ECONOMICS WORKING PAPER 03-06
Adoption of Maize and Wheat
Technologies in Eastern Africa:
A Synthesis of the Findings
of 22 Case Studies
CHERYL R. Doss, WILFRED MWANGI, HUGO VERKUIJL, AND HUGO DE GROOVE
E C O N O M I C S
Working Paper 03-06
Adoption of Maize and Wheat
Technologies in Eastern
A Synthesis of the Findings
of 22 Case Studies
Cheryl R. Doss, Wilfred Mwangi, Hugo Verkuijl, and Hugo de Groote*
International Maize and Wheat Improvement Center (CIMMYT), Apartado Postal 6-641, 06600 Mexico, D.F.,
Mexico. The views expressed in this paper are the authors' and do not necessarily reflect the views of CIMMYT.
CIMMYT (www.cimmyt.org) is an internationally funded, nonprofit, scientific research and training
organization. Headquartered in Mexico, CIMMYT works with agricultural research institutions
worldwide to improve the productivity, profitability, and sustainability of maize and wheat systems for
poor farmers in developing countries. It is one of 16 food and environmental organizations known as the
Future Harvest Centers. Located around the world, the Future Harvest Centers conduct research in
partnership with farmers, scientists, and policymakers to help alleviate poverty and increase food
security while protecting natural resources. The centers are supported by the Consultative Group on
International Agricultural Research (CGIAR) (www.cgiar.org), whose members include nearly 60
countries, private foundations, and regional and international organizations. Financial support for
CIMMYT's research agenda also comes from many other sources, including foundations, development
banks, and public and private agencies.
F U T U R E- Future Harvest builds awareness and support for food and environmental research for
HARV/E ST a world with less poverty, a healthier human family, well-nourished children, and a
better environment. It supports research, promotes partnerships, and sponsors projects
that bring the results of research to rural communities, farmers, and families in Africa, Asia, and Latin
International Maize and Wheat Improvement Center (CIMMYT) 2002. All rights reserved. The opinions
expressed in this publication are the sole responsibility of the authors. The designations employed in the
presentation of materials in this publication do not imply the expression of any opinion whatsoever on
the part of CIMMYT or its contributory organizations concerning the legal status of any country, territory,
city, or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries. CIMMYT
encourages fair use of this material. Proper citation is requested.
Correct citation: Doss, C. R., W. Mwangi, H. Verkuijl, and H. de Groote. 2003. Adoption of Maize and Wheat
Technologies in Eastern Africa: A Synthesis of the Findings of 22 Case Studies. CIMMYT Economics Working
Paper 03-06. Mexico, D.F.: CIMMYT.
Abstract: This paper synthesizes the findings of 22 micro-level studies on technology adoption carried
out by the International Maize and Wheat Improvement Center (CIMMYT) with national agricultural
research systems in Ethiopia, Kenya, Tanzania, and Uganda during 1996-1999. The authors found that
technology adoption is taking place across Eastern Africa but considerable scope remains to improve the
productivity of smallholder agriculture in higher potential regions with high levels of adoption.
Extension was the variable most highly correlated with technology adoption, and extension services
continue to play an important role in disseminating information on new varieties and how to manage
them. Despite the usefulness of the micro-study results, especially for priority setting and impact
assessment, future adoption studies can be improved by standardizing definitions across studies and
using sampling techniques that allow results to be generalized across wider areas. Finally, the paper
suggests that maize and wheat breeding research should be made more relevant to the preferences and
circumstances of farmers, that the link between research and extension should be strengthened and
include the private sector and non-governmental organizations, that policies should support the
development and expansion of efficient markets for inputs and outputs, and that rural credit systems
should be strengthened.
AGROVOC descriptors: Maize; Wheat; Innovation adoption; Technology transfer; Plant breeding; Small
farms; Case studies; Fertilizers; Private sector; Nongovernmental organizations; Research institutions;
International organizations; Ethiopia; Kenya; Tanzania; Uganda
AGRIS category codes: E14 Development Economics and Policies
F30 Plant Genetics and Breeding
Additional Keywords: Eastern Africa
Dewey decimal classification: 338.16
Printed in Mexico.
1 Eastern African Adoption Studies
3 To What Extent have Farmers Adopted Improved
Technologies for Maize and Wheat in Eastern Africa?
8 Which Farmers are Using Improved Technologies?
12 Farmer characteristics
16 Technology characteristics
17 Farmer objectives
17 What are the Main Obstacles to Farmers Adopting
18 Availability of information
24 Are Improved Seeds and Fertilizer Available?
28 Implications for Policy and Research
31 Appendix 1. CIMMYT/National System Eastern African
2 Table 1. Description of adoption studies, Ethiopia, Kenya, Tanzania, Uganda,
4 Table 2. Percentage of sampled farmers adopting improved technologies, Ethiopia,
Kenya, Tanzania, Uganda, 1996-1999
9 Table 3. Summary of estimations of adoption of improved wheat and maize, Ethiopia,
Kenya, Tanzania, Uganda, 1996-1999
11 Table 4. Summary of estimations of adoption of fertilizer, Ethiopia, Kenya, Tanzania,
20 Table 5. Availability and sources of information on improved technologies, Ethiopia,
Kenya, Tanzania, Uganda, 1996-1999
23 Table 6. Percentage of farmers using formal credit, Ethiopia, Kenya, Tanzania,
24 Table 7. Percentage of surveyed farmers reporting unavailability of seed or fertilizer,
Ethiopia, Kenya, Tanzania, Uganda, 1996-1999
The authors thank CIMMYT science writers Satwant Kaur and Mike Listman for their editorial
assistance and designer Antonio Luna Avila for his help with layout.
Adoption of Maize and Wheat Technologies in Eastern Africa:
A Synthesis of the Finding of 22 Case Studies
Cheryl R. Doss, Wilfred Mwangi, Hugo Verkuijl, and Hugo de Groote
The adoption of improved technologies for staple crop production is an important means to
increase the productivity of smallholder agriculture in Africa, thereby fostering economic
growth and improved well being for millions of poor households. Yet, for much of Africa,
basic descriptive data on the technologies used by farmers have not been available. In
contrast to many other parts of the world, many African governments do not collect or
report such data. Without basic, descriptive information about who is adopting technologies
and who is not, it is difficult to formulate policies for increasing agricultural productivity.
To compile data and improve the capacity of local institutions to conduct technology
adoption studies, the International Maize and Wheat Improvement Center (CIMMYT)
collaborated with national agricultural research systems in 22 micro-level studies of
technology adoption in four countries in Eastern Africa-Ethiopia, Kenya, Tanzania, and
Uganda-during 1996-1999. The studies examined the adoption of improved wheat and
maize varieties, as well as adoption of chemical fertilizers.
This paper synthesizes and analyzes the study results.1 It is organized around four key
questions: To what extent have farmers adopted improved technologies for maize and wheat
in Eastern Africa? Which farmers are using the improved technologies? What are the main
obstacles to farmers adopting improved technologies? Are improved seeds and fertilizer
available? The final section discusses policy implications and offers recommendations.
Eastern African Adoption Studies
Table 1 provides a brief description of the micro-studies. All 22 studies involved collecting
farm level data from households in survey areas. The data were representative of those
collected for most adoption studies-farmers' characteristics, and descriptions of farms and
technology use-and followed CIMMYT's manual for survey design (CIMMYT 1993).
Researchers from the national agricultural research systems (henceforth: "national systems")
in each country chose the study regions, which typically represented primary maize and
wheat growing areas. The sample size for each study was fairly small, ranging from 36-353.
In addition to recording descriptive data, most studies ran simple econometric models to
analyze determinants of adoption of improved varieties and/or fertilizer.
1 The research report for each of these studies is listed in Appendix 1.
Table 1. Description of adoption studies, Ethiopia, Kenya, Tanzania, Uganda, 1996-1999.
Improved Market Agricultural Population/ Farm
Country/region Year Crop varietiest Fertilizert accesstt potentialttt km2t size/haP
Bale Highlands 1997 Wheat Yes Yes Medium High 60 3-4
Central Highlands 1997 Wheat Yes Yes Medium High 102-171 2-3
Chilalo Awraja Wheat Yes No High High 4.0
Enebssie 1997 Wheat Yes No Medium High 131
Northwestern Province 1999 Wheat Yes Yes Medium High 91 5-7
Sidamo and North
Omo Zone 1997 Maize Yes Yes Low Medium 127-229 1.0
Western Oromia 1996 Maize Yes No Medium High 3-4
Wolmera Woreda Wheat Yes No High High 2-3
Coastal Lowlands 1998 Maize Yes Yes High Low 60-114 4.30
Embu District 1998 Maize Yes Yes High Low 456 1.68
Kakamega and Vihiga
Districts 1996 Maize Yes Yes High High 433-866 1.2
Kiambu District 1996 Maize No Yes High High 373 4-5
Narok, Nakuru, and
Uasin Gishu Districts 1997 Wheat Yes No High Low 24-164- 187 100-780
Central 1995 Maize Yes Yes Low Low 8-25
Eastern 1995 Maize Yes Yes Low Low 35 7-10
Lake Zone 1995 Maize Yes Yes Medium Medium 49.4
Northern 1995 Maize Yes Yes High High 2-4
Southern 1995 Maize Yes Yes Low Medium 19.3 2.6
Southern Highlands 1995 Maize Yes Yes Medium High 16.4 3.0
Western 1995 Maize Yes Yes Low Medium 10-11
(S. Highlands) 1997 Wheat Yes Yes Medium High 2-3
Iganga District 1995 Maize Yes No High 196 2.5
t Indicates whether the study analyzed the adoption of this technology.
tt Low access: seasonal markets are accessible but farmers have to travel at least one full day to get to large
markets (>50,000 people) and the nearest tarmac roads are about 6 hours away by foot. Medium access:
large markets are accessible and the nearest tarmac roads are about 1-2 hours on foot. Farmers have
access to large markets (>50,000 people). High access: very large markets are accessible (>100,000 people)
and tarmac road are less than 1 hour by foot.
ttt Low: Unreliable unimodal rainfall, <600 mm per year, poor soils. Medium: 600-1,200 mm rainfall/yr, unimodal
or bimodal. High: reliable bimodal rainfall distribution, >1,200 mm/yr. Soils are rich volcanic or alluvial.
t Population density is for the entire region. If a range is listed, it is for each of the two smaller units of area.
t Farm size is based on average farm size of sampled households.
