• TABLE OF CONTENTS
HIDE
 Front Cover
 Title Page
 Table of Contents
 Acknowledgement
 Summary
 Highlights of conditions found...
 Agricultural research and...
 Past experience: The evidence
 From past experience to lessons...
 Utilization of lessons learned...
 Appendix A: Literature cited
 Appendix B: Impact evaluation team...
 Appendix C: Summaries of impact...
 Appendix D: Proceedings of a workshop...
 Back Cover






Group Title: A.I.D. program evaluation report ;, no. 10
Title: Strengthening the agricultural research capacity of the less developed countries
CITATION PAGE IMAGE ZOOMABLE PAGE TEXT
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00053898/00001
 Material Information
Title: Strengthening the agricultural research capacity of the less developed countries lessons from AID experience
Series Title: A.I.D. program evaluation report
Physical Description: 1 v. (various pagings) : ; 28 cm.
Language: English
Creator: Murphy, Josette
United States -- Agency for International Development
Publisher: U.S. Agency for International Development
Place of Publication: Washington D.C.?
Publication Date: [1983]
 Subjects
Subject: Agricultural assistance, American   ( lcsh )
Agricultural research -- Developing countries   ( lcsh )
Genre: federal government publication   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Bibliography: Includes bibliographies.
Statement of Responsibility: by Josette Murphy.
General Note: Distributed to depository libraries in microfiche.
General Note: "September 1983."
General Note: "PN-AAL-020"--Cover.
Funding: Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
 Record Information
Bibliographic ID: UF00053898
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 001283040
oclc - 10368532
notis - AGD3701

Table of Contents
    Front Cover
        Front Cover
    Title Page
        Page i
        Page ii
    Table of Contents
        Page iii
        Page iv
    Acknowledgement
        Page v
    Summary
        Page vi
        Page vii
        Page viii
    Highlights of conditions found favorable to effective research systems
        Page ix
    Agricultural research and development
        Page 1
        Research, food production, and population growth
            Page 1
        AID's assistance to agricultural research
            Page 2
            Page 3
            Page 4
            Page 5
    Past experience: The evidence
        Page 6
        The findings of eight impact evaluations
            Page 6
            Evaluation methodology
                Page 6
            Kenya
                Page 7
                Page 8
                Page 9
            Central America
                Page 10
            Korea
                Page 10
            Guatemala
                Page 11
            Nepal
                Page 12
            Thailand
                Page 13
            Tunisia
                Page 14
            West Africa
                Page 14
        Other evaluations and studies
            Page 15
            Review of routine project evaluations
                Page 15
            Conference on impact of agricultural research, Leesburg, Virginia
                Page 16
            The Asia agricultural research review project
                Page 17
    From past experience to lessons learned
        Page 18
        Introduction
            Page 18
        Research should be oriented towards farmers' needs and constraints
            Page 19
            Page 20
            Page 21
            The impact of research on food production
                Page 22
                Page 23
                Page 24
                Page 25
                Page 26
                Page 27
            The impact of research on farmers' well-being and on rural equity
                Page 28
                Page 29
            Conclusions
                Page 30
        The utilization of research findings is dependent on a favorable political and economic environment
            Page 31
            Technological solutions alone cannot solve problems which are basically economic in nature
                Page 31
            Host government commitment is essential
                Page 32
                Page 33
        Characteristics of effective agricultural research systems
            Page 34
            The effectiveness of an institution depends on its place in the administrative structure and the resources it receives
                Page 35
            Institutional development and the concomitant training of scientists is a long-term, complex process of critical importance to the sustainability of a research project
                Page 36
                Page 37
            Linkages among national and international research institutions are essential
                Page 38
                Page 39
        Logistical difficulties should not be underestimated
            Page 40
    Utilization of lessons learned for future AID activities
        Page 40
        The changing relations between host governments and donor institutions
            Page 41
            Page 42
        Suggestions for AID assistance
            Page 43
            Planning assistance to agricultural research activities
                Page 43
                Page 44
            Project design
                Page 45
            AID management of research projects
                Page 46
            AID evaluation of its assistance to research
                Page 47
                Page 48
                Page 49
    Appendix A: Literature cited
        Appendix 1
        Page 1
        Page 2
    Appendix B: Impact evaluation team members and working group members
        Appendix 1
        Page 1
        Page 2
        Page 3
    Appendix C: Summaries of impact evaluation reports
        Appendix 1
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
    Appendix D: Proceedings of a workshop on impact of agricultural research
        Appendix 1
        Appendix 2
        Appendix 3
        Appendix 4
        Appendix 5
        Appendix 6
        Appendix 7
        Appendix 8
        Appendix 9
        Appendix 10
        Appendix 11
        Appendix 12
        Appendix 13
        Appendix 14
        Appendix 15
        Appendix 16
        Appendix 17
        Appendix 18
        Appendix 19
        Appendix 20
        Appendix 21
        Appendix 22
        Appendix 23
        Appendix 24
        Appendix 25
        Appendix 26
        Appendix 27
        Appendix 28
        Appendix 29
        Appendix 30
        Appendix 31
        Appendix 32
        Appendix 33
        Appendix 34
        Appendix 35
        Appendix 36
        Appendix 37
        Appendix 38
        Appendix 39
        Appendix 40
        Appendix 41
        Appendix 42
        Appendix 43
        Appendix 44
        Appendix 45
        Appendix 46
        Appendix 47
        Appendix 48
        Appendix 49
        Appendix 50
        Appendix 51
        Appendix 52
        Appendix 53
        Appendix 54
        Appendix 55
        Appendix 56
        Appendix 57
        Appendix 58
        Appendix 59
        Appendix 60
        Appendix 61
        Appendix 62
        Appendix 63
        Appendix 64
        Appendix 65
        Appendix 66
        Appendix 67
        Appendix 68
        Appendix 69
        Appendix 70
        Appendix 71
        Appendix 72
        Appendix 73
        Appendix 74
        Appendix 75
        Appendix 76
        Appendix 77
        Appendix 78
        Appendix 79
        Appendix 80
        Appendix 81
        Appendix 82
        Appendix 83
        Appendix 84
        Appendix 85
        Appendix 86
        Appendix 87
        Appendix 88
        Appendix 89
        Appendix 90
        Appendix 91
        Appendix 92
        Appendix 93
        Appendix 94
        Appendix 95
        Appendix 96
        Appendix 97
        Appendix 98
        Appendix 99
        Appendix 100
        Appendix 101
        Appendix 102
        Appendix 103
        Appendix 104
        Appendix 105
        Appendix 106
        Appendix 107
        Appendix 108
        Appendix 109
        Appendix 110
        Appendix 111
        Appendix 112
        Appendix 113
        Appendix 114
        Appendix 115
        Appendix 116
        Appendix 117
        Appendix 118
        Appendix 119
        Appendix 120
        Appendix 121
        Appendix 122
        Appendix 123
        Appendix 124
        Appendix 125
        Appendix 126
        Appendix 127
        Appendix 128
        Appendix 129
        Appendix 130
        Appendix 131
        Appendix 132
        Appendix 133
        Appendix 134
        Appendix 135
        Appendix 136
        Appendix 137
        Appendix 138
        Appendix 139
        Appendix 140
        Appendix 141
        Appendix 142
        Appendix 143
        Appendix 144
        Appendix 145
        Appendix 146
        Appendix 147
        Appendix 148
        Appendix 149
        Appendix 150
        Appendix 151
        Appendix 152
        Appendix 153
        Appendix 154
        Appendix 155
    Back Cover
        Back Cover
Full Text
o01. I-


A.I.D. Program Evaluation Report No. 10
Strengthening the Agricultural Research
Capacity of The Less Developed Countries:
Lessons from AID Experience


September 1983
U.S. Agency for International Development


PN-AAL-020















STRENGTHENING THE AGRICULTURAL RESEARCH
CAPACITY OF THE LESS DEVELOPED COUNTRIES:
LESSONS FROM AID EXPERIENCE




A.I.D. Program Evaluation Report No. 10







by

Josette Murphy
(Bureau for Program and Policy Coordination)


U.S. Agency for International Development

September 1983




The views and interpretations expressed in this report are
those of the author and should not be attributed to the Agency
for International Development.































A.I.D. EVALUATION PUBLICATIONS


A complete list of reports issued in the A.I.D. Evaluation
Publication series is included in the back of this document,
together with information for ordering reports.









TABLE OF CONTENTS


Page

Acknowledgments. ........ ............ ....... ... .......... .

Summaary .................................vii

Highlights of Conditions Found Favorable to Effective
Research Systems............... ........ ... ................ x

I. Agricultural Research and Development...................1

A. Research, Food Production, and Population Growth.....1
B. AID's Assistance to Agricultural Research..........2

II. Past Experience: The Evidence.......................... 6

A. The Findings of Eight Impact Evaluations............6
1. Evaluation Methodology.......................... 6
2. Kenya........................................... 7
3. Central America.............................. 10
4. Korea............ ...............................10
5. Guatemala............... ..................... .. 11
6. Nepal........................................... 12
7. Thailand.........................................13
8. Tunisia................................ ........ 14
9. West Africa..................................... 14
B. Other Evaluations and Studies .....................15
1. Review of Routine Project Evaluations..........15
2. Conference on Impact of Agricultural
Research, Leesburg, Virginia.................16
3. The Asia Agricultural Research Review
Project ..................................... 17

III. From Past Experience to Lessons Learned................18

A. Introduction....................................... 18
B. Research Should Be Oriented Towards Farmers'
Needs and Constraints........................... 19
1. The Impact of Research on Food Production......22
2. The Impact of Research on Farmers'
Well-Being and on Rural Equity............... 28
3. Conclusions..................................... 30
C. The Utilization of Research Findings Is
Dependent on a Favorable Political and
Economic Environment............................. 31










TABLE OF CONTENTS (continued)


Page
1. Technological Solutions Alone Cannot Solve
Problems Which Are Basically Economic in
Nature.............. ................. .... 31
2. Host Government Commitment Is Essential........32
D. Characteristics of Effective Agricultural
Research Systems .............. .................34
1. The Effectiveness of an Institution
Depends on Its Place in the Administrative
Structure and the Resources It Receives......35
2. Institutional Development and the Concomitant
Training of Scientists Is a Long-Term,
Complex Process of Critical Importance
to the Sustainability of a Research
Project .................................. 36
3. Linkages Among National and International
Research Institutions Are Essential..........38
E. Logistical Difficulties Should Not Be
Underestimated.................... ............. 40

IV. Utilization of Lessons Learned for Future AID
Activities.................................. ... .. 40

A. The Changing Relations Between Host Governments
and Donor Institutions........................... 41
B. Suggestions for AID Assistance.....................43
1. Planning Assistance to Agricultural
Research Activities ..........................43
2. Project Design ................................ 45
3. AID Management of Research Projects............46
4. AID Evaluation of its Assistance to Research...47


Appendixes

A. Literature Cited
B. Impact Evaluation Team Members and Working Group
Members
C. Summaries of Impact Evaluation Reports
D. Proceedings of a Workshop on Impact of Agricultural
Research










ACKNOWLEDGMENTS


This report summarizes the results of many activities
extending over several years. Many people have therefore con-
tributed to it, and it is my pleasure to acknowledge their
participation and assistance.

