of Farmer Groups in Extension
Farmer groups have played an important role both in the community and in
extension, and now appear to be taking on an even larger role. It is known that farmers
transfer knowledge and technologies to each other (Arbab & Prager; 1991, Gubbels,
1991; Maseko, Scoones & Wilson, 1991). Maize was spread throughout the African
continent long before any formal extension was in place. Rhoades and Booth (1982)
argue that farmers are beneficial sources of information and practices for other farmers.
In Kenya, the major source of agroforestry germplasm was other farmers, according to a
study in 1998 (Edouard, 1998). Farmers obtained germplasm from their own farms,
relatives and neighbors. Over 39% of the farmers interviewed exchanged agroforestry
germplasm with other farmers (Edouard, 1998).
Groups are considered by both the Kenyan government and donors to be vehicles
and entry points for new technologies and training for farmers. Extension workers in
Meru Central District find that their work is easier to handle when they deal with groups.
Groups can be a powerful tool for extension, especially because they present an efficient
way for extension staff to pass on information and technologies. The current Kenyan
extension program, National Agriculture and Livestock Extension Programme (NALEP),
encourages what are called "common interest groups."
Within a group context, one resource person can be trained, who will then be
empowered to pass on the information to the group. Groups are believed to extend
technologies faster than individual farmers. They have also been found to support fellow
members in adoption (Phiri et al. 2004). In the FARM-Africa project, 78% of the project
beneficiaries were said to be non-members of the FARM groups (Mutia, P., FARM-
Africa Meru Tharaka Nithi Dairy Goat and Animal Healthcare Project Progress Report,
January to June 1999). They were benefiting because of dissemination of information
and technologies by the dairy-goat groups, especially at the buck stations.
Farmers have some comparative advantages over what are seen as the more
conventional extension agents. Because they have similar circumstances, usually speak
the same mother tongue and have comparable educational backgrounds, farmers can
communicate well with and are trusted by fellow farmers. Farmer extensionists are able
to reach more people in a more timely fashion than regular agents (Nyakuni, 2001).
Farmers can be trained to lead community-based extension (Cooper & Denning, 1999), or
farmer exchanges can be facilitated in order to share information. Farmer Trainers are
already being educated in areas where the World Agroforestry Centre is working, since
they can effectively pass technologies on to fellow farmers (Cooper & Denning, 1999).
Farmer groups can be facilitated to network with other groups, forming strong farmers'
associations and giving farmers a voice with which to educate other farmers and to
demand services.
Along with advantages, farmers or community-based mechanisms of any kind have
some obvious disadvantages as extension players. They do not have the power or
authority to institute or regulate policy as governments do. They may lack capacity,
resources and the infrastructure that government or private organizations have. The
following issues come from Scarborough et al.'s 1997 book on farmer-led extension and
need to be addressed: The best way to choose farmer extensionists, defining their role,
remuneration for farmer-extensionists, and personal issues and jealousies that may play a
role at the community level.
Within Kenya, informal self-help groups have historically been an important tool of
community development. The colonial government used these groups to help promote
soil conservation, and formed the Department of Community Development to organize
such groups in 1948 (Tiffen, 1992, as cited in Wellard and Copestake, 1993). Following
independence, the harambee (let's all work together) movement brought about more
group formation in order to obtain government assistance. Place et al. (2002) found in
central Kenya that most adults belonged to groups. Women's groups especially are a
ubiquitous part of rural Kenya (Saito, 1994); over 25,000 are registered with the Ministry
of Culture and Social Services.
Due to the reasons discussed above, many are advocating community-based
extension through farmer groups as a means of scaling up technologies (Nyakuni, 2001;
Raussen, Ebong and Musiime, 2001; Wambugu, Franzel, Tuwei & Karanja, 2001).
However, little is known about how farmer groups work in disseminating technologies
and information. There is limited empirical evidence on the performance of groups
(Pretty & Ward, 2001). This points to a need to examine farmer-to-farmer technology
dissemination using farmer groups. The following studies show some of the research
findings on the role of farmer groups in disseminating technology. Factors that play a
role in farmer group success are also described.
Many studies on farmer groups attempt to find out why farmers join groups-what
benefits do they gain from being in a group? In an analysis of farmer groups in cereal
growing systems in the United Kingdom, Wibberley (1997) rated farmers' perceptions of
farmer group benefits in the categories of self-help, motivation, cohesion, and
performance. Some of the highest ratings were with regard to cohesion; giving
friendship, problem sharing and enjoyment received the highest marks. In Kenya, Alawy
(1998) found that women feel that they benefit from being in the group through training,
cash, financial assistance, knowledge gained, and food.
Farmer groups have proven to be a useful way to access a community and to extend
knowledge to other farmers. In Australia, Andreata (2000) found in her study of farmer
groups that they were an efficient way for farmers to share information and experience.
Rouse found in 1996 that being part of a group contributed to knowledge, empowerment,
confidence and ability to make decisions among members. Women's groups were shown
in Malawi to reach more smallholders than customary extension practices, and to be an
efficient way to reach women farmers (Sigman, Chibwana & Matenje, 1994). They are
an important component of farmer-to-farmer extension, helping to coordinate research
and extension. A study by Parkins in 1997 showed that 63% of farmers surveyed in
Embu preferred to approach groups, rather than individual farmers, for information on
tree planting.
Both public and private development partners can facilitate such groups to achieve
their goals by linking them with other groups and service providers (Cooper & Denning,
1999). Geran found in her 1996 study in Zimbabwe that group formation led to increased
links with service providers, as did Rouse in Zambia (1996). Such groups increase the
efficiency, effectiveness and equity of service provision and also help to empower
farmers (Esman & Uphoff, 1984; Geran, 1996).
However, being in a group does not guarantee equal access to services. There may
be differences among groups that lead to inequitable service provision. Alawy (1998)
conducted a study on the Kenyan coast where he examined factors influencing
accessibility of women's groups to extension services. He found that extension tended to
be biased toward male farmers, Christians and tribes from other areas. This was likely
due to the fact that extension workers are mostly male, Christians working in a Moslem
area, and from an "up-country" tribe.
Esman and Uphoff (1984) perhaps conducted the most comprehensive study on
groups. They analyzed a cross-section of local organizations (LOs) from around the
world. Data were gathered from various books, journals, and bibliographies on the
subject of local organization. From this large set of case studies, data was gathered and
analyzed.
The authors put forth the idea that LOs act as intermediaries in rural development;
they intermediate between individuals and the state. Rather than being a part of the
public or the private sector, local organizations rather make up a third sector. Local
organizations, according to Esman and Uphoff, can extend the outreach of public
services, increasing their efficiency. They can also aggregate the demands of rural
people and assist them to solve problems in appropriate ways.