It is important to note that the areas covered by these studies are not necessarily
representative of Eastern Africa overall: they cover a wide range of areas, especially in
Ethiopia and Tanzania, but most of the samples were purposively selected in major areas
for production of the particular crop. Thus, the levels of adoption presented here are not
representative of national adoption levels. Nor can adoption levels be directly compared
across sites, since the definition of "adoption of improved varieties" varies. Nonetheless,
the data show interesting patterns of adoption (Table 2).
To What Extent Have Farmers Adopted Improved
Technologies for Maize and Wheat in Eastern Africa?
Maize and wheat are the two chief staple crops in Eastern Africa. In Ethiopia, maize is the
most important crop in terms of production and yield, although teff is more widely
cultivated. Maize is the most important crop in all three respects in the zones in Ethiopia
where adoption of maize technologies were examined. Maize is a major crop cereal in
Tanzania, providing 60% of dietary calories (FSD 1992; 1996). In some areas of Tanzania,
maize is also an important cash crop, competing with cotton for land and labor. In Kenya,
maize is a major staple and the main source of income and employment for most
households. More than 70% of maize area in Kenya is cultivated by smallholder farmers
(<20 ha; CBS 1990). In Uganda, maize is a major staple crop and has the potential to become
a major export crop.
Overall, wheat is less important than maize in Eastern Africa but is important in some
areas. In Ethiopia, wheat ranks fourth in total crop area and production. Ethiopia continues
to be a net importer of wheat. Wheat is the second most important crop in Kenya and is the
most important crop in some areas. Wheat is of minor importance in Tanzania (Ekboir 2002)
The data suggest that improved varieties of maize have been widely adopted in many
maize-producing regions of Eastern Africa.
Ethiopia shows the lowest levels of adoption-0% to 56%-in the areas studied in 1996.
However, adoption has increased dramatically since 1992 (when almost no farmers were
growing improved maize varieties) due to the introduction of a new extension system
supported by the Sasakawa Foundation of Japan.
Rates of adoption of improved maize varieties were relatively high in the three regions in
Kenya and one region in Uganda examined. In Kakamega and Vihiga Districts, Western
Kenya, 50% of farmers used certified maize seed in the first season, although in the second
season all farmers grew only local varieties. In Embu District, 65% of the farmers surveyed
sowed certified seed for two consecutive seasons. However, only 30% of farmers on the
Table 2. Percentage of sampled farmers adopting improved technologies, Ethiopia, Kenya, Tanzania,
Seed adopters Fertilizer adopters
Country/region (%) (%) Definition of adoption
Adaba 34 83 Recently released (<10 yrs)
Ada MHHt 12 wheat 0 maize Any improved material
Ada-FHH 5 wheat 0 maize
Lume-MHH 39 wheat 13 maize
Lume-FHH 22 wheat 11 maize
Gimbichu-MHH 3 wheat 0 maize
Gimbuchu-FHH 0 wheat 0 maize
Chilalo Arwaja: 80% of harvested area Variety released or introduced
into study area between 1990-95.
Intermediate 32 70 Any improved material
High 36 27
Northwestern Province: 72 60 Any improved material
Sidama and North Omo Zone: Any improved material
Lowland 22 58
Intermediate 25 70
Chaliya 46 78 Any improved material and
Bako-Tibe 49 97 fertilizer for maize production
Bila-Sayo 56 88
Sibu Sire 39 79
Coastal Lowlands: 30 4.5 Certified purchased
seed on at least 1 acre
Embu District: 65 98
Kakamega and Vihiga Districts: 51 35 Certified purchased
seed on at least 1 acre
Kiambu District: 74 Any chemical fertilizer
Narok, Nakuru, and
Uasin Gishu Districts: Variety released in past 10 years
High potential 52
Low potential 22
Lowlands 78 across all 17 Any improved material
Intermediate 3 77
Lowland 85 17 Any improved material
Intermediate 95 8
t MHH = male-headed household. FHH = female-headed household.
Table 2. (cont'd)
Seed adopters Fertilizer adopters
Country/region (%) (%) Definition of adoption
Low rain 45 50 Any improved material
Intermediate rain 62 48
High rain 100 100
Lowland 89 64 Any improved material
Intermediate 92 44
Southern: 3 Any improved material
Intermediate 64 65 Any improved material
Highlands 44 79
High 55 66 Any improved material
Low 93 60
Mbeya District (S. Highlands): 79 40 Any improved material
Inganga District: 43 3 Recommended variety
(grown on own initiative)
Kenyan coast grew improved maize seed. In the Iganga District in Uganda, 43% of the
sample grew the recommended improved variety Longe-1, an open-pollinated variety
(OPV) released in 1991. Most Ugandan farmers surveyed who were not growing this variety
were growing Kawanda Composite A, a variety released in 1971. Composite A continues to
be grown despite the lack of maintenance breeding and the collapse of the seed
multiplication system in the early 1980s.
In Tanzania, most farmers surveyed were using improved materials. Two-thirds of the study
regions reported at least a 75% adoption rate. Most farmers used recycled hybrid seed. Only
in the Southern Highlands, a relatively high potential area, did most farmers report that they
purchased new seed each year. The proportions are much lower in other zones. In the
Central Zone, farmers in the lowlands reported that they recycled seed for 5-8 years; farmers
in intermediate altitude and highland zones said they recycled seed for 8-10 years.
The number of farmers who purchased hybrid maize seed was very low. We do not know
how frequently farmers replenish their seed in Uganda, where recommended varieties are
OPVs rather than hybrids. Regardless of problems in interpreting or comparing data, it is
noteworthy that farmers in all regions surveyed are using some improved materials.
More sampled farmers were using improved varieties of wheat than was true for maize
in Ethiopia. The figures range from 32% in Enebssie region to over 70% in the
Northwestern region. Adoption of improved wheat varieties was over 90% in 1992 in the
Central Highlands but was lower during 1997. Of those growing wheat, 70% of male-
headed households and 86% of female-headed households grew local varieties. One
explanation for the decline is that farmers who recycled improved seed considered it a
local variety, once it had been recycled. This illustrates the difficulty of analyzing cross-
country data when the definitions of terms are not harmonized beforehand. The report on
Chilalo Awraja2 noted that, although the majority of farmers grew improved wheat, only
2% used newly released varieties. The varietal replacement rate is very low in Ethiopia.
For sites where it was calculated, the weighted average age of varieties was 11-13 years.
This contrasts with averages of less than 4 years in the Yaqui Valley in Mexico to over 10
years in the Punjab of Pakistan, with a global average of 7 years. Still, it suggests a
willingness to use new varieties and to adopt new technologies.
In the Mbeya District (Southern Highlands), the only wheat growing area studied in
Tanzania, about 79% of the farmers sampled grew improved wheat varieties. The most
commonly adopted variety was Juhudi, which was released in 1987. All five varieties
recommended for the Southern Highlands were released in the 1980s.
In the only wheat growing are studied in Kenya-the Narok, Nukuru, and Uasin Gishu
Districts-28% of the farmers sampled used "new" varieties, defined as varieties released
within the previous 10 years.
Thus, a relatively large proportion of farmers, especially in high potential zones, have
adopted improved varieties. However, these may be relatively old varieties and may
have been recycled for many years.
The proportion of farmers using fertilizer, specifically chemical fertilizer, varies
tremendously across Eastern Africa. Fertilizer was more likely to be used in high
Not all the Ethiopian studies looked at fertilizer use, especially in areas where fertilizer
use was very low. Yet, high levels of adoption were found in areas where fertilizer
adoption was studied. Fertilizer adoption rates were higher in wheat farming areas than
among maize farmers. In the Bale Highlands, 95% of adopters of improved wheat and
75% of non-adopters used fertilizer. High levels of adoption of fertilizer-over 58%-
were reported among the maize farmers surveyed in Ethiopia.
In Kenya, most farmers in high potential areas used inorganic fertilizer: 98% in Embu and
74% in Kiambu. This was substantially less in low potential areas: 35% in Kakamega and
2 An "Awraja" is a sub-region in Ethiopia, and a "Woreda" is a district.
Vihiga Districts and only 4.5% on the Coast (Table 2). In Kiambu, 35% of farmers used
both organic and inorganic fertilizer and 29% used only inorganic fertilizer.
In the highest potential area in Tanzania, the Southern Highlands, 65-79% of farmers used
fertilizer.3 In the Southern Zone, one of the lower potential areas of Tanzania, only 3% of
farmers used inorganic fertilizer.
Fertilizer use was very low (3%) in the Iganga District in Uganda.
Thus, the range of fertilizer use is very wide. The proportion of sampled farmers using
fertilizer was higher in Ethiopia than in Tanzania. Clearly, there are complementarities
between fertilizer use and improved varieties. Although we typically expect that farmers
adopt new varieties of seed first and then fertilizer, this was not the case in all areas. In
Ethiopia especially, we saw areas where fertilizer adoption outpaced adoption of
Farmers have adopted improved varieties of wheat and maize and fertilizer, although it is
important to be careful in interpreting the numbers, given that the definitions of adoption
vary across sites and sites were not selected representatively for particular areas. Even
with the limited data, we can draw a couple of conclusions.
First, farmers did not appear to be resistant to using improved varieties of wheat and
maize. There did not seem to be strong cultural views against using these improved
varieties. In a later section, we discuss which farmers adopted improved technologies and
some of the reasons why some farmers did not adopt them. Similarly, farmers appeared to
be willing to use fertilizer.