A special mention should be made of Dr. Twig Johnson (AID/
PPC/E/S) who coordinated the first part of this series and se-
lected projects for field evaluation. Dr. Richard Blue (AID/
PPC/E) and Ms. Charlotte Suggs (AID/PPC/E/S) were involved in
the entire series. Dr. Frank Byrnes (IADS) organized a work-
shop in 1982 on the Impact of Agricultural Research.

The members of our impact evaluation teams took four weeks
off from their busy schedules for a field evaluation, which is
always exhausting and demanding; they are listed in Appendix B,
together with the members of an intra-agency working group who
assisted in the preparation of the workshop.

Special thanks are also due to the following individuals
who gave detailed and thoughtful comments on an earlier draft
of this paper: Joan Atherton (AID/PPC/PDPR), Frank Byrnes and
Kenneth McDermott (IADS), Peter Oram (ISNAR), Vernon Ruttan
(University of Minnesota), and Ram Yadav (IFPRI).


Josette Murphy









SUMMARY


The U.S. Agency for International Development (AID) and
its predecessor agencies have assisted the less developed coun-
tries in establishing and strengthening agricultural research
systems for over 30 years, and the United States is a major
contributor to international agricultural research centers.
Current AID policy reaffirms the Agency's long commitment to
agricultural research, an activity to which AID devoted over
$140 million in 1981, nearly 20 percent of its appropriation
for agriculture, rural development, and nutrition.

Because agricultural research continues to be a priority
activity, the Bureau for Program and Policy Coordination,
through its evaluation program, was asked to examine what les-
sons were learned from AID's past experience and suggest how
they can be incorporated in the Agency's policy, planning, and
implementation activities. This assessment is based on an
analysis of available project evaluations, field evaluations by
interdisciplinary teams of the impact of AID-funded projects in
eight countries, and discussions among AID officers, host gov-
ernment officials, and experts from the agricultural research
community during a three-day workshop.

The USAID assistance to regional and national research
institutions has been found to be highly successful in training
researchers and establishing or expanding research facilities,
but the effectiveness and sustainability of research activities
have often been hampered by managerial insufficiencies and by
unfavorable government policies, as well as by an inadequate
awareness of conditions in farming households.

The key recommendations that emerge from these studies are
as follows:

1. Host Government Commitment and Support to Research Is
Essential.

A real, long-term commitment to agricultural research on
the part of the host government determines the sustainability
of a research project as well as (indirectly) the utilization
of research findings. Therefore, a continuous dialogue among
politicians, administrators, and researchers will greatly in-
crease the likelihood of adequate support to research. This is
more likely to occur if the potential benefits of research pro-
grams for the government are clearly demonstrated. Research
institutions will be more effective if their mandate and au-
thority are clearly defined and agreed upon with the host
government.










2. Technological Solutions Alone Cannot Solve Problems
Which Have Political, Economic, and Social Dimensions.

Government policies and infrastructure determine in part
whether farmers can and will adopt improved technology and
practices, and whether necessary support services will be in
place and effective when needed. Therefore, agricultural re-
search programs should be selected within a much broader rural
development policy and planning framework.

Technological changes can have negative as well as posi-
tive impacts on rural household incomes and well-being, and can
sharpen inequity among households if adoption is dependent upon
a resource which is unequally distributed.

3. Research Should Be Farmer-Oriented.

If research activities are to increase the productivity of
food producers, the program designers and researchers, as they
establish the research program, should be aware of and under-
stand the existing farming systems and local agroecological and
economic conditions, and the resources available to the farm-
ers. This requires that some of the research activities be
interdisciplinary and include on-farm research. It will be
essential to establish, maintain, and use a two-way information
system among researchers, extension service agents, and the
farmers. Official linkages and feedback mechanisms among in-
stitutions and government entities with responsibilities in
research, extension, and the provision of services to farmers
should be established; they should also be established with the
educational institutions which train the researchers and exten-
sion staff.

4. Inadequate Management of Limited Resources, Especially
a High Rate of Attrition Among Skilled Staff, Can
Undermine the Effectiveness and Sustainability of an
Otherwise Satisfactory Program.

Training skilled researchers has been found to be the most
successful component of many research projects, but the train-
ing provided should be adapted to the realistic needs and capa-
bilities of the country, in choice of discipline, level of edu-
cation, and timing of the training. Training provided under a
project should be scheduled to complement its technical assis-
tance. Returning trainees should be assured of satisfactory
material, professional incentives, and rewards comparable to
those offered to other public servants.

National research institutions should not function in iso-
lation, but should maintain an active network of information
exchange with other national institutions in comparable ecolog-
ical zones, as well as with international research institutions.










5. Coordination Among Researchers and Other Development
Actors, From Farmers to Politicians, Is the Key to
Success.

Most of the issues outlined above share a common solution:
coordination and information flow. A research.system will be
most effective if the many actors who influence its success
(defined as the generation of improved technology that is
adopted by farmers and increases food production and incomes in
the country) are involved in a network in which their needs are
identified and through which the interaction between different
sectors of development are as synergetic as possible.

In many countries, the main difficulty in activating such
a network will be cultural. If the food producers are not
recognized as full members of the network, it will remain
insufficient. If the administration is highly centralized, if
a top-down, authoritarian approach to management is maintained,
the exchange of information will be hampered. Donor institu-
tions are part of this network by the very act of deciding
which activities they will support.

The importance of coordination is not specific to re-
search, but it may need particular emphasis in research activi-
ties because of the frequent assumption that science functions
in a world separated from daily reality. The food problem
exists in the real world of the small farmers, in the real
world of imperfect economies, and that is where the success of
any research program is tested. The remarkable contributions
of research to food production have been amply demonstrated all
over the world. Researchers will meet the challenges ahead if
the political and administrative structures and systems in
which they function make it possible for them to do so.











Highlights of Conditions Found Favorable to Effective
Research Systems1


A. Host Government


1. The host government is committed to pro-
viding sufficient human and financial
resources for research activities.

2. Pricing mechanisms and other government
policies are conducive to expanding pro-
duction of crops being researched.

3. Complementary services (such as exten-
sion, inputs, marketing, credit, roads,
Irrigation) will be functioning when
needed for adoption of research results.
The private sector is allowed to partici-
pate in the provision of such services.

4. Research priorities are established as
part of a comprehensive development plan.

5. Coordlnaton is encouraged among research,
extension, services, and training insti-
tutions.


C. Research Institution


1. The institution benefits from stability
in its research staff (i.e., sufficient
incentives to keep them) and from the
presence of competent managers as well as
knowledgeable researchers.

2. Funding and research priorities remain
assured and stable over the duration of a
research program.

3. The research staff forms a multidiscipli-
nary team including social as well as
technical expertise.

4. Linkages are established and maintained
with other related in-country, regional,
and international research institutions.

5. The research institution exchanges infor-
mation with the extension services and
agricultural training Institutions.





This table is based upon the documents listed in
documents prepared by Tom Niblock and Rich Feely,


B. AID Assistance


1. Assistance to agricultural research in-
stitutions is designed as a long-term
activity, preferably 10 years, with op-
tion to redesign or extend on the basis
of regular evaluations.

2. Assistance is integrated into the entire
program of assistance to the country.

3. The AID in-country mission is capable of
providing the required logistic support
and problem-solving assistance to the
project contractor and the host country,
and includes staff members with knowledge
and understanding of agricultural devel-
opment and research issues.

4. AID assistance Is implemented through a
government entity which can coordinate
Its activities with those of related
institutions and programs.

5. Training programs are adapted to future
needs and scheduled to complement on-the-
job training with foreign technical
assistants.


D. Research Program


1. The overall research takes into account
existing farming conditions and the
natural, economic, and social conditions
that affect change. This does not mean
that basic research may not also be
necessary.

2. Base line data on actual farming prac-
tices and results are necessary both to
establish research priorities and program
design and to verify the results
achieved.

3. The expected research results should
clearly be worthwhile from the farming
household's point of view.

4. The research program should include on-
farm testing of results, possibly in
coordination with the extension service.

5. Correct promotion of research results to
the farmers should be assured.



Appendix A and upon the analysis of these
ADL, for PPC/Evaluation.










I. AGRICULTURAL RESEARCH AND DEVELOPMENT


A. Research, Food Production, and Population Growth


The Green Revolution has demonstrated that high-yielding
varieties of food crops and improved technology could lead to
increased productivity in the less developed countries (LDC).
The overall rate of increase in food production in the LDCs
from 1961 to 1976 averaged 2.6 percent per year. While this
is a remarkable achievement, in more than half these countries
the increase in food production has not kept pace with popula-
tion growth, so on balance the situation is worsening. This is
especially true in Africa, the only region with a net loss in
production per capital (see Table 1). Food production must now
increase by an average of at least 4 percent per year if con-
sumption needs are to be met by 1990.