Their 1984 study was based off an earlier study conducted in 1974. The 1974
study indicated that local organizations were necessary for rural development. It also
showed that the most efficient local organizations functioned at more than one level.
Those organizations with links to political or administrative centers that provide
information were also more effective. Esman and Uphoff believe that characteristics of
the poorer members of the community prevented them from taking part in local
organizations. This is in contrast to Parkins' findings below.
Parkins (1997) conducted a study on the mechanisms of group extension of
agroforestry technologies in central Kenya. He termed this "innovation networking" and
found that networking varies by gender, attitude toward participation and recency of
migration. He found that formal organizations tended to provide information to farmers,
while informal organizations usually provided materials. Parkins expected to find that
group participants were the middle class of small-scale farmers, because the poorer
farmers might not be able to afford the financial and labor commitments. However, he
found that the poorer farmers actually were participating in groups along with those of a
more average wealth level. The wealthier farmers were not as heavily involved in
groups.
Another hypothesis in Parkins' study was that group-to-farmer contacts would be
more common than farmer-to-farmer contacts. Because 63% of farmers preferred groups
to individuals for information, this hypothesis was retained. However, respondents also
perceived that there were local experts available, and about half of them approached their
neighbors for networking purposes.
Factors for Group Success
In addition to the reasons why farmers join groups and the role that groups play,
effectiveness of groups is another important area of study. If groups are to be used to
help scale up technology dissemination and to extend more conventional extension, it is
important to know what factors make groups successful in group activities in general, and
extension in particular. Therefore several researchers have examined group performance.
One aspect that must be addressed at this point is the question of what is meant by
the term success. Presumably farmer groups have their own indicators of success.
However, outside agencies working with such groups may also have their definition of
success. From a project or donor viewpoint, adoption data, outputs or quantifiable
benefits from being involved with a group may be indicators. In this study, success was
examined in terms of dissemination of information and technology, not necessarily group
performance on the whole. Success was determined through the groups' own perceptions
of success in dissemination to other farmers, through variables such as number of other
farmers and groups trained, and through external ratings of the groups by FARM and
extension staff.
Stringfellow, Coulter, Lucey, McKone and Hussain conducted a broad study on the
effectiveness of groups in sub-Saharan Africa in 1997. They found that cooperation
among farmers was more successful with small cohesive groups, when conducting simple
activities and by liaising with service providers such as agribusiness. They also found
that groups need internal cohesion and a member-driven agenda. Cohesion was also
assisted by small group size, homogeneity of members and member accountability.
De Haan (1999) studied group dissemination of dairy goats in Tanzania. Her
research was a case study of Heifer Project International's goat group project. She found
that social capital was important in gaining access to goats. Success of the groups was
related to age of the group, spatial distance between members and group function. Older
groups with multiple functions were more successful at dissemination of the technology.
Hambly (2000) examined longevity in women's tree planting groups. Her findings
included the conclusion that unsuccessful groups were related to inequitable social
structures.
Morton et al. (2001) conducted a study on self-help groups and cooperatives in the
dairy industry in Kenya. They found that success of these groups was related to
homogeneity, group size and activities undertaken. In analyzing the Kenyan dairy sector,
both cooperatives and farmer groups were examined as to their success in Morton et al.'s
study. The structural features that contributed most to success were member
homogeneity and starting with a single activity. Group size was shown to have both
positive and negative effects upon success. There was more cohesiveness and sense of
ownership among small groups of farmers (8 to 25 members). However, larger groups
were more likely to function successfully when working with outside agents such as
agribusiness and banks. With relation to outside agents, a high degree of self-financing
led to greater success. Also, heavy external training inputs led to greater success.
Finally, having a member-driven agenda had a negative effect upon group success
(Morton et al., 2001).
Place et al. (2002) studied group performance among small-scale farmers in Central
Kenya. They examined 87 groups and 442 households, and using descriptive analysis
and regression models, were able to gain better understanding of factors that affect group
performance. In the empirical analysis, the authors focused on group structural variables
as factors affecting performance. They found that performance was not correlated with
any particular "easy-to-measure" group characteristic (Place et al., 2002).
They also found that groups were very dynamic and took on new activities. Many
different types of groups were able to take on diverse activities and be successful in them.
Significant factors in explaining the success of groups were purpose of the group and
whether the group purpose had changed over time.
Place et al. found that in certain analyses, group size affected performance.
However, it seemed like middle-sized groups were more successful than the large or
smaller groups. Age of the group was not linked to performance in any of the analyses
(Place et al., 2002).
Conclusion
Many feel that there are currently good agroforestry practices and technologies for
small-scale farmers, such as calliandra for fodder, that have been developed and are ready
to be taken "off the shelf' (Cooper & Denning, 1999; Wambugu et al., 2001). There are
many farmers, in Kenya and elsewhere, who could benefit from such technologies if they
could obtain the necessary information and germplasm. What is lacking is the means of
disseminating this technology to more farmers who could take advantage of it.
Government extension in Kenya today is unable to provide many of these small-
scale farmers with appropriate technologies and information to meet their needs and thus
help to bring about rural development. The issue then is how to extend, or scale up, these
technologies to benefit more low-resource farmers in spite of the limited government
extension. Many approaches have been developed since the reduction of government
extension, such as private extension services and those run by NGOs. Recently,
however, community-based extension has come to the foreground as a means of scaling
up these technologies to have a wider impact in the rural economies (Franzel, Cooper &
Denning, 2001; Noordin, Niang, Jama & Nyasimi, 2001). Farmer groups are an
important vehicle for community-based extension.
Today in Kenya, many technology dissemination approaches exist, with few
studies to show their effectiveness. One important need in the new extension paradigm
that includes community-based extension is to determine the role that community groups
and farmers play in extending technologies, and how they go about disseminating the
information to other farmers. Knowing these mechanisms will contribute to the effort in
scaling up the impact of agroforestry and other research.
In this chapter we have examined extension history and models used in Kenya and
theories of social capital with special emphasis on farmer groups. In Chapter 3 the
methods used to gather the data and research design will be discussed.
CHAPTER 3
METHODS
Deby m6nn gen m6nn.
(Haitian proverb meaning "Behind mountains, more mountains")
In the previous chapter, theories of extension, social capital and farmer groups were
discussed. This chapter describes the methods used to help understand and describe the
role of groups in extending new technologies. It covers the research design, the
population and subjects, sampling procedure, instruments, study variables, data analysis
and means for ensuring validity and reliability.
The goal of this study was to examine the role of farmer groups in technology
dissemination, and to assess what factors make groups effective in extending
technologies among small-scale dairy-goat farmers in Meru Central District of Kenya.