Second, although many farmers were using improved seed, much of the improved seed
used was recycled and came from old varieties, especially in Ethiopia and Tanzania. Thus,
not all of the benefits of hybrid maize were being realized. A recent survey of literature on
recycled maize seed use concludes that "...while advanced-generation hybrids may not
perform as well as crops grown from F1 seed, in many cases they significantly outperform
the variety that the farmer was growing previously" (Morris et al. 1999). This suggests
that farmers obtain some but not all agronomic benefits from improved varieties. Using
newly purchased seed would presumably increase output, but would also increase costs.
Many questions remain about the extent of adoption. We do not know how representative
surveyed areas are. The surveyed areas were chosen because they were in the crop
producing areas. Adoption rates were relatively high in many of these areas; researchers
expected to find these technologies in use. We would expect that farmers in more remote
and more marginal areas would be less likely to use improved technologies. Adoption
studies could be improved by standardizing definitions across studies (or providing
3 The higher number is for the highlands areas, while the lower number is for the intermediate altitude zone.
information using more than one definition) and by using sampling techniques that allow
results to be generalized. Despite their limitations, these studies indicate that even in higher
potential regions with relatively high levels of adoption, there is still considerable scope to
improve the productivity of smallholder agriculture in surveyed areas.
Which Farmers Use Improved Technologies?
Each micro-level adoption study provides descriptive data on farmer characteristics and
most studies estimated the probabilities of a farmer adopting a technology. Like most
adoption studies, these studies focused on a cross-section of the population and compared
adopters to non-adopters. Because of this, it was not possible to glean anything from them
about the characteristics of farmers at the time of adoption. Although regression results
such as these are often interpreted as representing the probability that a farmer will adopt
the technology, they are more appropriately interpreted as the probability that a farmer is
using the technology. In other words, the information that we are using is current
information on the farmer, not information on the farmer at the time of adoption.
Interpreted in this manner, the estimations presented in these micro-studies do provide
some information on the characteristics of farmers who were using the technology at the
time of the studies. Summaries of the econometric findings are presented in Tables 3 and 4.4
Factors that affected technology use fell into three general categories: attributes associated
with farmers and farms, characteristics of the technology, and objectives of the farmer.
Institutional and policy factors that may have affected the use of technologies were rarely
included in the analyses, usually because there was little or no variation across sampled
households except for use of credit and extension services. Some local level institutional
characteristics, including access to markets, were included only through a dummy variable
indicating the district or ecological zone. Most of the econometric analyses focused on the
effects of farmer and farm characteristics.
The CIMMYT/national system studies estimated the use of each technology separately,
although some included a measure of the probability of using improved varieties in the
estimation of the probability of using fertilizer.
The first technology considered was the use of improved maize and wheat varieties. Some
of the estimations simply examined whether the farmer used improved varieties (again, the
definitions of improved varieties vary across studies), while others examined the
proportion of land a farmer planted with improved varieties. The second technology
considered was the use of fertilizer. Again, some farmers were simply asked whether they
used any fertilizer, whereas others provide an actual figure for fertilizer application by area.
4 Full details on estimations are available in individual reports (Appendix 1). Tables 3 and 4 list the variables that
were included in each estimation and indicate which ones were significant. Since the definitions of variables and
the units used vary across studies, these summary tables do not attempt to indicate the size of the effect.
Table 3. Summary of estimations of adoption of improved wheat and maize, Ethiopia, Kenya, Tanzania,
Uganda, 1996-1999. t
Bale Central Northwestern Sidama and Western Wolmera
Highlands Highlands Enebssie Province North Omo Oromia Woreda
(wheat) (wheat) (wheat) (wheat) Zone (maize) (hybrid maize) (wheat)
Wheat area Livestock Zone Chemical Fert. Age Experience Experience
Age Coop Credit Farm size Education Education Extension
Education member Cultivated area Extension Family size Extension Education
Extension Household Extension Participation Farm size Field day Radio
Family size size Producer in demo TLU Farm size Family
Hired labor Extension cooperative Attend Off-farm Family size size
Credit Education Oxen agricultural income Hired labor Farm size
Livestock Km to Experience course Hired labor Livestock Zone
Disease market Education Credit Zone Off-farm
resistance Farm size Labor Illiterate Extension income
Bread quality Age Off-farm Elementary Credit High yield
Lodging resist. income Junior High Organization Lodging
Contact farmer Impurity
Hand hoe Seed
Ox plow condition
Embu and Vihaga Nakuru
Chilalo Coastal District Districts and Uasin
Awraja lowlands (maize) (maize) (maize) Gishun Districts
Age Age Mijikenda Age Age Seed source
Age squared Education tribe Education Primary District
Near AADE Permanent Trees Use credit Secondary Household size
farm employment Cattle Extension Cattle Seed selection
Near ESE Use Credit On-farm Organization Use credit Wheat price
farm Extension income member Extension Age of head
Near Organization Tractors Hires labor Organization Education
Res. stn. member Sells maize Male member Zone
Literacy Hires labor Maize acres Farm size Hires labor Farm Size
Campaign Female-headed Farmer training Use Farm size Seed
Primary ed. household course fertilizer Cash crop area retention
Second. ed. Farm size Listens to Coffee area Uses manure Years farming
Hosted Off-farm agricultural Zone Zone wheat
demo plot income programs
Extension District Organization
Farm size member
t Variables listed are those included in the econometric estimations. Variables in bold are statistically significant
at the .05 level or higher.
Table 3. (cont'd)
Lake District Southern
Central Eastern Zone (S. Highlands Northern Highlands Western
(maize) (maize) (maize) wheat) (maize) (maize) (maize)
Experience Experience Experience Age Farm size Zone Experience
Labor Labor Education Education Experience Farm size Labor
Education Education Extension Extension Education Hand hoe Education
Wealth Wealth Farm size Farm size Livestock units Ox-plow Wealth
index index Family Family size Family labor Extension Extension
Extension Extension labor Hired labor Hand hoe Experience Varieties
Zones Zones Hired labor Livestock Ox-plow Livestock units Zone
Varieties Varieties Livestock Off-farm Tractor Labor
Hand hoe income Nitrogen- Hired labor
Credit fertilizer Credit
Table 4. Summary of estimations of adoption of fertilizer, Ethiopia, Kenya, Tanzania, 1996-1999.t
Northwestern Sidama and North
Bale Highlands Province Omo Zone (maize)
Wheat area Farm size Age
Age TLU Education
Education Participation in demo Family size
Extension Field day Farm size
Family size Agricultural training TLU
Hired labor Credit Radio Off-farm income
Livestock Coop member Hired labor
Gender Credit Zone
Coastal Lowlands Kakamega and Vihiga
Coastal Lowlands (continued) Districts Kiambu District
Age Maize Age Age
Female-headed houseld Acreage Primary Extension
Permanet Extension Secondary Organization member
employment income Attend course Cattle Farm size
Education Listens to agricultural program Use credit Household size
Mijikenda tribe member Credit Extension Hired labor
District Organization member Organization member Livestock
Farm size Hires labor Off-farm income
Trees Farm size
Cattle Crash crop area
Hires labor Uses manure
On-farm income Zone
Central Eastern Lake Zone Northern
Experience Experience Zones Farm size
Labor Labor Farm size Experience
Education Education Hand hoe Education
Wealth index Sealth index Ox-plow Livestock units
Extension Extension Extension Family labor
Zones Zones Experience Hand hoe
Varieties Varieties Livestock Ox-plow
Hired labor Nitrogen- fertilizer
t Variables listed are those included in the econometric estimations. Variables in bold are statistically significant
at the .05 level or higher.
Table 4. (cont'd)
Southern Highlands Western District (S. Highlands)
Zone Experience Age
Farm size Labor Education
Hand hoe Education Extension
Ox-plow Wealth Farm size
Extension Extension Family size
Experience Varieties Hired labor
Livestock units Zone Livestock
Hired labor income
Farmer characteristics that might be associated with the use of improved technologies
include age or experience, education, wealth (including land), availability of cash or credit
to purchase inputs, access to information, and access to labor.
The CIMMYT/national system studies hypothesized that experienced farmers were more
likely to use improved technologies. We might also expect that younger, less experienced
farmers are less set in their ways and are thus more likely to try improved technologies.
The age of the farmer was often used as a measure of the farmer's experience, although
sometimes the number of years the individual had been farming was used. None of these
variables were statistically significant in any of the analyses on the use of improved maize
varieties. Age (or experience) was positively associated with the use of improved wheat
varieties in the Bale Highlands, Enebssie, and Chilalo Awraja in Ethiopia.
Years of farming experience related positively to fertilizer use in only the southern
highlands of Tanzania. In Kiambu, Kenya, age was negatively related to inorganic fertilizer
use but was not significant in determining combined use of inorganic and organic
fertilizer. On the coast of Kenya, age was negatively associated and education positively
associated with the fertilizer use.
Several measures of education were used. In the Tanzanian studies, it was the number of
years of education of the household head. Only in the Lake Zone was this correlated with
the use of improved varieties. A dummy variable was used in the Iganga District in
Uganda to indicate whether the household head was literate but this was not statistically
significantly related to the use of improved maize. In Ethiopia, the education variable was
positive and significant only for the use of improved maize in Western Oromia.
At least one measure of wealth was included in most estimations of technology use. It is
not obvious a priori what the relationship is between wealth and use of improved varieties,
although it is often assumed that wealthier farmers have greater access to inputs,
especially purchased seeds and fertilizer.5 But many researchers have argued that seed
technologies are scale neutral and thus available to farmers regardless of the size of their
farms or levels of wealth. We might not expect to find much of a relationship between
wealth and use of improved varieties, particularly in studies where the definition of an
adopter of improved varieties does not necessarily indicate that the farmer purchased new
seed. However, we might expect to see a stronger positive relationship with fertilizer use,
given that fertilizer must be purchased. To the extent that the adoption of a technology
increases a farmer's wealth, we might expect to see a positive relationship between current
use and wealth. We might also expect to see a relationship between farmers' wealth and
their willingness to try a new and unproven technology. Wealthier farmers may be more
willing and able to take risks. Wealth may also be an indicator of a farmer's access to
credit. For these reasons, we might expect a positive relationship between wealth and use
of improved technologies.