Table 1. Agriculture Production Indices
Per Capita, 1970, 1975, and 1980
(1969-1971 = 100)



Country 1970 1975 1980


Africa 100 95 89
Latin America 100 103 108
Asia 101 105 107
Near East 98 104 101
World 100 103 104


Source: Food and Agriculture Organization, Production
Yearbook 1980.







1Bachman and Paulino, 1979, p. 13.


2Oram et al., IFPRI No. 10, 1979.










I. AGRICULTURAL RESEARCH AND DEVELOPMENT


A. Research, Food Production, and Population Growth


The Green Revolution has demonstrated that high-yielding
varieties of food crops and improved technology could lead to
increased productivity in the less developed countries (LDC).
The overall rate of increase in food production in the LDCs
from 1961 to 1976 averaged 2.6 percent per year. While this
is a remarkable achievement, in more than half these countries
the increase in food production has not kept pace with popula-
tion growth, so on balance the situation is worsening. This is
especially true in Africa, the only region with a net loss in
production per capital (see Table 1). Food production must now
increase by an average of at least 4 percent per year if con-
sumption needs are to be met by 1990.


Table 1. Agriculture Production Indices
Per Capita, 1970, 1975, and 1980
(1969-1971 = 100)



Country 1970 1975 1980


Africa 100 95 89
Latin America 100 103 108
Asia 101 105 107
Near East 98 104 101
World 100 103 104


Source: Food and Agriculture Organization, Production
Yearbook 1980.







1Bachman and Paulino, 1979, p. 13.


2Oram et al., IFPRI No. 10, 1979.






-2-


Furthermore, researchers have found that the generation,
adoption, and results of improved agricultural technology are
complicated by economic, political, social, and institutional
constaints both at the farm and the national levels. More
funds and more technical assistance do not necessarily solve
these problems, even if it were feasible to increase the
amounts involved in assistance to LDC research institutions.

The world's annual expenditure on agricultural research
now stands at $5,000 million, about double what it was in 1975
in constant 1975 terms, and about $1,600 million of that
amount is spent in the less developed countries. Oram and
Bindlish computed the amounts and distribution of expenditures
on agricultural research in 47 less developed countries, to-
gether with the total number of agricultural scientists in each
region (see Table 2). They point out that total expenditures
seem to have stagnated since 1978-1979. The trend begun in the
early 1970s may be changing, especially as most donor countries
face internal economic difficulties.

Much effort has been directed toward institution-building
and training at the national level, and an effective network of
international agricultural research centers has been estab-
lished. In the context of increased need, a well-established
research network, and possibly limited financial resources, it
behooves agricultural scientists and rural development special-
ists to learn from past experience so that future financial and
human investments in agricultural research are as productive as
possible.


B. AID's Assistance to Agricultural Research


USAID and its predecessor agencies have assisted agricul-
tural research in less developed countries for more than 30
years. During the 1950s, the emphasis was on transfer of
Western know-how, characterized by assistance to extension
services and training institutions, especially universities.
As evidence mounted that Western know-how was not always suc-
cessful in the agroeconomic context of most LDCs, the emphasis
shifted in the 1960s from extension to assisting national and
regional research institutions through training and technical
assistance, and by providing these institutions with adequate
facilities. During that period, the achievements of the Green
Revolution demonstrated that agricultural research that was



3World Bank, 1981, p. 16.


40ram and Bindlish, 1981, p. 81.








Table 2. Change in Expenditures on Agricultural Research and Numbers of
Agricultural Scientists for 47 Countries, 1971, 1975, and 1980


$ millions Percentage Number of Percentage
(constant 1975 terms) Change Scientists Change

Region1 1971 1975 1980 1971/75 1975/80 1971 1975 1980 1971/75 1975/80


SOuth Asia (5) 41.2 73.3 139.7 78 91 2,529 6,120 12,293 42 101

Southeast/East Asia (5) 28.0 46.7 101.0 67 116 2,285 4,400 5,830 95 31

N. Africa/Middle East (5) 21.9 21.9 35.1 -1 60 1,432 1,163 1,375 -21 18

West Africa (6) 41.8 86.5 112.5 107 30 915 3,239 1,897 154 -42

East/Southern Africa (5) 18.0 18.9 27.9 5 47 513 605 861 18 42

Central America/ 18.6 22.7 59.9 22 86 967 1,393 1,680 44 21
Caribbean (11)

South America (10) 110.1 160.4 342.8 46 214 4,100 5,291 5,939 29 12

Total (47) 279.8 430.4 818.9 54 90 12,741 22,251 29,875 74 33


1Figures in parentheses denote the number

Source: Oram and Bindlish, 1981.


of countries in each region.






-4-


focused on commodity improvement (e.g., breeding rice varieties
whose yields were highly responsive to nitrogen and water
application) could indeed lead to production breakthroughs in
the less developed countries.

Since the.1970s, U.S. assistance has focused on smallhold-
ers and landless farmers. The Foreign Assistance Act (Section
103A) specifically requires that AID-assisted agricultural re-
search programs be adapted to the needs of small food producers
and include on-farm testing. The current AID policy on Food
and Agricultural Development (AID Policy Paper, May 1982) reaf-
firms the Agency's long term-commitment to research, citing as
one of the major areas of commitment of U.S. assistance the
developmentt of] human resources and institutional capabili-
ties, especially to generate, adapt, and apply improved science
and technology for food and agricultural development" (emphasis
in text, p. 2).

The policy paper recognizes the need for long-term assis-
tance and the importance of training, institutional develop-
ment, and policies that encourage the small farmers and private
entrepreneurs to increase agricultural productivity in their
country. Specific recommendations for implementation of AID
policy are developed in the Agency's Strategy Paper for Food
and Agricultural Development.

In 1981, USAID allocated about 20 percent of its appropri-
ation for agriculture, rural development, and nutrition to
agricultural research (see Figure 1). The actual expenditures,
which fluctuated considerably over the last few years, have
ranged between 13 and 19 percent of all appropriations for
agriculture. The funds, which include a contribution to the
international agricultural research centers, are about equally
divided between centrally funded and regional bureau- and
mission-funded projects (i.e., projects coordinated directly by
the Science and Technology Bureau of AID/Washington and those
coordinated by the regional bureaus). Projects funded through
the Science and Technology Bureau are usually specific research
activities in a commodity sector, while projects funded through
the regional bureaus and missions usually focus on institution-
building and human resource development.

Funding levels for the regional bureaus are tending to
increase. Currently, 24 missions have included agricultural
research as an area of particular importance in their Country
Development Strategy Statements for 1983, and the Africa and
Asia Bureaus have given clear priority to agricultural research
for their future programs. The Asia Bureau, which has a long
history of agricultural research activities, is conducting a
review of its past experience in agricultural research (Asian
Agricultural Research Review); results available to date are
presented in Section II. The Africa Bureau, conscious of the






-5-


Figure 1. U.S. Agency for International Development,
Agricultural Research Appropriation, 1978-1981
(in thousand U.S.$)

Amount Appropriated
(in thousand of
U.S. dollars)


150 -






125






100 -






75 -






50 -






25 -






A


AFR


AS IA-


LAC


NE





S&T*


1978


AFR




ASIA



LAC


NE







S&T*


1979


AFR




ASIA
LAC
NE






S&T*


1980


*Includes U.S. contributions to the international research centers,
indicated by -


AFR


ASIA




LAC


NE


S&T*


1981









particularly difficult situation in African nations, has re-
fined its strategy for agricultural research to incorporate
some of the lessons from experience which are substantiated in
this document, in particular, the AID long-term commitment to
strengthening national agricultural research institutions; the
need for coordination and feedback among scientists at the
regional level; the advantages of farming systems approach to
research; the importance of linkages between research, exten-
sion, the farmers, and education activities; and the necessity
of providing support services to the farmers.


II. PAST EXPERIENCE: THE EVIDENCE


AID activities in agricultural research can be documented
through the routine evaluations conducted for each project and
through special studies and evaluations conducted for projects
or programs of particular interest to AID. Examples of these
are the series of eight impact evaluations coordinated by the
Bureau for Program and Policy Coordination, Office of Evalua-
tion, Studies Division (PPC/E/S), as part of this review of AID
experience, and an in-depth evaluation of selected Asian
countries coordinated by Professor Vernon Ruttan for the Asia
Bureau.

AID officers and contractors have also accumulated much
experience and wisdom which is not recorded or published.
Discussions among evaluation teams, members of the Intra-Agency
Agricultural Research Working Group, and the participants to
the Workshop on Impact of Agricultural Research have been
incorporated throughout this report.

The evidence from past experience will be summarized in
this section as the basis for the discussion of key issues and
lessons learned for agricultural research activities which is
developed in Section III.


A. The Findings of Eight Impact Evaluations


1. Evaluation Methodology


In addition to a comparative analysis of existing evalua-
tion documents for all completed AID projects, eight projects
were selected for field evaluations. The decision was made to
limit the evaluations, for the time being, to projects funded
through AID's missions and regional bureaus: two in Africa,
three in Asia, two in Latin America, and one in the Near East.
The projects provided some form of assistance to national









particularly difficult situation in African nations, has re-
fined its strategy for agricultural research to incorporate
some of the lessons from experience which are substantiated in
this document, in particular, the AID long-term commitment to
strengthening national agricultural research institutions; the
need for coordination and feedback among scientists at the
regional level; the advantages of farming systems approach to
research; the importance of linkages between research, exten-
sion, the farmers, and education activities; and the necessity
of providing support services to the farmers.


II. PAST EXPERIENCE: THE EVIDENCE


AID activities in agricultural research can be documented
through the routine evaluations conducted for each project and
through special studies and evaluations conducted for projects
or programs of particular interest to AID. Examples of these
are the series of eight impact evaluations coordinated by the
Bureau for Program and Policy Coordination, Office of Evalua-
tion, Studies Division (PPC/E/S), as part of this review of AID
experience, and an in-depth evaluation of selected Asian
countries coordinated by Professor Vernon Ruttan for the Asia
Bureau.

AID officers and contractors have also accumulated much
experience and wisdom which is not recorded or published.
Discussions among evaluation teams, members of the Intra-Agency
Agricultural Research Working Group, and the participants to
the Workshop on Impact of Agricultural Research have been
incorporated throughout this report.