The specific objectives were to
* Examine participation in groups and identify what factors, if any, affect
participation in groups;
* Examine linkages and their outcomes, if any, between farmer groups and with other
extension stakeholders;
* Identify the mechanisms by which farmer groups and their members receive and
disseminate information and new technologies, especially fodder shrubs and
improved dairy-goat breeds;
* Identify the factors characteristic of groups successful in disseminating technology;
and
* Propose policy recommendations to extension and development organizations
regarding farmer groups' roles in extension.
Research Design
A mixed-methods, multiple-stage approach was used to obtain data for the study.
The study consisted of a preliminary phase (four months), survey research (four months)
and a follow-up stage that included stakeholder feedback (two months). The approach
used obtained both qualitative and quantitative information to answer research questions.
Many researchers in social science studies use the mixed-methods approach, also known
as triangulation (Jick, 1983). Triangulation not only allows for an enhanced description
of phenomena, but also helps to validate findings (Hinds & Young, 1987, in Bowen,
1996).
A main component of the research was an in-depth case study of dairy-goat farmer
groups in Meru. A case study is an in-depth look at one individual or social unit such as
an organization or community (Ary, Jacobs & Razavieh, 1996). Case studies look for
variables important to the phenomenon under study, and examine these variables and
their relationships within the social unit. Case studies involve a prolonged time frame-
for this study the researcher spent over 10 months in the field. The goal of case studies is
to understand and describe a phenomenon. Data, both qualitative and quantitative, were
collected through secondary documents and interviews, questionnaires and observations.
Much of the quantitative data were collected from formal surveys and secondary sources.
In light of the goals of the study, a mixed-methods approach was appropriate,
because in order to make associations, predictions, and inferences, one must first
understand a phenomenon. Using both qualitative and quantitative techniques helped to
strengthen and to add validity to the results of the study. Many researchers believe that
qualitative and quantitative methods are not competing, but complimentary (Bowen,
1996; Casey & Kumar, 1993), and indeed, can strengthen the perceived weaknesses of
both approaches.
Population and Subjects
The population of interest to the study was small-scale farmers in Meru Central
District in Kenya. According to FARM-Africa, small-scale farmers are those with
landholdings between 0.25 and 1.5 hectares (Meru Dairy Goat and Animal Health Care
Phase II April 1999-March 2002 Project Review, 2002). However, some of those
interviewed owned larger pieces of land, especially in the lower zones where there were
settlement schemes. The target population was those farmers who were involved in
dairy-goat groups through the non-governmental organization FARM-Africa in the
district, plus other similar farmers who were not members of the dairy-goat groups but
had benefited in some way from the groups. FARM-Africa intentionally worked with
those whom they consider to be the poorer farmers in the area, and with women. The
project also targeted farmers in the medium- and low-potential areas of the districts
(Meru Dairy Goat and Animal Health Care Phase II April 1999-March 2002 Project
Review, 2002), although other groups belong to higher-potential areas.
From that population of small-scale farmers, a sampling frame of farmers and
dairy-goat groups on FARM-Africa or government lists was put together. From the
sampling frame both purposive and random samples (described in Ary et al., 1996) of
individual households was drawn from FARM-Africa and government lists to elicit data
for the study. Purposive sampling was used for key informant interviews (n = 24).
Individual dairy-goat group members were chosen at random from a list of groups where
possible (n = 44). Also of interest to the study were farmers who have benefited from the
group (through group dissemination of information or technology). These non-dairy-goat
group farmers were sampled by asking the farmer groups for lists of people who had
benefited from their group, and then randomly selecting farmers from that list (n = 44).
Units of analysis thus included both farmers (dairy-goat members and non-members) and
the dairy-goat groups. Instead of randomly selecting groups, all of the current dairy-goat
farmer groups in the FARM-Africa project in the district were interviewed (n = 46).
There were other dairy-goat groups in the district that had either just formed or were
supported through another non-governmental organization (NGO) that were not
interviewed.
Sampling Procedure
The dairy-goat groups in Meru Central District that were part of the FARM-Africa
project were chosen for the study for several reasons. To reduce the number of variables,
it was decided to focus on one type of group rather than a variety of farmer groups who
may have had many different activities that they focused on. This also allowed for the
comparison of success in dissemination across groups. The dairy-goat groups were all
focused around the same activity, and there was an adequate number of them (46 in Meru
Central District). They were also in the same geographic vicinity, and worked with the
same organization (FARM-Africa).
Four other dairy-goat groups were also interviewed that were in the same area but
were supported through another NGO, Meru Drylands Farming Project, to see if the
project/NGO itself made any difference. However, their data was not included in the
analysis with the 46 FARM-Africa groups.
The Meru area is quite homogeneous ethnically, which also helped to reduce
variables in the study. The Meru tribe makes up most of the population, with its subtribes
the Tharaka, Imenti and Tigania. However, ethnicity was accounted for in the data
collection to see if it had an effect on the variables studied.
Although FARM-Africa is working with over 83 groups in Meru Central and Meru
South Districts of Kenya, it was decided to interview farmers and groups only in Meru
Central. This was decided for several reasons. Based upon preliminary research and
discussions with another research team that was covering both districts, it was decided
that there were not major differences between farmers in Meru South and Meru Central.
The Meru tribe was the majority in both districts. The districts were formed to cut across
the various agroecological zones, and so both districts had farmers in all the zones. It
was decided that there would be no major difference between the two districts.
Furthermore, it was decided that better quality data could be obtained by focusing on a
smaller yet similar area. The researcher could spend more time collecting data at each
point and more easily return to groups or individuals for clarification. Concentrating
solely on Meru Central would reduce expenses and increase efficiency of researcher time.
Therefore, all of the current FARM-Africa project dairy-goat groups in Meru
Central District were surveyed. Because the project was growing so rapidly, not all the
groups were interviewed. Therefore, there were dairy-goat groups that were not
interviewed, either in Meru South District (and still part of the FARM-Africa dairy-goat
project), or in Meru Central (with Meru Drylands Farming Project), or new groups that
had just recently started and will not be associated with FARM because the project was
ending.
Although FARM-Africa only had 20 dairy-goat groups in Meru Central, many new
groups were associated with the project, albeit without FARM support. These were
known as "extension" groups. They had bought a breeding buck (while "FARM" groups
were given one), and were trained by extension staff or other dairy-goat groups without
the normal FARM-Africa support. There were 26 extension groups as compared to 20
FARM groups.
Instruments
The study consisted of both qualitative and quantitative research methods.
Interviews, non-structured observation and document analysis were the main means of
collecting data that were used for descriptive purposes. A research journal was kept for
daily entry of data and observations.