A number of factors were used as a proxy for wealth. They vary, in part, because different
measures are appropriate in different contexts. Some wealth indicators used in the
analyses represent more than just measures of wealth and may be related to technology
use in other ways.
One of the key measures of wealth used in the studies was farm size. In the maize
adoption studies, farm size was only correlated with the use of improved varieties in the
Lake Zone in Tanzania. A relationship between farm size and the use of improved variety
was more frequent in wheat studies in Ethiopia-in two of the six wheat studies
(Northwestern Province and Enebssie Region) a positive correlation was found. In areas
where land is more abundant, farmers may increase productivity both by using fertilizer
and by increasing farm size. In areas such as the Southern Highlands, where land is
relatively scarce, farmers may increase productivity by using fertilizer. More intensive
agriculture on smaller plots may be the appropriate strategy under these circumstances.
Farm size was positively related to the use of fertilizer among maize farmers on the
Kenyan Coast and wheat farmers in Mbeya District (Southern Highlands), Tanzania, but
was negatively related to fertilizer use on maize in Tanzania's Southern Highlands.
Another frequently used measure of wealth is the number of livestock. Livestock
ownership was positively related to the use of improved maize in the Southern Highlands
and Lake Zone of Tanzania, and in Kakamega and Vihiga Districts, Kenya. Different
definitions of livestock were used. In Kenya, only the number of cattle was included. In
several Ethiopian studies, livestock were aggregated using tropical livestock units.6 In the
wheat studies, livestock were correlated only with use of improved varieties in the
Enebssie region of Ethiopia, where the number of oxen was the livestock measure.
Livestock may be a measure of wealth, but in the case of oxen particularly, they may also
be used as an input in the production process, allowing for greater area to be cultivated.
Animal manure may also substitute for purchased fertilizer.
5 Many of the CIMMYT/national research system studies make this assumption.
6 They are usually aggregated as follows: Oxen and cows 1; goats and sheep = 0.8; poultry = 0.2.
Other wealth measures included assets that may be owned by farmers, either entered
individually into the estimation or, as was done in three Tanzanian maize studies (Central,
Eastern, and Western), aggregated into a wealth index. This wealth index combined the
average number of livestock units, farm implements, and cultivated land for the past three
years-each divided by the sample mean for the item. When assets were combined, the
assumption was that they impacted similarly on decisions about technology use. When
the assets were included separately in the equations, the assumption was that various
forms of wealth affected input decisions differently. Since many forms of wealth are
related to agricultural production, including land, tools, and livestock, each may have
other effects than simply the wealth effect. Few of these variables were significant in
explaining the use of improved technologies.
To the extent that agricultural inputs must be purchased, we would expect to find a
relationship between the availability of cash or credit to a farmer and the use of improved
technologies. One source of cash is off-farm income. Current income is not appropriate to
examine initial adoption decisions because successful adoption of agricultural
technologies may change farming households' incentives to supply off-farm labor. We
may, however, be interested in the correlation between current off-farm income and
current technology use, because off-farm income may be important for farmers who wish
to purchase improved inputs. Off-farm income was correlated with the use of improved
maize in Oromia, Ethiopia, and was positively related with both improved seed and
fertilizer use in the Kenyan Coast. Off-farm income was not correlated to income in other
studies that examined it. No measures of remittances were used in any of the studies. We
might expect that remittances from family members who worked in the city would be
used to finance agricultural inputs.7
For similar reasons, use of credit may also be correlated with the use of improved inputs.
However, credit is only statistically significant in the estimation for two sites in Ethiopia-
North Omo and Sidamo for maize and Bale Highlands for wheat-and for the Kenyan
Coast. Credit is not available in many areas. Where it is available, the use of credit is often
highly correlated with wealth or farm size. Although many of the studies referred to
access to credit, the variable that is used is whether the farmer actually obtained credit.
Extension is the one variable that was statistically significant in many of the estimations.
The definition of extension varies-it may include simply whether or not the farmer had
any extension contact or it may be based on the number and frequency of visits. At most
study sites, farmers who grew improved varieties were more likely to have extension
contact than farmers who were not. Many extension offices also provided inputs, which
increased the correlation between extension contact and use of improved technologies.
Thus, extension effort is more than simply increasing knowledge. The provision of
extension services may also be correlated with infrastructure and market access.
7 For example, Francis and Hoddinott (1993) find this to be the case in Kenya.
When other information variables were included in the analysis, they were usually
significant. For example, variables that were significant in explaining the use of improved
seed included attending a field day in Western Oromia (for maize hybrids but not OPVs),
participating in demonstrations in Northwestern Ethiopia, and being a contact farmer (but
not hosting demonstration plots) in Chilalo Awraja. In addition, being near the Arsi
Agricultural Development Enterprise, the Ethiopian Seed Enterprise, or the research station
in Chilalo Awraja meant that a farmer was more likely to use improved wheat varieties. In
Kenya, access to extension increased the likelihood of using improved seed in Embu, while
on the Coast factors such as listening to extension programs on the radio, being a member
of a group, or participating in training courses were also significant. Membership in an
organization in Ethiopia and at the Kenyan Coast-usually a producer cooperative-was
often associated with use of improved varieties. This may be due to both the information
and access to resources that the organization provides. In addition, to the extent that
farmers choose to participate in these organizations, we would expect a greater likelihood
of their being adopters and innovators.
A final factor that may be correlated with the use of improved varieties is the availability of
labor. Small-scale farms obtain labor from household members or by hiring labor. One
measure used frequently in adoption studies to account for labor availability is household
size. This may be the total number of individuals, the number of adults, or the number of
adult equivalents (calculated by counting children as a fraction of an adult). Household
composition may also be important in determining household labor availability, but no
measure of this was included in any of the 22 studies.
We did not find a clear relationship between household size and use of improved varieties.
Where household size is a significant explanatory factor, it is sometimes positively and
other times negatively related to use of improved technology. When land and labor markets
exist, we would not necessarily expect to find any relationship between technology use and
The use of hired labor was related to the use of improved wheat varieties in Tanzania. In
Uganda and Kenya, the use of hired labor was related to the use of improved maize
varieties. Again, the use of hired labor is probably also correlated with the wealth of the
farmer and the size of the farm.
None of the studies explicitly included variables to account for the agricultural potential of
the land at the village or farm levels. Several included a dummy variable identifying an
agricultural zone. In Ethiopia, farmers in intermediate areas in Sidamo and North Omo
were more likely to allocate land to improved maize than farmers in the lowland areas. No
zone effect was found for fertilizer use. In the Enebssie region, farmers in highland areas
were more aware of improved varieties than farmers in intermediate areas, but this did not
affect their use of the technologies. The location of Wolmera Goro, a peasant association in
the Wolmera Woreda, was related to an increased probability of using improved bread
wheat varieties, but the authors assert that this is because it is closer to the research center
and thus has greater access to inputs, rather than due to a greater agricultural potential.
In Tanzania, four studies included measures of location. The location variable was
significant in determining the amount of land allocated to improved varieties except in the
Central Zone. In the Eastern Zone, lowland farmers were less likely to allocate land to
improved maize varieties than farmers in intermediate or highland areas. In the Southern
Highlands, farmers in intermediate areas were less likely to allocate land to improved
varieties than farmers in the highlands. In the Western Zone, farmers in low rainfall areas
were less likely to allocate land to improved maize. In general, farmers in the highlands
were more likely to use fertilizer than those in the lowlands.
It is worth noting that most of the studies failed to include any measure of farmers' gender.
The analyses of the Central Highlands of Ethiopia and the Kenyan Coast did disaggregate
the data, especially adoption data, by gender of the household head. Male-headed
households were found to be more likely to use improved wheat varieties than female-
headed households. In Tanzania, researchers claimed that there are few female-headed
households. The assumption is implicit in all studies that the male household head is the
primary farmer and decision-maker. This may not be true if the male head of household is
a migrant within the country, leaving his wife to manage the farm for much of the year.
There is increasing evidence that many economic decisions made within households are
dependent upon the characteristics of both men and women members, and that it is not
sufficient simply to model farming decisions made solely by male heads of household.8
The characteristics of the available technology also influences farmers' use of the
technology. Again the results presented in this section tell us less about whether or not
farmers will adopt new technologies than about which technologies farmers are using.
Only a few of the studies included measures of technology characteristics in the
Three of the Tanzanian studies-Western, Central, and Eastern-controlled for
technologies' characteristics by including dummy variables for whether the farmer grew
early- (3 months), intermediate- (3.5-4.0 months) or late-maturing (4.5-5.0 months)
varieties. The dependent variable in the estimation was the proportion of land allocated to
improved varieties. In all three regions, more land was allocated to improved varieties
when early-maturing varieties were grown. In Eastern Tanzania, this was also true when
intermediate-maturing varieties were grown. Thus, less land was allocated to improved
varieties when late-maturing varieties were grown. Only in Eastern Tanzania was maturity
type identified as significant in explaining the use of fertilizer.
In Western Oromia, Ethiopia, variables indicating farmers' preferences were included in
estimations. It was assumed that farmers who were concerned about high yield, resistance
to lodging, clean seed, and conditioned seed would allocate more land to improved
varieties. The estimations found that this was the case when looking at the amount of land
allocated to hybrid maize, but not for improved OPVs.
8 See Doss (1999) for a detailed discussion of women and agricultural technology in Africa.
Wheat farmers in the Bale Highlands were asked about their preferences on three types of
traits-disease resistance, bread baking qualities, and lodging resistance. Farmers' preferences
for each positively influenced the amount of land allocated to improved varieties.
We would certainly expect that farmers' objectives influence decisions on whether to use
improved technologies. However, it is hard to sort out some of the relationships. Farmer
objectives may also change as farmers become familiar with improved technologies and see
options that not previously available.