The evidence from past experience will be summarized in
this section as the basis for the discussion of key issues and
lessons learned for agricultural research activities which is
developed in Section III.


A. The Findings of Eight Impact Evaluations


1. Evaluation Methodology


In addition to a comparative analysis of existing evalua-
tion documents for all completed AID projects, eight projects
were selected for field evaluations. The decision was made to
limit the evaluations, for the time being, to projects funded
through AID's missions and regional bureaus: two in Africa,
three in Asia, two in Latin America, and one in the Near East.
The projects provided some form of assistance to national









particularly difficult situation in African nations, has re-
fined its strategy for agricultural research to incorporate
some of the lessons from experience which are substantiated in
this document, in particular, the AID long-term commitment to
strengthening national agricultural research institutions; the
need for coordination and feedback among scientists at the
regional level; the advantages of farming systems approach to
research; the importance of linkages between research, exten-
sion, the farmers, and education activities; and the necessity
of providing support services to the farmers.


II. PAST EXPERIENCE: THE EVIDENCE


AID activities in agricultural research can be documented
through the routine evaluations conducted for each project and
through special studies and evaluations conducted for projects
or programs of particular interest to AID. Examples of these
are the series of eight impact evaluations coordinated by the
Bureau for Program and Policy Coordination, Office of Evalua-
tion, Studies Division (PPC/E/S), as part of this review of AID
experience, and an in-depth evaluation of selected Asian
countries coordinated by Professor Vernon Ruttan for the Asia
Bureau.

AID officers and contractors have also accumulated much
experience and wisdom which is not recorded or published.
Discussions among evaluation teams, members of the Intra-Agency
Agricultural Research Working Group, and the participants to
the Workshop on Impact of Agricultural Research have been
incorporated throughout this report.

The evidence from past experience will be summarized in
this section as the basis for the discussion of key issues and
lessons learned for agricultural research activities which is
developed in Section III.


A. The Findings of Eight Impact Evaluations


1. Evaluation Methodology


In addition to a comparative analysis of existing evalua-
tion documents for all completed AID projects, eight projects
were selected for field evaluations. The decision was made to
limit the evaluations, for the time being, to projects funded
through AID's missions and regional bureaus: two in Africa,
three in Asia, two in Latin America, and one in the Near East.
The projects provided some form of assistance to national










(five) or regional (three) institutions, and all except one
(Guatemala) had been completed prior to the impact evaluation.
However, AID has continued to assist some of the institutions
after the projects evaluated here had ended.

The basic characteristics of each project (compiled from
the impact evaluation reports) are listed in Table 3. For ease
of presentation, each project will be referred to by its loca-
tion. Each project was evaluated by an interdisciplinary team
(see list in Appendix B) during a visit of about four weeks.
Agriculturalists, economists, social scientists, and develop-
ment generalists were present, with each team including one or
more AID officers. Outside consultants joined the teams where
the necessary expertise was not available within AID at the
time of the evaluation. Every team included members with pre-
vious experience in the country and with knowledge of a local
language.

To assess the impact of the project, each team interviewed
a sample of farmers as well as researchers and administrators,
spent a minimal time in the capital city, and traveled in rural
areas. The main goals of each evaluation were as follows:

To determine whether the institution that had received
assistance was functioning and whether the researchers
who had received training were active in research

-To assess the quality of the research program and its
applicability in actual farming conditions

-To determine the extent to which research findings
have been adopted by farmers, how food producers have
been affected by the new technology, and why

Each team had a common list of topics to cover as a frame-
work for its inquiry, but individual scopes of work were drawn
up because of the great diversity of project strategies.


2. Kenya


The Crop Production and Research Project, the starting
point of this impact evaluation, was only one among many activ-
ities funded by USAID and other donors which led to the breed-
ing and dissemination of hybrid maize lines in East Africa.
The first hybrid, bred at the Kitale Research Station in 1964,
produced a 40 percent increase in yields. It has been widely
adopted by both large and small farmers, in spite of the fact
that seeds need to be purchased every year, because no other
changes in practices were necessary to obtain a significant
increase in production.








Table 3. Characteristics of Eight AID Project Impact Evaluations


Title and No.
Project Funding Implementation Institutions Date of of Evaluation
Location Program Title (in millions) Dates Assisted Evaluation Report


Kenya




Central
America


Crop Production and
Research (618-0644,
618-0657)

Small-Farm Cropping
Systems (596-0064)


Guatemala Food Productivity and
Nutrition Improvement
(520-11-130-232)


Korea





Nepal





Thailand,
Northeast
Region


Tunisia


,West Africa


Agricultural Research
Project (DLC/P-2014,
489-11-088)


Food Grain Technology:
Agricultural'Research
in Nepal (367-11-110-054,
367-0054)

Agricultural Development,
Agricultural Research
(493-11-190-180.2)

Accelerated Cereals
Production (654-0205.1)
and related regional
projects (698-0173)

West Africa Rice Devel-
opment Association: Rice
Research and Development
(698-11-190-382, 698-0382)


$2.2


AID grant,
$1.633


AID, $1.7
(plus $1.0 in
earlier projects)


Loan, $5.0
Korean contribution,
$3.124


About $20.0 total





AID, $6.272
Thai Government,
$6.8


$1.715


AID, $5.166
WARDA, $0.3
(in kind)


1969-1981 East African
Community


1975-1979 Center for Trop-
ical Agriculture
Research and
Training (CATIE)

1975-1979 Institute of Agric-
cultural Science and
Technology (ICTA)


1974-1980 Office of Rural
Development,
Ministry of Agricul-
ture and Fisheries

1957-1974 Ministry of Food and
Agriculture, with
assistance to five
research stations

1966-1975 Thai Phra Agricul-
tural Research Center


1967-1977 Office of Cereals


1975-1980
(first phase)


West Africa Rice
Development Asso-
ciation (WARDA)


December 1979




February 1980


October 1979


January 1982





January 1982





February 1981


April 1982


October 1981


Kitale Maize: The
Limits of Success
(No. 2)

Central America: Small
Farmer Cropping System
(No. 14)


Guatemala: Development
of the Institute of Agri
cultural Science and
Technology and its Impact
on Agricultural Research
and Farm Productivity
(No. 27)

Korean Agricultural Re-
search: The Integration
of Research and Extension
(No. 30)

Food Grain Technology:
Agricultural Research
in Nepal (No. 33)


Agricultural Research in
Northeastern Thailand
(No. 34)

Tunisia: The Wheat
Development Program
(in preparation)


West Africa Rice Re-
search and Development
(No. 44)






-9-


As a result, maize production increased and Kenya has come
close to self-sufficiency in this staple food. The report
states that production increase could have been higher, how-
ever, had the Kenyan Government strengthened its marketing and
storage infrastructure to handle crop surpluses, and had nec-
essary inputs, especially credit and fertilizer, been available
in sufficient quantities. The private sector did play a cru-
cial role in the rapid dissemination of the hybrid; the Kenya
Seed Company assured seed multiplication and distribution,
while shopkeepers actively promoted the new hyrbrid.

As hybrid maize became more widely adopted, AID attempted
to assist the East African Agriculture and Forest Organization
in developing a research institution capable of coordinating
varietal trials of hybrid maize and other crops and of dissemi-
nating the results among African scientists in the East African
Community.5 This effort resulted in the identification of
improved hybrid maize with better potential than the original
one, but other technical components of the projects were not
successful. They included developing varieties suitable for
low rainfall areas and improving maize protein quality, a topic
which has been found especially difficult elsewhere.

As institution-building programs, the three projects were
failures, not simply because of the break-up of the East African
Community, but because from the beginning, each country had
avoided sending scientists to the regional institution. The
national research programs did not receive sufficient support
from their governments either, so the few African scientists
were not encouraged to stay with the research institution. Few
Africans were trained under the projects, and at the time of
the impact evaluation (1979) the breeding program had all but
died out after departure of the last American breeder.

Thus, the Kenya report presents the case of research
results being widely adopted, in spite of some unfavorable
nationwide economic conditions, but little permanent research
capacity remaining after some 15 years of technical assistance
with limited training.








5This was done under the Annual and Crop Production (618-11-
110-644) and Major Cereals and Legume Improvement (618-11-130-
652) projects from 1969 to 1972, and under the East African
Community Food Crop Research (618-11-110-657) from 1972 to 1977.






-10-


3. Central America


The Small Farmer Cropping System project was implemented
from 1975 to 1979 through the Center for Tropical Agricultural
Research Training (CATIE) so that scientists from CATIE could
"develop and demonstrate an innovative multidisciplinary metho-
dology for doing research on the cropping systems of the small
farmers of Central America." Both institutional and technical
results were expected from this project: development of a
regional institution capable of coordinating on-farm research
and training programs well adapted to the needs of the small
farmers, and some improved cropping systems adapted to various
ecological zones in the region, which could then be tested and
promoted by the national institutions.

The impact evaluation was conducted in 1980. At that
time, it was evident that the program was successful in devel-
oping methodologies for on-farm cropping systems research, but
only one set of recommendations had yet been verified on a
large scale before dissemination. The expected institutional
results had been reached, with CATIE providing the necessary
training program and coordination with national institutions.

The evaluation report discusses two sets of problems: the
division of labor between the regional institution and its
national counterparts, and the importance of socioeconomic
factors. The project called for the regional organization to
survey traditional practices and identify improved cropping
systems, which the national institutions would then verify and
disseminate. The team found this division arbitrary, as both
the regional and national institutions would benefit from coop-
eration. The team also emphasized the importance of taking
into account socioeconomic factors in planning and implementing
both research and extension programs. This requires a fully
multidisciplinary effort, with social scientists and farm man-
agement specialists, as well as agriculturalists.

A recent (April 1983) evaluation of current AID assistance
to CATIE confirmed that farm-level research is still on-going
in spite of the political instability in the region. It does,
however, confirm the impact evaluation concern about the very
limited outreach program through the extension services in the
national systems.


4. Korea


The Korea Agricultural Research project (1974-1980) aimed
at strengthening an existing research capacity within the
Ministry of Agriculture and Fisheries with a $5.0 million loan.