Topic guides were used for semi-structured interviews (Appendix B). These
included general questions with probes. More formal questionnaires for both individual
farmers and farmer groups were then developed based on this information and from
document analysis (Appendices D and E). The individual instrument included 48
questions, while the group questionnaire had 66.
The researcher developed questionnaires based upon similar survey instruments
from studies in Kenya. Advice and input from key informants and research colleagues
was also sought. Questionnaire content was mainly guided by initial information elicited
from farmers and farmer groups, however, during the preliminary phase of the study.
The Institutional Review Board of the University of Florida approved the study design
and instruments prior to data collection (Protocol #2003-U-371). The researcher
developed questionnaires for both individual farmers and farmer groups.
Survey research includes issues of face and content validity. Face validity means
that the questionnaire appears to be measuring what it purports to measure. Content
validity means that the instrument contains a good representation of the range of meaning
of the construct being measured (Babbie, 1986). An instrument that does not have
content validity will not give an adequate measure of a construct. Researchers
conducting similar research in Meru and in Central Kenya and experts in the United
States checked the instruments for face and content validity. This panel of experts
consisted of four North American researchers and five Kenya-based researchers (three of
whom were based in Meru).
A pilot test of the instruments helped to check for face, content, and criterion
validity. This also helped to ascertain the amount of time needed to administer them.
Five groups and seven individuals were used for pilot testing. Pilot testing was
completed when the questions seemed to make sense to respondents and no additional
questions or pre-coded answers were added. Unclear questions were removed or
changed. Data from the pilot test instruments were used only when the question was not
changed from the pilot tests to the final instrument.
Other threats to validity on study questionnaires included history, measuring
instruments, differential selection of subjects, subject's attitudes, importance of the topic
to respondents, and anonymity of respondents. Multiple indicators were used to measure
constructs on the instruments. The researcher used local languages to ensure that
respondents understood the questions. Indigenous categories and terms were used to
ensure understanding, for instance, in describing wealth categories. The translator was
trained and took part in the pre-test. All research objectives and methods were described
to him and other participants to make sure that they understood what was being sought.
Preliminary data collection ensured that the study was collecting information on matters
important to the farmers being studied. Respondents were assured of anonymity to
increase internal validity. Having multiple members present (e.g., group interviews) also
helps with validation, because the members present are validating each other. If someone
says something wrong, another member can correct the statement during group
interviews. Multiple methods of getting data (through groups, individuals, reports,
records, timelines, and chapati diagrams) were used. Also, at times the same question
was asked on surveys in different ways to make sure the results were valid. To limit the
phenomena of people trying to give an answer that they thought the researcher was
seeking, the researcher did not emphasize the connection with the World Agroforestry
Centre or with FARM-Africa. It was made clear that the researcher was just that, a
student, and not a possible founder, to limit bias in answers.
A further means of ensuring validity was through member checks. This was done
by presenting research results to study participants for verification on the findings. The
researcher followed Babbie's (1986) suggestion to enlist the assistance of others to
confirm or validate researchers' finding. This was done through a stakeholders'
workshop, where farmers, extension agents and NGO/international research center
personnel were present. Findings were presented to stakeholders to obtain feedback and
to ensure validity of results. This is known among anthropologists as "repatriation" of
information, and is a way to return knowledge to community. Reports in the vernacular
were also given to each group that had participated in the study. Reports were also left
with all of the other stakeholders, including farmers, FARM-Africa staff and extension
personnel.
Data Collection
Collection of data occurred in both the greater Meru area and Nairobi, and was on-
going throughout the study. As mentioned above, the study consisted of a preliminary
phase, a survey phase and a follow-up phase.
In line with the qualitative inductive approach, the study began with preliminary
data collection, which allows for the development of appropriate research questions.
Content analysis, direct observation, and semi-structured interviews with key informants
were used throughout the study, but especially during the preliminary phase to help guide
the inquiry, identify hypotheses, and collect rich qualitative data.
The more informal preliminary phase began with sondeo-like semi-structured
interviews. Sondeos (derived from "sounding out" in Spanish) are a type of rapid rural
appraisal where researchers quickly gain information in an informal, non-threatening
manner (Hildebrand, 1981). Topic guides were used which included basic questions and
probe questions to elicit yet more responses (Appendix B). Questions were open-ended,
and the interview followed the thought processes of the person being interviewed, rather
than sticking to a set questionnaire (as described in Chambers, 1997). This allows issues
to come out that perhaps the researcher missed during the literature review. However,
because the guide is structured, it allows the researcher to obtain consistent and reliable
data. Information obtained was mostly qualitative. These interviews were conducted
with key informants who were well informed and recognized in the community.
Farmers, farmer groups, extension workers and NGO personnel, and international
research center personnel were all interviewed to obtain and triangulate data.
Content or document analysis was another important tool during the preliminary
phase. This is the use of existing records and documents to obtain information on a
community or group of subjects. It allowed the researcher to assess the project prior to
semi-structured interviews and survey research. In this study such records included
FARM-Africa and World Agroforestry Centre (WAC) reports and government literature
such as the District Development Plan.
Qualitative techniques were used during the formal survey research as well.
Participatory techniques such as social mapping and timelines were used during the
survey phase to examine variables, and to provide both qualitative and quantitative data.
Social mapping is the construction of a representation of the services and facilities
that are available to a particular community. Venn (or chapati) diagrams were used
during the survey phase to allow the groups to show the relative importance of various
organizations or individuals within a community, and how they are interrelated. This was
done by giving respondents circles of paper of three different sizes to show relative
importance of the groups or individuals, and to show how they related to each other by
their proximity to the group and to one another.
Group timelines were used to elicit information on the history of the farmer groups.
This was done by using a piece of flip chart paper to write down group events while
asking them questions such as, what have been the changes in membership over the years
(has it increased, declined, stayed the same)? What have been your successes and/or
failures? When did the group start, how did it start, what have your activities been, has
gender composition changed and what events impacted your group? Has the group
changed focus since it started?
These details helped to give a history of the group and the major events affecting
them (AMREF, 1997). The timelines answered many of the questions on the instruments
using a relaxed way of story telling, rather than question-and-answer format with the
researcher appearing to extract information from the informants. This method (used at
the beginning) helped to put the group at ease, and assisted them to remember many
details about their group that helped to jog their memories later on in the interview. The
timelines also provided valuable information about the groups that helped to show where
they have come from and why they were the way they were.
Following the preliminary qualitative data collection, questionnaires were
developed and then administered via a translator to answer research questions, obtain
quantitative data and to gather more qualitative information. This helped to further
explore the themes that were brought out through the initial data.
A translator/research assistant was used to collect information from farmers using
questionnaires due to the language barrier. An assistant was procured through FARM-
Africa who was experienced in field data collection in the area and was of the same
ethnic group as the respondents. The assistant spent several years working for a similar
NGO and received training in participatory methods of working with communities. The
assistant was further trained for this particular study, and tested for accuracy by having
him ask the questions in Kimeru, the local language, to another person who then repeated
what he said in English.