A measure of the area planted to cash crops was included in the Kakamega and Vihiga
District study in Kenya, but was not significantly correlated with use of improved inputs. In
the Bale Highlands, Ethiopia, a measure of the area under wheat was included. This may
capture the extent to which wheat is an important crop for the farm, but because no other
measure of farm size was included, it may simply be capturing the effect of farm size. In the
Central Highlands of Ethiopia, distance to the market was included as a measure. This
provides a proxy for how easy it would be for the farmer to produce for the market. This
variable, however, was not statistically significant in explaining the use of improved
Extension is clearly the variable most highly correlated with the use of improved
technologies. It is not always clear, however, what the extension variable is actually capturing.
It may be related to the provision of both inputs and information. The extent of extension
services may also be picking up infrastructure issues: farmers in more accessible, less remote
areas may receive more frequent extension visits. This provides further evidence of the need
for great care in interpreting variables.9
What are the Main Obstacles to Farmers
Adopting Improved Technologies?
Farmers cited several reasons for not adopting improved technologies. The first was simply
being unaware of the technologies or that they could provide benefits; this may included
misconceptions about the related costs and benefits. The second reason was that the
technologies were not profitable, given the complex sets of decisions that farmers make about
how to allocate land and labor across agricultural and non-agricultural activities. This may be
due to the fact that appropriate varieties for farmers' agroecological conditions were not
available or that farmers preferred characteristics found only in local varieties. It may also be
due to institutional factors, such as the policy environment, which affect the availability of
inputs (land, labor, seeds, fertilizer) and markets for credit and outputs. These institutional
factors also affect input prices. It may also be that use of improved technologies may increase
9 For further discussion on the interpretation of variables, see Doss 2003.
production risks: if crops fail, the financial losses will be higher. Finally, in some instances,
technologies were not adopted because they were simply not available. The availability of
improved seeds and fertilizer will be discussed in the following section.
In a number of studies, farmers (both adopters and non-adopters) were asked to identify
constraints to the adoption of improved technologies. Most of the Tanzanian studies did
not list farmers' responses on this point, so the following section uses only data from the
other three countries. Several Tanzanian studies, however, did report on farmers'
explanations regarding constraints to fertilizer use.
Availability of Information
The first issue is whether lack of information is a constraint to the adoption of new
technology. Overall farmers did not identify "lack of information" as a key constraint to
adopting improved varieties or fertilizer. Only in a few areas, such as the Kenyan Coast,
did farmers mention the lack of information as an important constraint. However,
additional information on extension services and farmers' awareness of extension
recommendations suggests that many farmers were not aware of recommendations or
technologies. If a farmer does not know about a technology, he or she will not say that the
lack of information is the constraint! For example, the study on Chilalo Awraja, Ethiopia,
reports that only 8% of the farmers interviewed were able to identify or had information
about new varieties.
Farmers were asked to list constraints to adoption of improved technologies in their areas.
For improved maize varieties, 28% of non-adopters in the Uganda sample said that lack of
information was a constraint. At the Kenyan Coast, the proportion was 14%. In Ethiopia,
lack of information was mentioned as a constraint by 12% of lowland and 2% of
intermediate zone farmers in Sidamo and North Omo, and 5% of adopters and 20% of non-
adopters in the Bale Highlands. In Mbeya District (Southern Highlands), Tanzania, 21% of
non-adopters of improved wheat varieties said that lack of information was a constraint.
The range of opinion among farmers was wider for fertilizer. In the Uganda study, 39% of
adopters of improved varieties and 64% of non-adopters reported lack of information as a
constraint to fertilizer use. In two woredas in Western Oromia, Ethiopia, no one reported
that lack of information was a constraint, but in the other two woredas, 13% and 50% of
the farmers identified it as a constraint. In Sidamo and North Omo, Ethiopia, information
was mentioned as a constraint by only 2-3% of sampled farmers. All areas that reported
lack of information as a constraint to fertilizer use were maize farming zones.
A second way to look at this issue is to examine where farmers obtain their information.
Table 5 presents data on the availability and sources of information on improved
technologies. The Tanzanian studies asked farmers whether they received information on
improved varieties and fertilizer and, if so, where they received it. Extension was clearly
one of the key sources of information, except in the lowlands of the Northern Zone.
These results should not necessarily be viewed as saying that extension is effectively
getting the message out about improved technologies. In studies broken down by adopters
and non-adopters, more adopters received extension services. It may be surprising,
however, to note the percentage of non-adopters who received extension services: 77% in
the Bale Highlands, Ethiopia; 31% in Kakamega and Vihiga Districts, Kenya; 32% on the
Kenyan Coast; 35% in Uganda; and 57% in Mbeya District (Southern Highlands), Tanzania.
Instead, these results should be interpreted as indicating that farmers do not generally view
lack of information as an important constraint. Many farmers who adopted improved
technologies did not follow recommended practices. It is not at all obvious from the data
whether this is because they misunderstood recommendations, because other constraints
prevented them from fully implementing recommended practices, or because the
recommendations were inappropriate for their situation. In addition, farmers may not have
realized that they lacked the information and thus may not have listed this as an important
Several studies reported additional information about extension services. In the Central
Highlands of Ethiopia, information on extension was reported according to the gender of
the household head. Similar numbers of male- and female-headed households were taught
to use fertilizer in two of the three woredas surveyed (Ada and Lume), whereas the
differences were more pronounced for the use of improved seeds. Considerably more
farmers received information from extension services about fertilizer than about improved
seeds, regardless of the gender of the household head.
Of the farmers in Kiambu, Kenya, 80% who used manure as fertilizer, 59% who used only
inorganic fertilizer, and 40% who used both said there were various problems with
extension services, including infrequent visits, unavailability, and unclear messages.
Farmers' perceptions of the effectiveness of extension services were not reported in other
The results suggest a continued, important role for extension services to inform farmers
about new varieties and how to manage them.
The second set of reasons for non-use of improved technologies was that they were not
profitable, given farmers' constraints. Profit for farmers is the value of the output minus the
cost of production, including the opportunity cost of their time. Farmers rarely talk about
their decisions in these terms. Instead, they focus on one aspect of the equation-the price
of inputs. Most surveys, including the CIMMYT/national system studies from Eastern
Africa, do not ask farmers specifically about profitability, although some contain sufficient
data to allow researchers to calculate profitability.
The high price of improved seed was frequently listed as a constraint to adoption. Among
maize farmers in Ethiopia, 62% of lowland farmers and 45% of intermediate altitude
farmers in Sidamo and North Omo reported price as a constraint. Even adopters of
Table 5. Availability and sources of information on improved technologies, Ethiopia, Kenya, Tanzania,
Country/region information -varieties information -fertilizer
ETHIOPIA Extension any Contact 1-2 times/month >2 times/month
Adopters 92 38 25
Non-adopters 77 30 12
Ada-MHH 13 38
Ada-FHH 11 37
Lume-MHH 51 100
Lume-FHH 17 100
Gimbichu-MHH 30 88
Gimbuchu-FHH 15 52
Wolmera Goro 78 20
Robe Gebeya 68 15
KENYA Any extension visit
Kakamega and Vihiga Districts
Kiambu District: 47
Use fertilizer 92
Use both 94
TANZANIA All sources Extension All sources Extension
Lowlands 100 62 64 53
Intermediate 91 78 75 45
Highlands 100 67 91 86
Lowlands 91 86 65 56
Intermediate 94 86 67 89
Low rain 87 60 95 71
Intermediate 77 60 82 67
Lowlands 100 44 100 29
Intermediate 85 56 80 55
Southern: 44 6
Intermediate 98 73 96 74
Highlands 96 79 96 80
High 77 98 60
Low 79 100 95
Mbeya District (S. Highlands): None Rarely Regularly
Adopters 41 28 31
Non-adopters 57 18 24
High rain 100 57 100 92
MHH=male-headed households. FHH=female-headed households.
improved wheat seed (68%) in the Bale Highlands in Ethiopia claimed high seed prices as
a constraint; the proportion of non-adopters reporting price as a constraint was even
higher (73%). In Chilalo Awraja, 35% of surveyed farmers cited the high price of improved
seed as a constraint.
Cost was also often mentioned as a constraint to fertilizer use, especially in Ethiopia. The
cost of fertilizer was a constraint for 78% of adopters and 88% of non-adopters of improved
wheat varieties in the Bale Highlands, 8% of intermediate altitude farmers, and 16% of
highland farmers in the Enebssie region; between 40% and 67% of farmers sampled at the
four study sites in Western Oromia, and 83% lowland farmers and 63% of upland farmers
in Sidamo and North Omo. Related to the issue of cost is the low price of wheat,
mentioned as a constraint by 10% of both adopters and non-adopters of improved wheat
varieties in the Bale Highlands.
In the Iganga District in Uganda, 31% of adopters of improved maize varieties and 13% of
non-adopters reported cost of fertilizer as a constraint (note that the lack of information
was a key constraint for non-adopters). In Tanzania, the proportion of farmers who
mentioned cost as a constraint varied tremendously: 10% of lowland farmers, 35% of
intermediate farmers, and 5% of highland farmers in the Central Zone; 94% of farmers in
intermediate and 90% of farmers in highland areas in the Southern Highlands, a high
potential zone; 92% of highland farmers, 88% of lowland farmers, and 100% of
intermediate zone farmers in the Eastern Zone. Only 21% of farmers in the Southern Zone
mentioned price as a constraint. It should be noted again that a much higher proportion of
farmers in the Southern Highlands than in the Southern Zone actually used chemical
fertilizer, even though they reported high price as a constraint. Few farmers in the Eastern
Zone (15%) used fertilizer, while the rates were higher in the Western Zone (63%). Thus,
there does not seem to be a pattern in farmers in Tanzania reporting price as a constraint to
fertilizer use and their actually using it. It may be that farmers who use some fertilizer are
aware of its benefits and would like to use more but are constrained by price. Those not
using fertilizer may or may not say that price is the reason.
Price may be a constraint because farmers cannot purchase the inputs due to limited credit
markets or because the marginal levels of output from improved varieties do not justify
the use of improved inputs. More information on whether prices vary across regions
would also be useful in understanding these relationships. In particular, the cost of
fertilizer varies due to transportation costs: not only do costs vary across villages; farmers
may also face different transportation costs to get the materials to their fields.