-10-


3. Central America


The Small Farmer Cropping System project was implemented
from 1975 to 1979 through the Center for Tropical Agricultural
Research Training (CATIE) so that scientists from CATIE could
"develop and demonstrate an innovative multidisciplinary metho-
dology for doing research on the cropping systems of the small
farmers of Central America." Both institutional and technical
results were expected from this project: development of a
regional institution capable of coordinating on-farm research
and training programs well adapted to the needs of the small
farmers, and some improved cropping systems adapted to various
ecological zones in the region, which could then be tested and
promoted by the national institutions.

The impact evaluation was conducted in 1980. At that
time, it was evident that the program was successful in devel-
oping methodologies for on-farm cropping systems research, but
only one set of recommendations had yet been verified on a
large scale before dissemination. The expected institutional
results had been reached, with CATIE providing the necessary
training program and coordination with national institutions.

The evaluation report discusses two sets of problems: the
division of labor between the regional institution and its
national counterparts, and the importance of socioeconomic
factors. The project called for the regional organization to
survey traditional practices and identify improved cropping
systems, which the national institutions would then verify and
disseminate. The team found this division arbitrary, as both
the regional and national institutions would benefit from coop-
eration. The team also emphasized the importance of taking
into account socioeconomic factors in planning and implementing
both research and extension programs. This requires a fully
multidisciplinary effort, with social scientists and farm man-
agement specialists, as well as agriculturalists.

A recent (April 1983) evaluation of current AID assistance
to CATIE confirmed that farm-level research is still on-going
in spite of the political instability in the region. It does,
however, confirm the impact evaluation concern about the very
limited outreach program through the extension services in the
national systems.


4. Korea


The Korea Agricultural Research project (1974-1980) aimed
at strengthening an existing research capacity within the
Ministry of Agriculture and Fisheries with a $5.0 million loan.






-11-


It included training, some equipment, and technical assistance,
but no new infrastructure. The training component was found to
be the most useful, and long-term technical assistance the
least. The spread of high-yielding rice varieties has been
very rapid in Korea, largely because of the effectiveness of
the extension service and the hierarchical social tradition.
By 1975, Korea reached self-sufficiency in rice. This success
backfired in 1979 and 1980, however, when disease followed by
unfavorable temperatures greatly lowered production. The re-
port questions the wisdom of relying heavily on any one variety
of a staple crop and points out that this vulnerability stemmed
in part from the very strength of the extension service, which
assured a rapid diffusion of the Tangil variety before cold-
resistant varieties could be developed. The report also raises
questions regarding the choice of other crops for research
activities, even though some of these crops (wheat, soybeans,
and potatoes) are not economically well adapted to the farming
community, partly because of the price structure.


5. Guatemala


The Food Productivity and Nutrition Improvement project is
one of several projects since 1970 that has provided assistance
to the Institute of Agricultural Science and Technology (ICTA)
in Guatemala. This project had both institutional and techni-
cal objectives.

The project has been successful in strengthening the re-
search capacity of ICTA through training and technical assis-
tance. With the assistance of experts from international
agricultural research centers and support from the Inter-
American Development Bank and the Rockefeller Foundation, as
well as that of USAID, ICTA has developed new varieties of
maize, beans, and sorghum and has tested them under farm condi-
tions with the participation of local farmers. Improved farm-
ing practices have been identified, and a seed service has been
organized which provides a regular supply of good quality,
improved seeds.

ICTA is an unusual institution among those evaluated in
this series because it has responsibility for research and for
determiningn] farmer acceptance or nonacceptance by introduc-
ing these new technologies to farmers directly and inorpor-
ating farmers evaluations into the research effort." The
existing extension service within the Ministry of Agriculture
retains responsibility for large-scale dissemination. Research


16Report No. 30, p. 4.






-12-


personnel work in close cooperation with collaborating farmers
to test new varieties and farming practices in real farm condi-
tions, and to draw upon the farmers' knowledge in identifying
possible areas for improvement.

The evaluation report notes the success of the ICTA ap-
proach, stating that "ICTA has come to represent a new model
for agricultural research that planners and researchers in
other countries are studying and attempting to replicate. If
there is continued and increased support from the Government of
Guatemala, i will be able to sustain and expand its present
activities." This note of caution stems from some problems
caused by the Government organizational structure. Researchers
are penalized under the existing salary schedule, resulting in
a high attrition rate among ICTA personnel.

As could be expected from its very mandate, conflicts have
arisen between ICTA and the extension service, since the divi-
sion of labor between the two is unclear. At the time of the
evaluation, the two institutions were discussing a more coordi-
nated approach to their activities.


6. Nepal


The Food Grain Technology project in Nepal was the longest
among those evaluated, lasting from 1957 to 1974 (a follow-on
project is still being implemented in 1983). This project also
had the largest budget (about $20.0 million total); it included
training, commodities, infrastructure, and technical assis-
tance.

While the project goal remained that of increasing produc-
tion by promoting improved farm technology, the project design
was flexible, and project activities shifted over time from
general agricultural development, to the development of im-
proved technology for food grains, and finally to a combination
of development of new technology and some coordination with the
extension service.

The project has had sustainable results: 600 Nepalese
have been trained, five research stations have been built and
expanded, and a research system has been put in place and is
functioning. The research stations are specialized by commod-
ity. The rate of adoption of new technology has been high,
leading especially to an increase in cropping intensity (from
one to two and sometimes three crops a year in part of the


Report No. 30, p. viii.





-13-


southern plain of Nepal) and a dramatic increase in wheat cul-
tivation. This has not been accompanied by a significant in-
crease in yields, however, and improved rice varieties were
used on only about 25 percent of acreage in 1980. The report
raises questions regarding the equity impact of improved tech-
nology, which depends heavily on the availability of irrigation,
and about long-term effects on soil fertility. The technical
package calls for chemical fertilizer. Many farmers are reluc-
tant to use the recommended levels, because uncertainties in
water supply make the levels economically risky. On the other
hand, the higher cropping intensity has had a negative effect
on the size of the herds kept by these farmers, and therefore,
on the amount of manure available to them.


7. Thailand


The Agricultural Development-Agricultural Research project
assisted the already existing Tha Phra Research Center in
Northeastern Thailand from 1966 to 1975. The AID project in-
cluded training in the United States for 118 Ministry employees,
constructing and supplying equipment for research laboratories,
and establishing research programs and extension activities.

The original mandate of "the Center was to be a multidis-
ciplinary research facility focusing on the Northestern region
and responsive to the needs of the farmers. In addition, it
was to support and coordinate the work of the Ministry's [of
Agriculture and Cooperatives] 112 small research centers and
stations in Northeastern Thailand."

The project was successfully implemented, and "by 1975,
laboratories were well established, and substantial research
work was underway." Since then, the team found that although
an innovative extension and training program is now active, on
the whole, the research role of the Center has not been as ef-
fective as expected, mainly because of bureaucratic constraints.

Part of the difficulties are due to "bureaucratic conflict"
between the Center and the Ministry of Agriculture, which dis-
agree on research programming, and to several changes in the
mandate of the Center, with more emphasis on planning and co-
ordinating the work of the regional Ministry Agency and imple-
menting development projects. The activities of the Center are
further hampered by its insufficient budget and by staffing
difficulties.


8Report No. 34, p. iii.






-14-


This project is an example of one that successfully
strengthens a research institution, providing it with adequate
facilities and staffing, but whose long-term impact has been
lowered because the institution's role was later modified by
its ministerial authorities.


8. Tunisia


The Accelerated Cereals Production program had a dual
technical and institution-building purpose. It was a long-term
(1967-1977) effort, funded by USAID, the Ford and Rockefeller
Foundations, the International Maize and Wheat Improvement
Center (CIMMYT) in Mexico, and the Government of Tunisia.

The program was designed shortly after independence, at a
time when the Government of Tunisia needed both to establish
its own research and extension activities and to reverse the
decline in food production due to the departure of the French
farmers. The Wheat Development program proposed to adopt the
new semi-dwarf, high-yielding wheat varieties recently devel-
oped at CIMMYT and to establish a Tunisian research institution
in the process. Both objectives have been reached; five years
after the end of the project, the evaluation team found a suc-
cessfully operating Tunisian research institution and wide-
spread use of an improved wheat technology that resulted in
increased yield and production. The evaluation report points
out that much of the positive impact of the project became evi-
dent only after the project itself had ended, showing that a
long-term perspective is essential when assessing the impact of
a research project.

The report points out that the choice of a semi-autonomous
institution for implementation of the project gave project de-
signers and implementors freedom from some of the bureaucratic
constraints of the Ministry of Agriculture, and also separated
the research function from that of extension. Good cooperation
between individuals in the two entities has enabled an ad hoc
coordination between research and extension, but it is not in-
stitutionalized and therefore remains vulnerable.


9. West Africa


The West Africa Rice Development Association (WARDA),
operating in 15 countries, was created in 1970 to adapt the
improved technology developed for Asian rice production to West
African agroecological conditions. WARDA is assisted by many
donors; AID assistance was at first focused on training and on
adaptive research for mangrove rice (in Sierra Leone) and for






-14-


This project is an example of one that successfully
strengthens a research institution, providing it with adequate
facilities and staffing, but whose long-term impact has been
lowered because the institution's role was later modified by
its ministerial authorities.


8. Tunisia


The Accelerated Cereals Production program had a dual
technical and institution-building purpose. It was a long-term
(1967-1977) effort, funded by USAID, the Ford and Rockefeller
Foundations, the International Maize and Wheat Improvement
Center (CIMMYT) in Mexico, and the Government of Tunisia.

The program was designed shortly after independence, at a
time when the Government of Tunisia needed both to establish
its own research and extension activities and to reverse the
decline in food production due to the departure of the French
farmers. The Wheat Development program proposed to adopt the
new semi-dwarf, high-yielding wheat varieties recently devel-
oped at CIMMYT and to establish a Tunisian research institution
in the process. Both objectives have been reached; five years
after the end of the project, the evaluation team found a suc-
cessfully operating Tunisian research institution and wide-
spread use of an improved wheat technology that resulted in
increased yield and production. The evaluation report points
out that much of the positive impact of the project became evi-
dent only after the project itself had ended, showing that a
long-term perspective is essential when assessing the impact of
a research project.