The formal questionnaires were written in English. However, the interview was
conducted orally in Kiswahili and Kimeru with both groups and individuals. The
researcher asked the questions in Kiswahili and the research assistant translated into
Kimeru. Many of the farmers spoke Kiswahili but were more comfortable in Kimeru.
Answers were then recorded in English. Closed questions are typically used in survey
research, but some open-ended ones were asked as well.
The survey research took place with both groups and individual respondents.
Group interviewing is beneficial in that it is low-tech, rapid, and low-cost in comparison
to individual interviews. Group interviews also capitalize on group dynamics and take
advantage of the synergistic effect of people's conversations, where ideas can stimulate
more and richer responses (AMREF, 1997; Debus, 1986). They are a means of
validation as well; by having multiple members present, veracity of answers is better
ensured.
For the group interviews in this study, typically four to six farmers from one dairy-
goat group were interviewed together over a period of about two hours. The group
questionnaire consisted of 66 questions and participatory activities. During the
interviews, group timelines and Venn (chapati) diagrams were drawn up by the group
members. The interviewer sought to allow everyone to be heard and to ensure that no
one person would dominate. In addition to group members, the researcher, research
assistant/translator and an extension staff member were present. Extension staff had
worked with the group over a period of years training them, and thus were able to guide
the researchers to the group and also to maintain the rural protocol regarding
introductions of visitors to the groups. Although this was something that could contribute
to bias on the respondents' part, it was a necessary part of protocol that could not be
avoided. The research team attempted to avoid bias in having the agents present by
plainly stating that although the agent was present, they wanted the groups to answer
truthfully. The only role that the agent played was to introduce the group; he or she was
not to answer any questions. Finally, follow-up interviews with individuals were
conducted with no extension agent present to obtain further information and to confirm or
deny what the group had said.
Following the group interview, individuals were sought to both corroborate the
group information and to obtain information at the household level. One group member
and a non-member who had benefited in some way from the group were sought at each
interview site. The individual questionnaire consisted of 48 questions, and took between
30 to 60 minutes to complete. Typically, group members were chosen randomly from a
list obtained from FARM-Africa. Non-members were randomly selected from a list
provided during the group interview.
Data Analysis
Quantitative data from the questionnaires were entered into the Statistical Package
for the Social Sciences (SPSS) software (George & Mallery, 2001) and analyzed.
Descriptive analyses of the data were a major outcome. Correlational techniques and
measures of association such as correlation coefficients (Pearson's product moment) and
multiple linear regression were used to examine and predict relationships among the
study variables. The logistic regression model was also used to deal with binary
responses. Comparisons of groups were made using contingency tables and cross-
tabulations, and tested for significance with tests such as chi-square.
Qualitative data was analyzed by hand by reducing them to workable categories.
The researcher then sought to discover themes, patterns, associations, explanations and
general statements about the relationships among categories of data (Marshall &
Rossman, 1999).
Another tool that was used for data collection was GIS (geographic information
systems). GIS is basically a software package that combines maps and database
information in a single analytical tool. With GIS, the researcher was able to map the
dairy-goat groups. Further information on Meru such as agroecological zone, altitude,
markets, forests, rivers and roads were added to this information to determine how they
were all related.
Validity and Reliability of the Results
Validity
Validity essentially means the closeness of a research finding to physical reality
(Chambers, 1997). The validity or integrity of qualitative data is measured by
trustworthiness, dependability and credibility. Threats to validity in the research study
might come from lack of understanding by respondents or enumerators, from unclear
questions, non-random sampling, failure to pay attention to the theoretical basis, and
failure to record and describe the research process and study area. Threats to validity of
the instrumentation process were discussed under the instruments section. Other threats
to validity of the overall study are discussed below.
One possible threat to validity was through history. Perhaps respondents had
certain attitudes toward FARM, government extension or even other farmers. Also, since
the goat project originated in the United Kingdom where many people are white, seeing a
white researcher may lead farmers to think that answering questions in a certain way
would get them benefits. Issues of threats through history were avoided by carefully
explaining the research project and the position of each person present with the research
team. It was also stressed that the researcher was essentially a Kenyan, with a home and
family in Kenya. Local languages were also used to stress this fact. It was also plainly
stated that there would be no benefits, except for information, from the research study.
Respondents were assured that a report would be sent to them in the local language at the
end of the study. All participatory materials developed by the farmers (such as chapati
diagrams and timelines) were left with the group.
Internal validity is yet another type of measure of the study's trustworthiness.
Internal validity mostly relates to experimental research, but can be a factor in non-
experimental research designs as well (Ary et al, 1996). Internal validity of the research
design was improved through triangulation of data sources and the merging of qualitative
and quantitative methods. Furthermore, having multiple people validating the data also
contributed to internal validity.
External validity refers to the extent that results of the study can be generalized
(Ary et al., 1996). Population external validity, if high, means that results can be
generalized to the larger population (Ary et al., 1996). In this study it would mean that
results seen in a random sample of farmers in Meru could be inferred or generalized to
the general population of farmers there. Generalizations or inferences can be made to the
population only if a random sample was selected for the study. In qualitative research,
generalizability is not always a goal, and the main purpose is more exploratory than
explanatory. However, many social scientists do recognize the importance of designing
studies so that findings may also help in understanding other situations.
From the study population, purposive and random samples were drawn to elicit
data for the study. Survey data was collected from farmers chosen randomly where
possible from a list of farmers from FARM-Africa and/or government lists. This allowed
for higher population validity. Purposive sampling was used for the more qualitative
aspects of the study, especially to interview key informants.
High ecological external validity means that the same results would appear if the
study was conducted in another location or setting (Ary et al., 1996). High ecological
validity was ensured by carefully describing the study location, farming systems, and
farmer groups. By understanding the location and also all of the methods used, the study
should be able to be replicated elsewhere.
Finally, high external validity of operations means that if another researcher were
to conduct the study, the results would be the same (Ary et al., 1996). This was ensured
through the detailed description of the study daily in the research journal.
Validity in the qualitative portion of the research was handled overall through the
field research. Because the researcher "is the data-gathering instrument" (Ary et al.,
1996, p. 478), what she does is very important. A document trail of interviews, research
processes and findings was kept in the form of a research journal. This record of events
and observations, together with the raw data, can be audited and serves as a means of
checking the legitimacy of the study. High validity of operations was ensured by the
careful use of a research journal where all observations and details of the study are kept.
Anyone reading this should then be able to replicate the study.