Related to the issue of price is whether farmers have access to cash or credit to purchase
inputs. If farmers report that lack of credit is a constraint, we can assume that the problem
is not that the technologies are unprofitable, but that the lack of a credit market makes it
impossible for them to take advantage of available opportunities. The cost of credit is not
addressed in the statement that credit is not available. Thus, it may be that credit is simply
not available at any cost or that it is available only at high cost.
Many farmers reported credit as a constraint to using improved seed. In Ethiopia, lack of
credit was a constraint for 26% of adopters of improved wheat varieties and 31% of non-
adopters in the Bale Highlands, and for 5% of lowland and 12% of intermediate farmers in
Sidamo and North Omo. Many farmers in Sidamo and North Omo also noted that lack of cash
was a constraint (6% of lowland farmers and 19% of intermediate zone farmers).
When asked about constraints to adopting fertilizer, farmers did not mention credit. In
Enebssie region, Ethiopia, however, 51% of intermediate zone farmers and 34% of highland
farmers said that shortage of cash was an important constraint to fertilizer adoption. Reports
of lack of credit and cash shortage are related to each other and to the price of fertilizer.
Table 6 provides details on the use of credit where this information was reported.
The use of credit overall was much higher in Ethiopia than elsewhere. Credit is provided
through the State in Ethiopia either through State-run banks or cooperatives. The proportion
of people using credit to purchase fertilizer is especially high. It is clear that the majority of
people who use fertilizer purchase it on credit. In Tanzania, however, the use of formal credit
is much more limited, even in areas where fertilizer is heavily used. The Tanzania studies
report rates by zone of 0-44% of farmers using credit. Within Tanzania, there are fairly wide
differences in use of formal credit. Farmers in higher altitude and rainfall areas were more
likely to use credit. It is not clear whether credit is more widely available in these areas or
whether farmers there are more likely and able to take advantage of available credit.
Many farmers reported that improved seeds and fertilizer were either not available, or, in the
case of fertilizer, that delivery was too late. These issues will be discussed further in the next
section, but they imply that, among other reasons, farmers do not use these technologies
because they are simply not available.
To the extent that lack of information is the binding constraint, there continues to be an
important role for extension services. These services may be provided through the
government or through non-governmental organizations (NGOs). It is clear that farmers are
eager for more and better information.
To the extent that farmers do not adopt improved technologies because they are not profitable,
given the state of the technology and their circumstances, there are two directions that policies
can take. The first is to increase the productivity of improved varieties, thereby increasing
output. The second is to reduce input costs for farmers. Subsidizing costs is not sustainable,
and it is crucial to think about how to reduce input costs by changes in infrastructure,
transportation, credit availability, and markets.
It is difficult to determine which factors are behind farmers' decisions not to use new
technologies. Farmers often report that input prices are too high, but this means that prices are
too high given their knowledge and expected returns. Seeds and fertilizer may be unavailable
in a particular region in part because they cannot profitably be sold and used in that area.
Inputs may not be available if transportation costs for inputs and outputs are too high.
Table 6. Percentage of farmers using formal credit, Ethiopia, Kenya, Tanzania, Uganda, 1996-1999.
Country/region Formal credit Credit for seed Credit for fertilizer Use of chemical fertilizer
Adopters 48 85 95
Non-adopters 9 60 75
Ada-MHH 100 100
Ada-FHH 96 100
Lume-MHH 100 100
Lume-FHH 96 100
Gimbichu-MHH 100 100
Gimbuchu-FHH 100 100
Intermediate 30 70
Highlands 18 27
Adopters 37tt 55 70
Non-adopters 14 86 27
Sidomo and North
Lowland 21 26 58
Intermediate 26 34 70
Adopters 3 46
Non-adopters 9 23
Lowlands 7 17
Intermediate 0 77
Highlands 13 17
Lowlands 6 17
Intermediate 14 8
Low rain 3 50
Intermediate 2 48
High rain 0 100
Lowlands 8 64
Intermediate 19 20 25 44
Southern: 0 3
Intermediate 22 65
Highlands 20 79
High 44 0 16 66
Low 32 60
Mbeya District (S. Highlands):
Adopters 11 40
Non-adopters 0 0
Adopters 17 3
Non-adopters 16 3
MHH=male-headed household; FHH=female-headed household.
t Only credit received from the state is reported here. Credit may also be obtained from others.
tt The numbers reported are for credit for seed and fertilizer combined.
Are Improved Seeds and Fertilizer Available?
In addition to understanding constraints that farmers face, it is important to know whether
improved technologies are actually available to farmers in their villages. In this section, we
examine this by looking at the seed and fertilizer industries in each of the four countries
and farmers' perceptions of whether the lack of availability of seed and fertilizer is a
constraint to adoption. When seeds or fertilizer are unavailable, it is challenging to ascertain
whether the issue is a problem with the distribution network or lack of effective demand.
In a few of the studies, farmers were asked
directly to list constraints to adoption of
improved varieties. Table 7 summarizes
responses on whether seeds and fertilizer
are available. In most areas of Tanzania,
farmers did not report that either seeds or
fertilizer were unavailable. The exceptions
were in the low and intermediate rainfall
areas of the Lake Zone, where relatively
high numbers of farmers reported that
improved seeds were unavailable. Non-
adopters of improved wheat seed in Mbeya
District (Southern Highlands) also reported
that seed was unavailable.
Each of the four countries has a different
system for research and development and
for distributing seed and fertilizer. For each
of the four countries, we will briefly discuss
the seed sector and the seed and fertilizer
distribution systems. These systems differ
markedly across the four countries, although
all are moving towards greater reliance on
the market and increased privatization.
The four main sources of seed for Ethiopian
farmers are purchase, other farmers,
extension, and recycling own seed. The
distribution of these sources varies widely
across study regions. In the Bale Highlands,
51% of the farmers reported purchasing new
seed, while 31% obtained seed through
extension and 15% from other farmers.
Farmers did not mention obtaining seed
from their own fields, although this was a
primary source of seeds for farmers in other
Table 7. Percentage of surveyed farmers reporting
unavailability of seed or fertilizer, Ethiopia, Kenya,
Tanzania, Uganda, 1996-1999.
Country/ Seeds Fertilizer
region unavailable unavailable
Adopters 30 10
Nonadopters 27 10
Chilalo Arwaja: 42
Sidamo and North Omo Zone:
Low 9 3
Intermediate 11 4
Western Shewa-Challya NA 11
Western Shewa-BakeTibe NA 0
Eastern Wellega Bila-Sayo NA 40
Eastern Wellega-Sibu Sire NA 0
Low rainfall 63 0
Intermediate rainfall 94 0
High rainfall 0
Intermediate 0 15
Highlands 0 5
Mbeya District (S. Highlands):
Adopters 0 6
Non-adopters 76 5
Adopters 14 0
Non-adopters 47 0
The Uganda study did not look at fertilizer adoption
None of the other information for this table was
reported for Kenya.
areas. Seed purchases were much lower in areas that reported their seed source. Most
obtained seed either from other farmers or from their own fields. Extension services provide
as little as 5% to 34% of seed for sowing, depending on the area.
The Ethiopian seed industry continues to be dominated by the public sector. Improved
varieties are developed by the national agricultural research system and development
programs or introduced from outside. Public institutions are responsible for producing and
distributing seed to farmers, although some private companies are now entering the seed
industry and have begun research on hybrid seed production, marketing, and distribution.
Seed supply has been constrained by inefficient public seed enterprises, poor seed
promotion, poor transportation, and inappropriate agricultural and pricing policies. The
limited availability of fertilizer further constrained the use of improved seed. Most seed in
the smallholder sector is still produced by farmers (Hailu 1992).
Since its establishment in the 1950s, the Ethiopian national maize research program has been
introducing germplasm from outside sources. Ethiopia began participating in the Eastern
Africa Cooperative Maize Trials in 1967. Hybrids from Kenya and Zimbabwe had a 30%
yield advantage over tested varieties, so State farms were encouraged to import Kenyan
hybrids. In the early 1980s, Ethiopia began a national maize breeding program that
developed a number of hybrids suitable for Ethiopian conditions.
The Ethiopian Seed Enterprise (ESE), which was incorporated in 1979, is run as an
interministerial seed board with autonomous status to function as a profit making
enterprise. It dominates the production, multiplication, processing, and distribution of seed.
Since a policy reform in 1991, ESE has obtained seed from research centers and makes
contractual arrangements with private investors, the Ethiopian Agricultural Research
Organization, seed farms, and State farms for the production and multiplication of improved
maize seed. The ESE has acknowledged that it is not meeting demand for maize seed.
All three Kenyan studies on the use of improved varieties included information on where
seed was purchased. (It is not reported how the rest of the seed was obtained). Seventeen
percent of smallholder (<20 ha) wheat farmers in the Naruk, Nakuru, and Uasin Districts
reported that they obtained seed from the Kenya Seed Company (KSC); 9% from the Kenya
Farmer's Association, and the majority (56%) from other farmers. The patterns were similar
with large-scale farmers: almost 50% reported obtaining seed from other farmers. On the
Coast, 54% used their own recycled seed while 33% purchased seed. Sixty-five percent of
purchased seed was certified seed.
The Kenya maize seed industry, especially the hybrid seed development of the 1960s and
1970s, has been hailed as one of the success stories of agricultural development in Africa.
Yet, maize seed sales have stagnated for the past 15 years. An attempt was made recently to
liberalize the sector in the hope that competition would improve the availability of quality
seed and thereby increase adoption of improved varieties. Although it is too early to judge
its impact, some changes can already be observed.
Maize research in Kenya started before independence in the 1950s. In 1955, the first maize
breeder was posted in Kitale to develop hybrid varieties (Gerhart 1975). The seed industry
started in 1956 when European settlers formed the KSC to produce seed, initially for pasture.
In 1963, the Kenyan government contracted with KSC to produce new maize hybrid seed.