The report points out that the choice of a semi-autonomous
institution for implementation of the project gave project de-
signers and implementors freedom from some of the bureaucratic
constraints of the Ministry of Agriculture, and also separated
the research function from that of extension. Good cooperation
between individuals in the two entities has enabled an ad hoc
coordination between research and extension, but it is not in-
stitutionalized and therefore remains vulnerable.


9. West Africa


The West Africa Rice Development Association (WARDA),
operating in 15 countries, was created in 1970 to adapt the
improved technology developed for Asian rice production to West
African agroecological conditions. WARDA is assisted by many
donors; AID assistance was at first focused on training and on
adaptive research for mangrove rice (in Sierra Leone) and for






-15-


deepwater rice (in Mali). In a second phase of assistance, AID
is now working with WARDA to develop its analytical capacity to
identify problems and to make research suggestions for the
countries involved.

The mandate of WARDA, which is stated predominantly in
technical terms, has been found too restrictive to address a
problem which is economical as well as technical. Varieties of
rice adapted to local conditions do not guaranty an increase in
production if pricing regulations make rice production unfavor-
able in the first place. This is partly being corrected, be-
cause WARDA "on account of its scientific professionalism .
has discovered a politically acceptable way of targeting
project identification and research design on specific
situations that are not only ecologically but economically
conducive to expanded rice production."

The research projects under the first phase have had mixed
results, but the training (rice production course) has been
found very useful (the U.S. training component is not yet
completed).

The evaluation report discusses the pros and cons of a re-
gional research entity, a topic of crucial importance at this
time, as African and donor countries are planning long-term
research activities in a coordinated fashion.


B. Other Evaluations and Studies


1. Review of Routine Project Evaluations


All AID projects are normally evaluated during implementa-
tion and after the end of the project. A comparative analysis
of these evaluations for 48 agricultural and research projects
identified a number of recurrent problems at three levels:
selection of research topics, implementation and management of
the project, and difficulties because of inadequacies in re-
lated support and services. The results of this analysis were
not as definitive as expected, because it was found that rou-
tine evaluations in early years were often uneven in scope and
quality and thus difficult to compare in a systematic manner.
The review found that while most projects were supposed to
focus on research activities that benefit small farmers, not
all of the evaluations even considered whether this was actu-
ally the case during implementation. Of those evaluations that


Report No. 44, p. viii.






-15-


deepwater rice (in Mali). In a second phase of assistance, AID
is now working with WARDA to develop its analytical capacity to
identify problems and to make research suggestions for the
countries involved.

The mandate of WARDA, which is stated predominantly in
technical terms, has been found too restrictive to address a
problem which is economical as well as technical. Varieties of
rice adapted to local conditions do not guaranty an increase in
production if pricing regulations make rice production unfavor-
able in the first place. This is partly being corrected, be-
cause WARDA "on account of its scientific professionalism .
has discovered a politically acceptable way of targeting
project identification and research design on specific
situations that are not only ecologically but economically
conducive to expanded rice production."

The research projects under the first phase have had mixed
results, but the training (rice production course) has been
found very useful (the U.S. training component is not yet
completed).

The evaluation report discusses the pros and cons of a re-
gional research entity, a topic of crucial importance at this
time, as African and donor countries are planning long-term
research activities in a coordinated fashion.


B. Other Evaluations and Studies


1. Review of Routine Project Evaluations


All AID projects are normally evaluated during implementa-
tion and after the end of the project. A comparative analysis
of these evaluations for 48 agricultural and research projects
identified a number of recurrent problems at three levels:
selection of research topics, implementation and management of
the project, and difficulties because of inadequacies in re-
lated support and services. The results of this analysis were
not as definitive as expected, because it was found that rou-
tine evaluations in early years were often uneven in scope and
quality and thus difficult to compare in a systematic manner.
The review found that while most projects were supposed to
focus on research activities that benefit small farmers, not
all of the evaluations even considered whether this was actu-
ally the case during implementation. Of those evaluations that


Report No. 44, p. viii.






-16-


did, problems were identified in setting clear research priori-
ties, implementing multidisciplinary research activities, and
conducting on-farm testing, even though these three factors
were found to relate positively to a (subjective) assessment of
"better than" or "satisfactory project performance" by the
evaluations.

Finally, the routine evaluations of agricultural research
projects (as indeed those of most projects) manifested many
managerial problems during implementation. Difficulties arose
with AID contractors and host government personnel with almost
perfect regularity. While many of these difficulties are not
specific to agricultural research projects, several character-
istics of these projects make them especially vulnerable to
management problems: they involve procurement of large amounts
of equipment, they involve high-level training (usually in the
United States), and they usually involve long-term programs
which cannot show concrete results during the life of an AID
project.

The first two characteristics, not surprisingly, lead to
frequent difficulties with delays and resulting scheduling
problems: delays in construction and procurement that hamper
research activities; delays in identifying and preparing can-
didates for overseas training; and discrepancies in scheduling
the training of host country nationals to coincide with the
presence of expatriate technical assistance, so that all too
often, the technical assistant runs the program while his
"counterpart" is overseas, with little, if any, overlap upon
the trainee's return for on-the-job training.

It is important to note that the only factor which was
considered by almost all the project evaluations and found
positively related to good project performance was the host
government support to agricultural research activities, as
reflected in the government allocation of funds and staff, in
policies that influence the food producers, and in the flexi-
bility and control over its own activities given to the
research institution.

Finally, problems with the performance of the implementing
contractor are not unusual, with difficulties in identifying
qualified experts and delays in fielding the most often cited.

2. Conference on Impact of Agricultural Research, Leesburg,
Virginia

AID/PPC/E organized a conference near Leesburg, Virginia,
from June 13 to 17, 1982, on the impact of agricultural re-
search. More than 100 participants from 32 countries discussed
the findings of the impact evaluations in the context of their
own experience and knowledge. The participants (listed in






-17-


Appendix D) included officers from AID/Washington and 24 over-
seas missions, host government officials, and representatives
of-donor and research institutions. The key findings and sug-
gestions are presented in Section III.


3. The Asia Agricultural Research Review Project


The Asia Bureau of USAID, seeking to measure the relation-
ship between USAID assistance to national research systems and
changes in agricultural productivity, is funding a review of
its past activities in selected countries, conducted by the
University of Minnesota under the leadership of Professor
Vernon W. Ruttan. Through the work of the Minnesota team and
its collaboration with Yale and Cornell Universities, and the
East-West Center, the study will provide an assessment of the
contribution of AID research investigation to agricultural
productivity and its impact on equity at the farm and regional
levels, in quantitative terms whenever possible. This is under
way for the Philippines, Indonesia, Bangladesh, South Korea,
and India.

The Minnesota work highlights10 the importance of the
research institution's location in the administrative structure
of the country, as well as that of coordination among institu-
tions and entities involved in the generation and diffusion of
research findings, two points which are also well illustrated
in the Thailand and the Korea impact evaluations.

The report also cites the "lack of information and a nly-
sis that goes into establishment of research priorities," a
point that is recurrent among the PPC/E impact evaluations.
Ruttan mentions, for example, the Bangladesh Rice Research
Institute's goal of developing improved varieties of deepwater
rice yielding one ton per hectare, an objective which turned
out to be about half of what the farmers were already produc-
ing. A similar lack of knowledge about existing farm condi-
tions had been identified in the Korea impact evaluation.

The review of "te income distribution effect of the Green
Revolution in India" shows that while improved technology can


10University of Minnesota Economic Development Center, Bulletin
No. 81-2, March 1981, p. 12 ff.

llIbid, p. 16.

12Economic Development Center Bulletin No. 82-5, April 1982,
p. 37.






-18-


be adopted by small farmers as well as by the larger ones, the
diffusion of high-yielding varieties is "closely interlocked
with the nature and level of their [the farmers' region] devel-
opment in physical and institutional infrastructure." The in-
teraction among agricultural, social, and economic constraints,
and the danger of planning research in isolation from its con-
text at the farm and national levels are recurrent throughout
the impact evaluations.

Finally, the Minnesota case studies, like the impact eval-
uations, are constantly citing managerial problems and, especi-
ally, the high rate of attrition among skilled research staff
because of insufficient material and professional rewards.


III. FROM PAST EXPERIENCE TO LESSONS LEARNED


A. Introduction


In the past, many strategies have been followed for gener-
ating research results that will lead to increases in food
production. Among the impact evaluations alone, some projects
worked through regional institutions (WARDA in West Africa,
CATIE in Central America, the East African Community), others
through a national ministry (Thailand, Nepal, Korea) or through
a parastatal institution (Guatemala, Tunisia). All these proj-
ects had the dual objective (albeit not always clearly ex-
pressed) of technology transfer and institutional development,
with the basic assumption that a host institution can be cre-
ated or strengthened in a sort of "on-the-job institutional
training" as Western research technology is being introduced.

A major lesson learned from these evaluations and from the
workshop discussions--one that will permeate this section--is
that the key difficulties in increasing food production are not
solely agricultural or technical, but lie in political, socio-
economic, and managerial constraints that influence the re-
search system on the one hand, and the adoption of research
findings on the other.

Technology transfer alone is not sufficient to assure food
security and increase food availability per capital. The LDCs
need an effective network of regional, national, and interna-
tional institutions and must be willing and able to revise
their policies to encourage increases in food production.






-18-


be adopted by small farmers as well as by the larger ones, the
diffusion of high-yielding varieties is "closely interlocked
with the nature and level of their [the farmers' region] devel-
opment in physical and institutional infrastructure." The in-
teraction among agricultural, social, and economic constraints,
and the danger of planning research in isolation from its con-
text at the farm and national levels are recurrent throughout
the impact evaluations.

Finally, the Minnesota case studies, like the impact eval-
uations, are constantly citing managerial problems and, especi-
ally, the high rate of attrition among skilled research staff
because of insufficient material and professional rewards.