Validity was also verified by triangulation-the use of multiple sources of data,
methods of collecting data (various people, times, and settings), multiple investigators
and drawing upon multiple theoretical bases (Ary et al., 1996). Groups, individuals and
organizations were interviewed to triangulate the data. Debriefing with peers, member
checks and the use of expert consultants were other means of verification of validity (Ary
et al., 1996; Bottorff, 2000). Regular supervision while in Kenya was provided through
the international research center World Agroforestry Centre, and debriefing also took
place among fellow researchers in Meru who were working on a similar project at the
same time. Using inductive analysis and grounded theory helped to increase credibility,
because the researcher was basing her hypotheses or findings on what had been found in
the field. Finally, a prolonged engagement in the research setting helps to establish more
credibility or validity. The researcher spent over 10 months in the field collecting data.
In summary, validity was assured by
* Triangulation;
* Submitting questionnaires to a panel of experts;
* Pre-testing the instrument;
* Comparing observations to the literature;
* Training assistants;
* Assuring anonymity of respondents;
* Using random sampling where possible;
* Adequately describing the setting; and
* Keeping a complete record of design and methods.
Reliability
Reliability is the extent to which an instrument is consistent in measuring, or to
which a particular technique will always yield the same result (Babbie, 1986). It can be
compared to precision.
Error that affects reliability can come from various sources. The respondents may
have been tired or ill, or not be in the mood to talk to the researcher. On the
questionnaires, there may have been ambiguous questions or an enumerator may not have
understood what information the researcher was attempting to obtain. The atmosphere in
which the instrument was administered may also have affected reliability. Sometimes
during the interviews, it began to rain, or other types of interruptions occurred during
interviews. In such cases the interruption was noted and the data collection continued.
Finally, there may be errors in entering the data (Dedrick, 1997, Foundations of
educational research, unpublished manuscript, University of South Florida).
Inter-observer or inter-rater reliability is also important for reliability of qualitative
data. This can be increased by making sure that the enumerators are well trained by the
researcher, which was done. Because the researcher used just one assistant, this also
helped to reduce variability in observation. Also, following each interview, discussions
were held among the researcher, assistant and local extension staff as to the findings.
This helped to ensure that researcher observations were correct.
There is no true way to assure reliability in a purely qualitative study. In this study
it has been established mostly through documentation. A document trail of interviews
and findings was kept in the form of a research journal to ensure reliability. This record
of events and observations, together with the raw data, is a means of checking the
reliability of the study. For instance, during data analysis, if there is a question about the
responses, the researcher can return to the field notes to see what other factors may have
been coming into play. If there are any questions regarding the reliability of the findings,
the researcher can return to the record trail to show what was done. Random samples can
also be taken from the journal. The assumption is then made that the study is reliable
overall if random samples of this information are determined to be reliable.
Reliability was further assured by having multiple indicators to measure the
constructs on the questionnaires. The researcher also sought to only ask questions that
respondents are likely to know the answer to (Babbie, 1986). This prevented them from
making guesses or falsifying information to please the researcher.
An important consideration in research conducted in another language is the use of
translators to collect data with farmers who do not speak the same language as the
researcher. Ensuring that translators understand the questions and can translate them
with the right degree of meaning is also important. This was ensured by having the
translator ask the questions to another party in Kimeru, and the third party translated back
into English so that the researcher could check that he was translating accurately.
Reliability was further ensured by having the same translator in all except one of the
interviews, where it was not possible for him to be there.
In administering a questionnaire, reliability is a function of the length of
instrument, heterogeneity of the population, ability of the respondent, the nature of the
variables, and the number of items (Ary et al., 1996). Longer tests are more reliable. If
respondents are more heterogeneous, reliability will be higher. If respondents do not
understand a question because of their ability, they may guess and so affect the reliability
coefficient. Some variables are easier to measure and thus give higher reliability.
Limited redundancy was built into the questionnaire to assess the consistency of
responses to a particular question.
Reliability was ensured in the following manner:
* Pilot testing the instruments;
* Training the assistant;
* Maintaining a document trail of research findings;
* Utilizing triangulation; and
* Obtaining reliability coefficients from test constructs.3
This chapter has examined the research design and methods used, the qualitative
and quantitative paradigms, the population sampled, instruments used, the procedures,
3 Cronbach's alpha for the adoption index, a measure of dairy goat technology adoption by group
neighbors, was 0.69.
73
data analysis and issues of validity and reliability. The next chapter will examine the
results related to the objectives of the study.
CHAPTER 4
RESULTS
Before we started we were doing nothing and now we're doing something.
-Farmer in Meru Central
In Chapter 3 the methods used in the study were discussed. This chapter will
discuss the research findings, presenting information on the FARM-Africa dairy-goat
project and the groups that have been formed as a result of the project. It will also
describe the subjects of the study. Finally, the first four study objective results will be
presented, covering who participates in groups, group linkages, mechanisms of
dissemination, and factors that make groups successful in extension.
The Food and Agricultural Research Management (FARM)-Africa Project and
Dairy-Goat Groups
FARM-Africa is a British non-governmental organization (NGO) that has been
working with projects in eastern and southern Africa since 1985. FARM-Africa initiated
a dairy-goat project in Meru Central and South Districts of Eastern Province of Kenya in
1996, targeting poorer farmers in the middle and low potential zones. The project was
called the Dairy Goat and Animal Healthcare Project, and attempted to improve the
productivity of farm animals in the area through breed improvement, better management
and sustainable animal healthcare.
Following preliminary surveys in 1994 and 1995, the project was implemented in
1996. The poorest members of the communities in the greater Meru area were targeted
for participation in the project. Such people were identified following awareness and
sensitization meetings at the community level, which included FARM-Africa and
extension staff, local administration (chiefs), and community members. Criteria for
poverty, based on the local communities' indicators, included an inability to send
children to school, lack of regular income, temporary housing, having no cattle, and small
land size relative to the area.
FARM-Africa chose to work with organized farmer groups to efficiently target
communities. However, they did not use established groups, because they were targeting
the poor people of the communities. Therefore, most of these groups were formed for the
purpose of the project, resulting in about 44 groups that were created with assistance
from the local communities, chief, and extension staff. These dairy-goat groups were
then trained on group dynamics, goat management, housing, and breeding. Selected
members from each group were trained as buck keepers and community animal health
workers (CAHWs). Extension staff were also trained at this time using DELTA
(Development, Education, Leadership, Training and Action), a community mobilization
training program.
One of the main mechanisms for improving animal productivity was through the
establishment of pure Toggenburg buck stations located within the communities for
breeding with local goats. Genetic improvement of local goats was to occur through the
breeding of local and Fi crosses to the pure Toggenburg bucks. The project was unique
in that the breeding was to be done at the local level instead of the typical government or
private breeding stations. The bucks were given to the groups at no cost, with the
understanding that they belonged to the project, and could be retrieved in cases of
mismanagement.