Later, in the 1980s the Kitale program developed new varieties jointly with KSC and the
government obtained a majority share in the company. Maize hybrids were very successful
and seed sales increased from 235 tons in 1963 to more than 20,000 tons in 1985 (Ndambuki
1998). Hybrids showed a clear yield advantage over other varieties, but their deployment
was also supported by extensive agronomic research (to combine hybrid seed with fertilizer
and other management practices), extensive demonstration trials, and a large extension
effort including radio dissemination and agricultural credit to purchase inputs. Distribution
was organized through the Kenya Farmers' Association and other distributors.
After the initial success of improved varieties, sales have hovered around 20,000 tons per
year since 1985, although they reached a low of 13,202 tons in 1996-the last year for which
figures are available (Ndambuki 1998). Liberalization of the seed sector began the same year.
The seed sector was split between government services and companies; new seed
companies, both local and multinational, have proliferated along with new distribution
outlets stockistss). Quality control was transferred from the Kenya Agricultural Research
Institute (KARI) to a new regulatory agency, the Kenya Plant Health Inspectorate Service
(KEPHIS). The Kenya Agricultural Research Institute's close link with KSC has also been
severed and several disputes around property rights of old hybrids have soured the
relationship. The KSC still has a quasi monopoly on late-maturing highland varieties, but
new companies are moving in fast in intermediate-maturity varieties for the midaltitudes.
The KSC has a strategy of a single price regardless of seed type or place of sale, whereas
newcomers are much more flexible. With the arrival of new players, market information is
much harder to come by. Despite the liberalization and delinking of quality control, seed
regulations are still cumbersome and expensive.
The liberalization of the agricultural sector in general, which started in the late 1980s, also
affected the demand for improved seed and other inputs. The number of distributors of
agricultural inputs has mushroomed, making inputs widely available. However, subsidies
have been abolished and the agricultural credit system has collapsed. Fertilizer subsidies as
well as import taxes have also been abolished and competition to import and distribute is
heavy (no fertilizers are produced in Eastern Africa). However, transport costs remain high
and have increased the effective cost of fertilizers, although fertilizer has become more
available. Since hybrids are less effective without fertilizer, this is assumed to decrease the
demand for hybrid seed.
Four of the seven Tanzanian studies discussed how farmers obtained their seeds. In the high
rainfall area of the Lake Zone, 85.7% of farmers reported purchasing seed regularly. The
primary source for seed was NGOs. In the low rainfall zone, only 29% of farmers purchased
seed regularly from the cooperative union or local market. No farmers in the intermediate
zone reported purchasing seed regularly, but purchases, when made, were primarily from
the cooperative union. The Southern Highlands again stands out as being different from
other zones in Tanzania. Ninety percent of farmers in the highlands and 97% in intermediate
areas purchased maize seeds regularly. The majority reported that their source of seeds were
stockists. In Western Tanzania, only 14% of respondents reported purchasing new seed
regularly. Forty-seven percent of wheat farmers in Mbeya District (Southern Highlands)
obtained seed from other farmers and 31% from the market; the remainder was split between
research, extension, NGOs, and retaining one's own seed.
The seed industry in Tanzania has been involved in maize breeding in Tanzania for more
than 20 years. Open-pollinated varieties that were released in the 1960s are still widely used
in some areas. In 1974, the National Maize Research Program was launched, which released
two hybrids and six OPVs in the 1970s and 1980s.
The input market was liberalized in 1990 and the number of businesses engaged in selling
inputs has increased dramatically. In many areas, inputs are available through village
stockists and many shortages have been reduced. Inputs are available in a timely manner, but
at a much higher price than previously. This corresponds to the fact that no farmers reported
a lack of availability of seed or fertilizer.
The public sector seed source, Tanseed, markets locally bred hybrids. They have increasingly
had competition from the private seed companies, Cargill and Pannar. By 2001, the public
sector had released about 15 improved varieties and the private sector had released 12. Even
though private companies produce purer, more uniform, and higher yielding seed, prices are
also higher and grains have poorer storage, pounding, and taste qualities.
Uganda suffered a near total collapse of the agricultural research system, seed multiplication
capacity, output markets, input distribution networks, and extension services in 1986. The
recommended variety Longe 1 was released in 1991 and was made available to farmers
primarily though on-farm trials. Many farmers continue to grow Kawanda Compsite A, a
variety released over 20 years ago and no longer distributed. Both of these varieties are OPVs.
The only study was in Iganga District, where there have been verification and demonstration
trials for maize technologies since the 1980s, many conducted in farmers' fields. Thus, we
would expect a high proportion of farmers in this District to have adopted the recommended
variety. Most farmers who grew Longe 1 (51%) purchased or retained it from the previous
harvest (46%). Seed for the other variety grown (not considered an "improved variety" since
it was released in 1971) was saved from previous harvests.
The Ugandan Ministry of Agriculture multiplies breeder seed, certifies seed, and markets
through private entrepreneurs and extension services. Associated agricultural inputs, such as
fertilizer, are available mostly through private suppliers in urban areas. A large number of
non-adopters (47%) reported that seeds were unavailable, although only 14% of adopters said
that this was so.
The availability of seed and fertilizer varies from Kenya and Tanzania, where they are
widely available locally through private shops, to Ethiopia, where seed is less readily
available for purchase. Although the simple adoption numbers as reported in Table 2 do
not necessarily reflect patterns of adoption by country or, thus, by availability of seed, it
does seem that more farmers purchase seed in areas where seed is available. Causality
should not necessarily be inferred-it may be simply that the private sector is more willing
to supply seed in areas where farmers would choose to purchase it.
Implications for Policy and Research
One of the advantages of having a number of similar studies that vary across
agroecological zones and also across countries is that it allows us to ask how institutional
factors affect adoption of technology. These institutional factors include government
policies, such as the privatization of input sectors discussed above and broader institutional
factors, such as market access. Comparisons across areas with differing agricultural
potential should also be possible.
The CIMMYT/national system micro-level studies provide some insights into these issues.
Even though they are not detailed enough for formal hypothesis testing, they do provide
some directions and suggestions for further research.
One interesting question is whether areas with higher agricultural potential are more likely
to adopt improved technologies. The highest levels of adoption in Ethiopia were in Chilalo
Awraja and Northwestern Province, areas at the highest altitude among those surveyed.
High altitudes tend to get more rainfall, although not all studies report on rainfall. Yet in
other areas with relatively high potential, especially the Central Highlands, there was low
adoption of improved wheat and maize, although 100% of the farmers used fertilizer.
Within survey areas, farmers at higher altitudes were more likely to adopt improved
technologies than those at lower altitudes. A similar pattern seems to hold in Tanzania,
where high potential areas included the Northern Zone and the Southern Highlands. The
Northern Zone had high levels of adoption of improved maize varieties and the Southern
Highlands had lower levels of adoption, but both are areas where farmers were most likely
to purchase new seed. Thus, farmers in the South may have been taking better advantage of
improved technologies than simple adoption proportions suggest. It is hard to see a pattern
emerging from altitude and rainfall data alone, within study areas in Tanzania. This may be
because other factors were more important or because other factors influencing agricultural
productivity are not captured in the rainfall and altitude data.
A second question is whether market access plays a role in determining the level of
adoption. Market access could affect adoption decisions in two ways. First, farmers would
be more likely to produce for the market and it would be cheaper and easier to get goods to
market. Second, market access reduces the costs-both in time and money-of obtaining
inputs. In Ethiopia, the areas with the greatest market access were Chilalo Awraja and
Wolmera Woreda. The former had a relatively high level of adoption of improved wheat.10
In Tanzania, even areas with low market access had relatively high levels of adoption-in
part because the term "adoption" in Tanzania includes all improved materials, not just
recommended or recently released variety. The high market access Northern Zone had one
of the highest levels of adoption of improved seed. There does not seem to be a direct link
between market access defined across survey sites and level of adoption, possibly because
of significant uncontrolled variation within surveyed zones.
A third question is whether intensification results in higher levels of adoption. All studies
listed either the population density for the region as a whole, the average farm size of
surveyed households, or both. Farm size is included in most regression analyses; however,
we might expect that the average farm size for a region reflects the relative abundance or
scarcity of land. There were no clear relationships between population density or farm size
and adoption levels at aggregate levels for which data were available. The aggregate data
may mask many relationships to be found in much more localized data.
The most important finding from these studies is that technology adoption is taking place
across Eastern Africa. This is important, especially given that some researchers are skeptical
about the willingness of farmers in the region to innovate and adopt new approaches. There
was no evidence that farmers are failing to take advantage of these technologies where it is
economically advantageous for them to do so.
There are no simple correlations across all studies that indicate one factor as key in the
adoption process. Extension is the variable that is most statistically significant in explaining
technology use. The lack of a clear correlation between individual indicators across sites
may be that these relationships are difficult to pick up in the data, especially since many of
the explanatory variables are highly correlated. It may also be that policies, institutions, and
infrastructure are important determinants of adoption levels and micro-studies do not pick
up these factors, even within regions.
The descriptive information available from these studies will be of use to policy makers
within the region. These data have not previously been available. This basic information
about the levels of adoption is critical for priority setting and impact assessment.
A number of lessons have been learned about how to carry out adoption studies across sites
to allow us to address some bigger issues. It is crucial to standardize what we mean by an
adopter. Clearly, the use of improved seed could be better defined in different categories
such as: purchase of improved seed (in a certain period), use of recycled hybrids, and the
use of recycled, improved open-pollinated varieties. Adding the number of cycles can
further refine recycling. Moreover, adoption studies should cover clearly defined areas and
the sampling procedure should result in a representative sample of the target area.
10 Adoption levels are not available for Wolmera Woreda.
A number of policy implications can be drawn from these studies. Although the
econometric estimations do not necessarily give clear policy implications on their own, they
can be used in combination with the descriptive data and the discussions with staff at
national research centers, to draw some inferences.
Research and extension are vital for the development and dissemination of new
technologies and these services need to be strengthened. Research, especially maize and
wheat breeding research, needs to be more relevant to farmers' circumstances and
preferences. This suggests that there is a continuing need to link research and extension.