III. FROM PAST EXPERIENCE TO LESSONS LEARNED


A. Introduction


In the past, many strategies have been followed for gener-
ating research results that will lead to increases in food
production. Among the impact evaluations alone, some projects
worked through regional institutions (WARDA in West Africa,
CATIE in Central America, the East African Community), others
through a national ministry (Thailand, Nepal, Korea) or through
a parastatal institution (Guatemala, Tunisia). All these proj-
ects had the dual objective (albeit not always clearly ex-
pressed) of technology transfer and institutional development,
with the basic assumption that a host institution can be cre-
ated or strengthened in a sort of "on-the-job institutional
training" as Western research technology is being introduced.

A major lesson learned from these evaluations and from the
workshop discussions--one that will permeate this section--is
that the key difficulties in increasing food production are not
solely agricultural or technical, but lie in political, socio-
economic, and managerial constraints that influence the re-
search system on the one hand, and the adoption of research
findings on the other.

Technology transfer alone is not sufficient to assure food
security and increase food availability per capital. The LDCs
need an effective network of regional, national, and interna-
tional institutions and must be willing and able to revise
their policies to encourage increases in food production.






-19-


The purpose of much investment in this area is to develop
a research capacity in a country by strengthening existing in-
stitutions or by creating new ones, so that the ultimate goal
of increased food production can be reached. What matters when
identifying and planning a development program is to understand
that the research capacity in a country is not a simple sum of
well-trained researchers, adequate buildings, and well-equipped
laboratories. These are means, not ends. The research capac-
ity in a country depends upon how well these means can be made
to function and fulfill the mandate of providing farmers with
tools (improved practices and technology) that can lead to in-
creased food production, and whether the political, economic,
and social environments (at national and local levels) allow
these means to become effective.

While research can provide the required technology im-
provements, a research program will be more effective if it is
not planned in isolation, but as part of the political, social,
and economic system that it must serve. Assistance to agricul-
tural research must take into account necessary linkages
between a research capacity--the macropolicy and the institu-
tional environment in which research institutions function--and
the farming community that research is to assist. A focus on
research institution-building is not likely to be sufficient.


B. Research Should Be Oriented Towards Farmers' Needs and
Constraints


The impact of agricultural research on food production is
ultimately decided not by researchers but by the farmers them-
selves, who decide on their farming practices for each crop
season. As a background to the following discussion, it is
prudent to first review the various factors that the farmers
integrate when reaching a decision about the package of inputs
and practices they will use in a given crop season. Research-
ers must be aware of these constraints in order to identify
improvements that make sense from the farmers' point of view.

The farmers are knowledgeable about the microenvironment
(soil, climate) in which they work, more so than the research-
ers working at the national or even regional level. The farm-
ers also are well aware of the resources available to their
household (land, labor, irrigation, equipment, cash or credit).
These resources vary among households even within the same en-
vironment, and from year to year for the same household. While
farmers may not be cognizant of the details of legislation
bearing on agricultural production, they are well aware of cur-
rent prices and regulations pertaining to agricultural inputs
and to the marketing of their crops.






-20-


Farmers, functioning as managers, integrate the informa-
tion available to them on the various constraints described
above and choose the strategies best adapted for their particu-
lar circumstances, goals, and incentives (Figure 2). These
differ among the farmers and the researchers. Traditionally,
researchers use yield as the standard of success: the higher
the yield, the better the research. Yield is not the only
standard of success for farmers. They have more complex goals:
to achieve maximum well-being for the household and, in the
less favorable climates, to avoid a catastrophic series of crop
failures. This means achieving the best possible combination
of sufficient food and sufficient income while avoiding exces-
sive (economic and human) costs and risks of production.

Incentives also differ between farmers and researchers.
To be respected by his peers, a farmer must first be a good
provider and, if possible, better the household's economic
status. Putting one's entire fields into a new variety which
could yield a bumper crop but could also fail would be con-
sidered irresponsible. There is no need to call upon some
"risk aversion" inherent to farmers in the LDCs, for this is
perfectly in line with Western principles of good husbandry.

For a researcher, however, reaching higher yields under
experimental conditions is a recognized way to make his name
known and obtain the consideration of his peers. A crop fail-
ure is an expected but temporary set-back and does not influ-
ence his salary or the food available to his family. The
researcher's training influences the type and level of sophis-
tication of research activities he or she would like to under-
take.

The goals and incentive structure of the researchers de-
termine the research programs in which they would like to par-
ticipate, although in many research systems the researchers are
limited in their choice of activity by administrative and fund-
ing constraints. But the goals and incentive structure of the
farming community determine which of these findings have a
chance of adoption. Were the researchers to become aware of
the goals and incentives that apply to the farming community,
the research programs would become more effective.

Investments in agricultural research are more likely to
achieve their optimal rate of return if the research programs
are established as follows:

-- Researchers and decision-makers are made aware of the
farmers' priorities and constraints.

-- The research program is integrated into a broader plan
for agricultural development, so that all necessary
services are available.






-21-


Figure 2. Changes in Farming From the Farmer's Point of View




ENVIRONMENT


Biological Political/Economic
(micro-level)
--Macropolicy
--Infrastructure


KNOWLEDGE HOUSEHOLD
OF RESOURCES
TRADITIONAL
AND --Land
IMPROVED _FARMING --Labor
TECHNOLOGIES HOUSEHOLD --Water
--Equipment
--Credit
--Etc.


ALLOCATION OF RESOURCES

--Farming/Livestock/Other Activities
--Cropping System
--Farming Practices






-22-


-There is a systematic information feedback mechanism
among researchers, extension agents, and farmers.

-The research focus is on identification of improved
"modules," components that can be used alone, rather
than on an improved package that only works well as a
unit.

-- Research findings are tested in real farm conditions.

Such a research program cannot be implemented by agronomists
isolated in a research station. Farm-oriented research re-
quires an interdisciplinary approach, with agricultural econ-
omists and sociologists/anthropologists joining an array of
technical scientists. It also requires working outside the
experimental stations with the farmers, to ascertain their
needs and constraints and to have them test suggested improve-
ments in real farming conditions.

A number of lessons for the design and implementation of
research activities and for the desirable structure of the
research institutions which derive from this situation are
illustrated in the impact evaluation reports.


1. The Impact of Research on Food Production


One has read frequently of the miracles of the Green Revo-
lution over the last 20 years, and indeed AID's experience
includes success stories of research results being quickly
adopted and leading to increased productivity and to positive
economic returns for the country. These stories are often
mitigated by some drawbacks in the actual impact of improved
technology, which make them all the more worth considering for
lessons learned. Among the impact evaluations, Tunisia, Korea,
and Kenya are examples of particularly widespread use of re-
search results.

The Tunisia Wheat Development Program, which sought to
adapt wheat varieties developed by Mexico's Center for Maize
and Wheat Improvement (CIMMYT) to the Tunisian environment and
climate, was implemented to counteract the decline in wheat
production resulting from the departure of the French estate
farmers at Independence. The program was successful in both
technical terms (development by Tunisian scientists of new
varieties during and after the project) and economic terms.
Some of the new varieties, which were quickly accepted, have
led to an overall increase in wheat production of more than 5.3
million metric tons for the 11 years 1971 through 1981, com-
pared with the previous 11 years. Despite population growth,
annual per capital production of cereals increased from 104






-23-


kilograms in 1970 to 160 kilograms in 1980. The evaluation team
calculated that the increased production saved the Government
of Tunisia almost $126 million a year in grain importation
costs.

This is not to say that Tunisia has become self-sufficient
in cereals, a goal which the evaluation team calls illusory,
pointing out that the best utilization of natural resources is
more important than a drive for self-sufficiency.

Several factors which facilitated success of the program
in Tunisia should be noted. First, the need and opportunity
were clear for increasing production of the staple food at a
time when the foreign estate owners had left and land was being
redistributed. However, early Government attempts to organize
cooperative cultivation of the estates failed, and the rate of
adoption of improved technology did not increase until individ-
ual farms became the norm. The evaluation team identified two
key factors in the success of the program--a strong research-
extension link and the training component. The training compo-
nent assured the sustainability of the research effort, and
indeed the varieties developed by Tunisi scientists after the
project had ended are now the most used. The high rate of
adoption, however, is attributed to good coordination and feed-
back between research and extension.

Lesson 1. A two-way information system between the researchers
and extension service and the farmers is essential in program-
ming and implementing research activities.

This lesson goes beyond the recommendation found in every
one of the eight impact evaluations that research and extension
need to be more closely linked. This may seem obvious since
there is no point in developing improved technology for farm-
ers' use if there is no coherent effort to inform them of its
existence or of how to use it. Yet, making research results
available to farmers is not always easy, especially when there
is little cooperation--or even outright rivalry--between the
research institutions and the extension service of a country.
However, if a new technology is worth using, the first farmers
who learn of it will pass on the word and the adoption rate
will likely be high and fast, with or without further interven-
tion by the extension service. This was clearly shown in
Kenya. Awareness of a new technique, however, is not suffi-
cient to ensure its proper use.





13Lessons for institutional development are discussed in
Section III.D below.






-24-


It also is important that researchers be informed of how
farmers are receiving the new information, what reasons they
give for not adopting the extension agents' recommendations or
for adopting only part of them, and eventually how they modify
these recommendations for their own purpose. The researchers
should be involved in obtaining feedback from the farmers.

In addition to Tunisia, this lesson is particularly clear
in the case of Korea, Kenya, and Nepal, for different reasons,
each of which provides additional lessons learned:

In Korea, success was assured by a strong extension
service.

In Kenya, success was made possible by the technical
simplicity of the research findings and by the avail-
ability of needed services (through the private
sector).

In the Nepal Plain, success was hampered by the com-
plexity of the recommendations (technical package) and
by the insufficiencies of support services, a clear
reverse of the Kenya situation.

The Korea evaluation found that a significant increase in
production (due to widespread adoption of a new rice variety)
could be attributed partly to the fact that research and exten-
sion are closely linked. Improved varieties of rice have been
widely adopted; their use has increased from 16 percent of rice
acreage in 1972 to 60 percent in 1979. The Tongil variety in
particular has become ubiquitous because it yields more than
previous varieties under farmers' conditions. This rapid in-
crease is due in great part to the extension service, which is
effective and very comprehensive. The team cited "the integra-
tion of research and extension" as a key to the project's wide
impact. Extension activities included the monitoring of farm
trials, training programs, and demonstration plots.