These buck stations were thus owned and managed by groups of farmers. Group
members were to bring their local goats to the buck station to be served for free or for a
small fee. Community members could also obtain breeding services for a slightly higher
fee. The resulting offspring (Fis and F2s) were known as "improved" goats. Each of the
44 dairy-goat groups received a buck for their buck station. The project also included
breeding stations to breed more pure Toggenburg dairy goats. Breeding stations
consisted of a pure buck and four pure does. They were given to certain groups on the
condition that the pure goats were to be repaid to the project in kind from their offspring.
These repayment goats were then used to start additional breeding stations. At the time
of the research, there were 83 buck stations and 48 breeding stations within Meru Central
and South Districts (the project area).
The project also established two umbrella organizations, the Meru Goat Breeders'
Association (MGBA) and the Meru Animal Health Workers' Group (MAHWG). These
were to provide a forum and support for the goat farmers, CAHWs, animal health
assistants (AHAs), and veterinarians.
The activities of the MGBA included safeguarding the Toggenburg and its
upgrades, buck rotation, training new groups, organizing shows and auctions, sourcing
markets for members, and management of the seed bank. The MGBA also provided seed
loans to member farmers due to drought in the area. Maize and bean seed was provided
to the dairy-goat groups, which was then repaid in cash with 10% interest. The members'
goats were used as collateral on the loan. Goats in the project were sold through the
MGBA, which received a commission. The MGBA also assisted in agricultural shows.
Groups and/or individual farmers were encouraged to formally register with the MGBA.
The Meru Animal Health Workers' Group was provided with a loan of Kenya
shillings (Ksh.) 200,000 (USD 2,857) to start a credit scheme for its members to open
drug shops, purchase drug supplies, and so forth. It also helped members by purchasing
veterinary supplies in bulk. MAHWG was also meant to be a forum for experience
sharing among members.
In 2004, FARM started phasing out the project. The MGBA was therefore being
coached to take over many of its responsibilities such as extension activities and rotation
of breeding bucks. At the time of the research, the MGBA was starting to charge a small
fee for services such as training. The MGBA also began collaborating with private
breeders and with other NGOs such as the Meru Drylands Farming Project (MDFP), part
of another organization called SOS-Sahel.
Although 44 groups were formed and trained through FARM-Africa, there were
over 100 dairy-goat groups in the greater Meru region at the time of the study. Through
sensitization of communities and training held at agricultural shows, many more farmers
became interested in joining the project, but they were not chosen due to the criteria
mentioned above. Many farmers therefore formed their own groups, arranged to receive
training, and purchased a buck themselves. Extension staff, the MGBA or the older
FARM-Africa groups trained most of these newer groups. Because of this, the original
44 dairy-goat groups were called "FARM" groups and the new ones that were not
officially supported by FARM were called "extension" groups. Both extension and
FARM-Africa were involved with all of the groups; however, FARM groups were not
exclusively working with FARM nor extension groups working with extension only.
This terminology was used to distinguish the original project groups from the newer ones.
Ultimately, all of the groups were under the Meru Goat Breeders' Association, which was
composed mostly of farmers, but had staff from FARM-Africa and from the Ministry of
Livestock Development and Fisheries.
Groups were located throughout the old Meru and Tharaka-Nithi Districts, which,
during the project, were subdivided into Meru North, Meru Central and Meru South
Districts. This study surveyed only groups in Meru Central due to time and financial
constraints. Staff from FARM, extension and the World Agroforestry Centre (WAC) felt
that this was acceptable due to the fact that there were no major differences between
Meru Central and Meru South Districts. Agroecological zones and tribal composition
were much the same in Meru Central and South.
Therefore 50 groups were interviewed during the research. This included 20
FARM groups, 26 extension groups and four Meru Drylands Farming Project (MDFP)
groups (Figure 4-1). The MDFP groups were interviewed to compare the 46 FARM-
Africa groups with other groups that were in the same district but connected with another
NGO. The MDFP groups are the four groups at the top of the map (Kamarete,
Kamakambi, Mumiri and Karimene). These groups were not included in any of the data
analysis in this chapter, but will be discussed in Chapter 5.
Groups were located in three divisions in the district (14 groups in Abothoguchi
Central, 13 in Abothoguchi East and 19 in Miriga Mieru East). The four MDFP groups
were in the nearby Buuri division. Average age of the groups was six years. Average
size was 23 members. Fifty-nine percent of group members were female.
The groups were distributed across most of the agroecological zones of Meru
Central District. Group locations ranged in altitude from 905 to 1795 m, with the average
Figure 4-1. Dairy-goat group distribution within Meru Central District
being 1279 m. Many of them were concentrated in the coffee and marginal coffee zones,
where there is a higher population density. On the slopes of Mount Kenya where tea and
coffee were grown, land sizes were much smaller, while in the drier, lower zones, farmers
tended to have larger pieces of land. Bonferroni post hoc procedures revealed significant
differences in land sizes owned by individual farmers (both group members and non-
members) between two of the three divisions studied in the district. Abothoguchi
Central, on the mountain slopes, averaged 2.94 acres, while Miriga Mieru East averaged
5.55 acres (p < .02). Abothoguchi East division had an average land size of 4.13 acres.
Altitude was also significantly different in the three divisions. Abothoguchi Central at
1431 m was significantly different from Miriga Mieru East at 1222 m (p < .01).
Abothoguchi Central was also significantly different from Abothoguchi East at 1198 m (p
< .01). Miriga Mieru East and Abothoguchi East were not significantly different,
according to Bonferroni analyses.
Fodder Training and Dissemination in the Project
Farmers were taught during the project to house their goats in zero grazing units.
Water and fodder were brought to the goats, and manure could be collected below the
unit. The units consisted of a raised wooden structure with a roof. Project stakeholders
(farmers, extension officials, and FARM personnel) discovered that feed was a limiting
factor for dairy-goat farmers, especially during the dry season. Because of these issues,
fodder was an important component of the project.
Fodder training and germplasm dissemination in the project and district were a
confusing issue because of the number of organizations involved. Players involved in
fodder shrub technologies included FARM-Africa, the World Agroforestry Centre, Meru
Drylands Farming Project, Kithima Tree Nursery, Kamutune Tree Nursery, Kenya
Agricultural Research Institute (KARI), the KARI project ATIRI (Agricultural
Technology and Information Response Initiative), and government extension; specifically
the Ministry of Livestock Development and Fisheries. Also, the Meru Central Dairy
Cooperative gave out fodder for dairy cattle in the district, and British American Tobacco
(BAT) gave out leucaena for firewood to cure tobacco in the lower zones. (This may
have turned some farmers against leucaena as a fodder shrub, since the species given out
by BAT-probably Leucaena leucocephala-was weedy, and some farmers therefore
were reluctant to plant any fodder.)