Although governments should increase budgetary support for extension services,
strengthening these services may also involve private sector and NGO participation.
Policies should also support the development and expansion of efficient markets for inputs
and outputs. The recent liberalization of these markets in many Eastern African countries
should be continued, with emphasis on finding ways to decrease input costs. Although the
private sector clearly plays a role in this, the government also has a role by providing
physical infrastructure. Finally, the rural credit system should be strengthened to improve
availability and access to credit.
These studies have also opened up a number of bigger questions about technology adoption
in Eastern Africa; specifically, relating to the role of policies, institutions, and infrastructure
in the adoption process. The CIMMYT/national system studies suggest these may be
important and encourage us to move beyond micro-level studies to answer these questions.
CBS. 1990. Crops focus survey. Nairobi, Kenya: Central Bureau of Statistics, Ministry of Planning and National
CIMMYT.1993. The adoption of agricultural ~..,.1,.... A guide for survey design. Mexico, D.F.: CIMMYT.
Doss, C.R. 1999. Twenty-five years of research on women farmers in Africa: Lessons and implications for agricultural
research institutions; with an annotated bibliography. CIMMMYT Economics Program Paper No. 99-02.
Mexico D.F: CIMMYT.
Doss, C.R. 2003. Understanding farm level technology adoption: Lessons learned from micro surveys in Eastern
Africa. CIMMYT Economics Program Working Paper No. 03-07. Mexico, D.F.: CIMMYT
Ekboir, J. (ed.). 2002. CIMMYT 2000-2001 World Wheat Overview and Outlook: Developing No-Till Packages for
Small-Scale Farmers. Mexico, D.F.: CIMMYT.
Francis, E. and J.Hoddinott. 1993. Migration and differentiation in Western Kenya: A tale of two sub-locations.
Journal of Development Studies 30(1):115-45.
FSD. 1992. Comprehensive Food Security. Dar-Es-Salaam, Tanzania; Food Security Department, Government of the
United Republic of Tanzania, and the United Nations Food and Agriculture Organization (FAO).
FSD. 1996. Tanzania food security bulletin No. 2, April/May 1996. Dar-Es-Salaam, Tanzania: Food Security
Department, Ministry of Agriculture.
Gerhart, J. 1975. The lttrn -.. of hybrid maize in Western Kenya. Mexico, D.F: CIMMYT.
Hailu, G. 1992. Availability and use of seed in Ethiopia. Addis Ababa, Ethiopia: Program support unit, Canadian
International Development Agency.
Morris, M., J. Risopoulos, and D. Beck.1999. Genetic change in farmer-recycled maize seed: A review of the evidence.
CIMMYT Economics Working Paper No. 99-07. Mexico, D.F.: CIMMYT.
Ndambuki, F M. 1998. Kenya Seed Company: Activities and organization. In: Mwangi W. M. (comp.). Seed
production and supply policy: Teaching notes for the Training Workshop on Seed Production and Supply Policy.
Addis Ababa, Ethiopia: CIMMYT.
Appendix 1. CIMMYT/National System Eastern
African Adoption Studies
Tiruneh, A., T. Tesfaye, W. Mwangi, and H. Verkuijl. 2001. Gender differentials in agricultural production and
decision-making among .I:,.illl., '. i in Ada, Lume, and Gimbichu Woredas of the Central Highlands of
Ethiopia. Mexico, D.F.: CIMMYT and Ethiopian Agricultural Research Organisation (EARO).
Getahun Degu, W.Mwangi, H. Verkuijl, and Abdishekur Wondimu. 2000. An assessment of the adoption of
seed and fertilizer packages and the role of credit in smallholder maize production in Sidamo and North
Omo Zone, Ethiopia. Mexico, D.F.: CIMMYT and Ethiopian Agricultural Research Organisation
Abdissa G., G. Aboma, H. Verkuijl, and W. Mwangi. 2001. Farmers' maize seed systems in Western Oromia,
Ethiopia. Mexico, D.E: CIMMYT and Ethiopian Agricultural Research Organisation (EARO).
Tesfaye Zegeye, Girma Taye, D. Tanner, H. Verkuijl, Aklilu Agidie, and W. Mwangi. 2001. Adoption of
improved bread wheat varieties and inorganic fertilizer by small-scale farmers in Yelmana Densa and Farta
Districts of Northwestern Ethiopia. Mexico, D.F.: CIMMYT and Ethiopian Agricultural Research
Bekele Hundie Kotu, H. Verkuijl, W. Mwangi, and D. Tanner. 2000. Adoption of improved wheat technologies
in Adaba and Dodola Woredas of the Bale Highlands, Ethiopia. Mexico, D.F.: CIMMYT and Ethiopian
Agricultural Research Organisation (EARO).
Alemu Hailye, H. Verkuijl, W. Mwangi, and Asmare Yalew. 1998. Farmers' wheat seed sources and seed
management in Enebssie area, Ethiopia. Mexico, D.F.: CIMMYT and Institute of Agricultural
Research (IAR now EARO).
Hailu Beyene, H. Verkuijl, and W. Mwangi. 1998. Farmers'seed sources and management of bread wheat in
Wolmera Woreda, Ethiopia. Mexico, D.F.: CIMMYT and Institute of Agricultural Research (IAR now
Regassa Ensermu, W. Mwangi, H. Verkuijl, Mohammad Hassena, and Zewdie Alemayehu. 1998. Farmers'
wheat seed sources and seed management in Chilalo Awraja, Ethiopia. Mexico, D.E: CIMMYT and
Makokha, S., S. Kimani, W. Mwangi, H. Verkuijl, and F. Musembi. 2001. Determinants of fertilizer and
manure use in maize production in Kiambu District, Kenya. Mexico, D.F.: CIMMYT and KARI.
Salasya, B.D.S., W. Mwangi, H. Verkuijl, M.A. Odendo, and J.O. Odenya. 1998. An assessment of adoption of
seed and fertilizer packages and the role of credit in smallholder maize production in Kakamega and Vihiga
Districts. Mexico, D.F: CIMMYT and KARI.
Ouma, J., F. Murithi, W. Mwangi, H. Verkuijl, M.Gethi, H. De Groote. 2002. Adoption of seed and fertilizer
technologies in Embu District, Kenya. Mexico, D.F.: CIMMYT and KARI.
Gamba, P., C. Ngugi, H. Verkuijl, W. Mwangi, and F. Kiriswa. 2001. Wheat farmer's seed management and
varietal adoption in Tanzania. Mexico, D.F.: CIMMYT and KARI
Wekesa E., W. Mwangi, H. Verkuijl, K. Danda, H. De Groote. 2002. Adoption of maize production technologies
in the Coastal Lowlands of Kenya. Mexico, D.F.: CIMMYT and KARI.
Mussei, A., J. Mwanga, W. Mwangi, H. Verkuijl, R. Mongi, and A. Elanga. 2001. Adoption of improved wheat
technologies by small-scale farmers, Southern Highlands, Tanzania. Mexico, D.F: CIMMYT and the
United Republic of Tanzania.
Kaliba, A.R.M., H. Verkuijl, W. Mwangi, A.J.T. Mwilawa, P. Anandajayasekeram, and A.J. Moshi. 1998.
Adoption of maize production technologies in Central Tanzania. Mexico, D.F.: CIMMYT, United
Republic of Tanzania, and Southern African Centre for Cooperation in Agricultural Research
Bisanda, S., W. Mwangi, H. Verkuijl, A.J. Moshi, and P. Anadajayasekeram. 1998. Adoption of maize
production technologies in Southern Highlands of Tanzania. Mexico, D.F.: CIMMYT, United Republic
of Tanzania, and Southern African Centre for Cooperation in Agricultural Research (SACCAR).
Katinila, R., H. Verkuijl, W. Mwangi, P. Anadajayasekeram, and A.J. Moshi. 1998. Adoption of maize
production technologies in Southern Tanzania. Mexico, D.F.: CIMMYT, United Republic of Tanzania,
and Southern African Centre for Cooperation in Agricultural Research (SACCAR).
Nkonya, E., P. Xavery, H. Akonaay, W. Mwangi, P. Anandajayasekeram, H. Verkuijl, D. Martella, and A.J.
Moshi. 1998. Adoption of maize production technologies in Northern Tanzania. Mexico, D.F.:
CIMMYT, United Republic of Tanzania, and Southern African Centre for Cooperation in
Agricultural Research (SACCAR).
Kaliba, A.R.M., H. Verkuijl, W. Mwangi, D.A. Byamungu, P. Anadajayasekeram, and A.J. Moshi. 1998.
Adoption of maize production technologies in Western Tanzania. Mexico, D.F.: CIMMYT, United
Republic of Tanzania, and Southern African Centre for Cooperation in Agricultural Research
Kaliba, A.R.M., H. Verkuijl, W. Mwangi, A. Moshi, A. Chilangane, J.S. Kaswende, and P.
Anadajayasekeram. 1998. Adoption of maize production technologies in Eastern Tanzania. Mexico, D.F.:
CIMMYT, United Republic of Tanzania, and Southern African Centre for Cooperation in
Agricultural Research (SACCAR)..
Mafuru, J., R. Kileo, H. Verkuijl, W. Mwangi, P. Anandajayasekeram, A. J. Moshi. 1999. Adoption of maize
production technologies in Lake zone of Tanzania. Mexico. D.F.: CIMMYT, United Republic of
Tanzania, and Southern African Center for Cooperation in Agricultural Research (SACCAR).
Ntege-Nanyeenya, W.M. Mugisa-Mutetikka, W. Mwangi, and H. Verkuijl. 1997. An assessment of factors
iff. l i.: adoption of maize production technologies in Iganga District, Uganda. Mexico, D.F: CIMMYT
and National Agricultural Research Organization (NARO).
INTERNATIONAL MAIZE AND WHEAT IMPROVEMENT CENTER
Apartado Postal 6-641, 06600 Mexico, D.F., Mexico
Worldwide Web site: www.cimmyt.org