Germplasm was therefore provided to the dairy-goat groups by all of the above
institutions. The main species disseminated and planted for fodder were Calliandra
calothyrsus, Leucaena trichandra and Sesbania sesban. These species that were
developed and promoted especially for fodder for dairy animals were known as
"improved" trees or shrubs. All germplasm in the project was provided free of charge.
Bulking plots were also developed at certain dairy-goat groups. A bulking plot was
a site where cut fodder could be collected to use as animal feed. The bulking plots were
also to supply germplasm for other farmers and to provide income to the groups through
sale of fodder. The land for the sites was donated public land.
In 1997 and 1998 there was a little work done on fodder trees under the FARM-
Africa project. The main distributions through FARM were in 2000 and 2002. During
the rains (October through December) FARM distributed 30,000 seedlings to most dairy-
goat groups through the ATIRI project. Between July and December 2002, 90 dairy-goat
farmers were trained in nursery establishment and fodder conservation. During the same
time 27 nurseries were established and 2,800 seedlings distributed, and 80 farmers went
on a study tour to KARI-Embu, a nearby research center.
During the entire project, 200,000 seedlings were distributed, according to FARM-
Africa (Ahuya, Okey, Kitalyi, Mutia & Oduo, 2003). Germplasm for the dairy-goat
groups was obtained through Kithima Tree Nursery and Kamutune Farm in Meru Central
District. The British Department for International Development (DFID) funded the
seedling distribution through FARM. FARM reported that there were 500 people who
had planted forage in their project area (Mutia, P., FARM-Africa Meru Tharaka-Nithi
Dairy Goat and Animal Healthcare Project Progress Report, January to June 1999).
World Agroforestry Centre reported about 665 farmers in the district using the fodder
tree technology (Franzel & Wambugu, 2004).
World Agroforestry Centre's work also opened the way for ATIRI-Agricultural
Technology Information Response Initiative, which was run through KARI. This
initiative promoted training, cross-region tours, and on-farm field days. The ATIRI
project then went to the same dairy-goat groups that the World Agroforestry Centre
trained, making it difficult to differentiate between the different fodder trainers and
suppliers. The ATIRI staff involved both extension and FARM staff for their training.
With ATIRI the fodder training expanded to more divisions than what FARM/WAC had
covered. Trees were given out in July of 2000; 70 to 100 trees per household. KARI-
Embu conducted training of both group members and non-members at the chief baraza
(public meeting) in both districts. Table 4-1 shows the different organizations that were
involved in training the dairy-goat groups in fodder tree technologies.
Table 4-1. Source of fodder training for the dairy-goat groups (n = 46)
Source of training f Percent of groups
FARM-Africa 21 46
Government extension 18 39
KARI 12 26
Other 5 11
aPercentages do not total 100 because more than one group could provide training
FARM-Africa was the main source of training, followed by government extension.
KARI also played an important role. Other training sources included individual farmers
and members of the MGBA.
Sixty percent of individuals interviewed (both dairy-goat group members and non-
members) had planted improved fodder. It appeared, however, that the amount of
improved planted fodder was quite limited except in a few cases. KARI and WAC
recommended 150 to 200 trees per goat. However, the average number planted by
individual farmers interviewed was only 34, with a range of 0 to 500 (Figure 4-2). On
83
the other hand, farmers interviewed (both members and non-members) had an average of
three improved goats (Figure 4-3). Farmers seen during the fieldwork were almost
always using local fodder for feed for their goats. Napier grass (thara or Pennisetum
purpureum) was a popular feed. Many of the goats were fed banana leaves and local
hedge shrubs such as Tithonia diversifolia and Lantana camera. Mulberry (ntaratare or
Morus rubra) was also being used. Many farmers did not know what types of improved
fodder trees they had planted nor how many trees they possessed.
70
60
50
40
I-
S30
20
10 Std. Dev = 84.07
Mean = 34.0
0 N = 88.00
0.0 100.0 200.0 300.0 400.0 500.0
50.0 150.0 250.0 350.0 450.0
Total improved fodder planted
Figure 4-2. Distribution of fodder planted by individual farmers (n = 88)
4U 1
30
C 20
10
Std. Dev = 2.90
Mean = 3.4
0 N = 88.00
0.0 2.5 5.0 7.5 10.0 12.5 15.0
Number of improved goats
Figure 4-3. Distribution of improved goats owned by individual farmers (n = 88)
Seventy percent of the dairy-goat groups reported that all of their members had
planted at least some amount of improved fodder. Forty-six percent of groups said
"many" members had planted 1 to 50 trees. Fifty percent of groups said "some"
members had planted between 51 and 100 trees. Thirty-nine percent of the groups said
"some" members had planted over 100 trees.
Regarding their dissemination of fodder tree technologies, the dairy-goat groups
replied as follows during the interviews: 60% of their non-member neighbors had
planted "some" fodder shrubs, and 29% had planted "many." Although many of the
group members had planted fodder trees, often it was just a few trees. Unfortunately,
after one of the major seedling distributions there was a prolonged drought, and many
farmers and groups reported that the trees they had planted died. Others were given seed
that did not germinate (possibly due to lack of knowledge on germination techniques;
seed such as Calliandra calothyrsus requires soaking in water for 48 hours prior to
germination).
Description of Area and Subjects
Individuals interviewed had an average of 4.9 persons per household (Table 4-2).
In Meru the households were complex, as they are for small-scale farmers worldwide.
Therefore, when households were discussed with individual farmers, informants were
referring to those people who slept on the farm and were fed by the person or household
being interviewed.
Most of those interviewed lived in a house with a permanent (usually tin or mabati)
roof and timber walls (Table 4-3). Most respondents obtained water from a pipe, stream
or borehole/well. The other 5% obtained water from other sources, usually piped water
or a borehole at their neighbor's house.
Table 4-2. Descriptive data for individual farmers (n = 88)
Variable Minimum Maximum M SD
Years of education 0 14 6.68 3.55
Total household members on-farm 1 9 4.90 1.86
No. improved cattle 0 9 1.75 1.63
No. local cattle 0 10 0.49 1.52
No. improved goats 0 15 3.41 2.90
No. local goats 0 4 0.76 0.96
Total no. cattle and goats owned 0 20 6.39 3.77
Total no. animals sold last year 0 8 1.41 1.61
Individuals interviewed had an average of 6.7 years of education, varying from 0 to
14 (Figure 4-4). The Kenyan education system has eight years of primary school, four
years of secondary and four years of university. There are also two-year technical
colleges.
|