This widespread use of Tongil, which even led to a de-
crease in cultivation of other crops, was also the result of a
higher official farmgate price for rice. While these were
positive economic results for the Korean farmers, the use of
Tongil rice also made them more dependent on that one source of
income and therefore more vulnerable. Since 1977 the profit-
ability of Tongil has decreased as yields declined because of
the occurrence of rice blast disease and several years of un-
favorable cold weather.






-25-


Lesson 2. A simple change in input or practices is more likely
to be quickly adopted than a complex technical package.

Kenya is a clear example of the introduction of a techni-
cal improvement, a high-yielding hybrid maize, which was
quickly accepted by the farmers because it fitted easily into
their traditional practices and did not change the schedule of
farming activities. Simply switching to the hybrid resulted in
higher yields. Many Kenyan farmers promptly adopted the hy-
brid, even though new seed had to be bought each year. The
evaluation team hypothesized that the farmers could assign less
land to maize, their staple food crop, and still assure an ade-
quate food supply for the household. That left available land
that could then be used for a cash crop. The introduction of
hybrid maize enabled Kenya to become self-sufficient in that
crop for the first time.

The same rapid rate of adoption is likely when a new pack-
age of practices is obviously beneficial to the farmers, as was
the recommendation in Guatemala to lower seed density and fer-
tilizer rate.

Lesson 3. Support services must be available to the farmers.

The rapid rate of adoption in Kenya and Guatemala was fa-
cilitated by a concomitant improvement in needed support ser-
vices, so that seeds and other inputs were available to respond
to the increased demand. In both cases, the bottleneck in seed
multiplication and input distribution was avoided by the in-
volvement of the private sector. In Kenya, seed multiplication
was taken up by the Kenya Seed Company. Because hybrid maize
seed must be renewed each year, the company is assured of a
steady market.

In Guatemala, seed multiplication and distribution has
been transferred from the State to private growers. The Insti-
tute of Agricultural Science and Technology (ICTA) is involved
in quality control only for the first generation, and it rents
its processing and storage facilities to the growers. The
first generation seed is sold as "ICTA certified," but there
are no controls for second generation seed, a potential danger.
For the time being, the evaluation found the multiplication and
distribution system effective and calculated "that seed devel-
oped by ICTA was worth at least $10 million to Guatemalan agri-
culture in 1979, compared to the ICTA budget of $4 million....
Seed sales considerably decrease the Guatemalan foreign ex-
change levels previously spent on seed import.... [In] the






-26-


coastal area ... 95 percent of the farmers now use ICTA-
developed varieties, compared with less than 50 percent in 1975
using improved varieties."

Often though, the situation is more complex. The farmers
may want to adopt some new technology but find it difficult
because the necessary ancillary services are not yet available
(or not sufficient). This can occur for high-yielding variety
seed multiplication and distribution; the availability of in-
puts, especially fertilizer; and the availability of water,
machinery, repair services, and storage, processing, and mar-
keting facilities.

The farmers also may be selective in adopting improved
technology and practices because of conflicts with other farm
or household activities.

Lesson 4. The project designers and researchers should under-
stand the existing farming system and be aware of local agro-
ecological conditions and of the resources available to the
farmers as they establish the research program. This requires
an interdisciplinary effort.

Throughout the impact and routine evaluations, there are
numerous examples of research activities producing results
which are technically perfectly valid but which are not adopted
by the farmers as expected. Insisting, as did a Nepalese re-
searcher during an impact evaluation, that "those farmers
simply have to be convinced [to use higher doses of fertili-
zer]" is not constructive.

The impact evaluations showed that awareness of improved
technology is not a problem. In addition to the diffusion of
information through the extension services, including eventu-
ally radio programs and leaflets, information can be spread
quickly by word-of-mouth among the farmers themselves. Aware-
ness however, does not guaranty understanding of correct utili-
zation of a new input or practice, not does it guaranty its
adoption.

In the case of Nepal, the evaluation team found that farm-
ers in the Tarai plain were well aware of the advantages of
fertilization. They also quickly understood that application
of fertilizer on wheat was not profitable if one could not be
sure that water also would be available on time. Many farmers,
having learned the hard way that they could not control the
timely availability of water, cut back on the use of fertili-
zer. This was true in some areas not because of the lack of


14Report No. 27, p. 7.






-27-


irrigation facilities but because of frequent power shortages
during which the irrigation pumps could not be used. Further-
more, the high-yielding wheat varieties perform best if planted
in early November, a time which conflicts with the rice
harvest. Therefore wheat often is planted too late, and fer-
tilizer application would not offset the loss in production
potential because of late planting. Finally, the distribution
system for seeds and fertilizer, which is controlled by the
Government, is not efficient and has not been able to respond
to the increased demand in the plain. The situation is much
worse in the hills and mountain regions, where transportation
is exceedingly difficult.

These constraints have not prevented widespread adoption
of improved varieties--acreage in high-yielding rice varieties
increased from 0.6 percent in 1965-1966 to 25 percent in 1979-
1980, and for wheat from 4 percent to about 85 percent in the
same years (wheat is a new crop for most farmers). These con-
straints did however, prevent the farmers from adopting the
entire technical package, and therefore from reaching the ex-
pected yields.

Another example of the selectivity of Nepalese farmers was
found with maize. The improved varieties of maize yield more
than the local strains, and the farmers know it, but the ears
do not keep as well. Many producers compromise by planting
part of their land to improved maize for immediate sale as a
source of cash income, and the rest to local maize for house-
hold consumption.

It is not the farmers who need to "be convinced," but the
researchers who must look for improvements which are effective
in real farm conditions, taking--as the farmers do--the entire
set of resources, priorities, and environment of the household
into consideration.

Two of the impact evaluations looked at projects using
such a farming system approach, and both are optimistic regard-
ing the project impact on food producers. The Guatemala evalu-
ation lists this as its first lesson learned:

"Farming system research" has been almost romanti-
cized by some students of agricultural research.
This evaluation serves as one of the first studies to
bring hard data to this new topic. The ICTA approach
to technology development demonstrates clearly the
positive benefits derived from this unconventional
approach for generating acceptable small farmer tech-
nologies and practices.


15Report No. 27, p. 12.






-28-


The evaluation of the small farmer cropping systems pro-
gram coordinated by CATIE in Central America also emphasizes
that the system approach was conducive to the development of
the improved technology adapted to farmers' needs. Having
established this factor, both reports make specific recommenda-
tions for assuring maximum effectiveness of a farming system
strategy. They emphasize the need for feedback from the farm-
ers to the researchers and for more active involvement of the
farmers than simply allowing use of their land for on-farm
testing. They also specify that such research can be conducted
only by an interdisciplinary team with technological, economic,
and sociological expertise. The implications of these recom-
mendations for institutional development will be discussed in
Section D below.


2. The Impact of Research on Farmers' Well-Being and on Rural
Equity


Technical improvements by themselves should not be ex-
pected to lead to a more equal income distribution among the
population. Macropolicies, especially land tenure rights and
access to means of production and support services, will
determine which way research results will influence income
distribution. These resources vary among households even with-
in the same environment. This is perhaps the most complex of
all problems faced by the researchers, as improved technologies
often assume the availability of resources, such as water,
which not all households can obtain, and may require an inten-
sification of land use, thereby increasing labor and input
requirements. These technologies, by their very nature, may be
practical only for the better-off households.

The question of equity, i.e., giving all farmers equal
access to benefits from the project, is difficult for several
reasons. Governments often place a higher priority on assuring
the food supply of the urban populations than on bettering the
income distribution among farmers. It is also a difficult
question from a technical viewpoint because many new or im-
proved farming technologies simply are not efficient on a small
scale, or demand a level of investment in tools, inputs, water,
or labor beyond the reach of the smaller farmers, especially
those who are tenants.

In Nepal, farmers with some irrigated land have had imme-
diate advantage over those with only rainfed land in using the
improved varieties of wheat and maize. Farmers who were better
off in the first place were more likely to be able to finance
the necessary inputs. Tenant farmers were disadvantaged be-
cause they did not qualify for credit to buy inputs, and proba-
bly had less incentive to invest in the land.






-29-


Even in Kenya, where the overall output of maize was
greatly increased as a result of research, the impact on equity
within the country probably was negative. Disparity increased
between the large and small farmers because the smallest farm-
ers were reluctant to adopt the hybrid. Their main concern was
to minimize the risk of crop failure (which the hybrid maize
did not do) rather than to increase production. In addition,
they were not able to finance inputs; even the need to buy new
seed each year was a problem.

In contrast, the project in Korea contributed positively
to equity among farmers because of the price subsidies provided
by the Government and relatively equitable land distribution.

Lesson 5. Technological improvements can sharpen inequity
among households with different resource bases.

The Tunisia report describes a mixed equity impact for the
Wheat Project. On one hand, farms of all sizes gained access
to more productive technology and reached higher yields. The
more intensive mode of production has made mechanization more
profitable; this does not necessarily lead to a negative impact
on smaller farms that are too small to support capital-
intensive farming, because some small farmers invest in heavy
equipment and work other farmers' land as well as their own.
On the other hand, the report mentions a decrease in labor
demand in rural areas and a rural exodus, especially among
younger people. The evaluation team especially raises the
issue of negative impact on women, because of changes in labor
demand, and on the nutritional status of the family. The in-
crease in overall cereal production has been accompanied by a
shift from hard wheat to bread wheat varieties, with a subse-
quent nutritional loss.

Lesson 6. Technological improvements can have both positive
and negative impacts on a rural household's income and well-
being.

As the Tunisia case has shown, one should remember that an
increase in production of one crop does not necessarily lead to
better overall well-being for the household. When a farmer
switches to a high-yielding variety, the cost of production
usually increases, and more labor is required from the family
and eventually from hired labor. The opportunity cost of land
and labor should be taken into account, as often a change in
farming practices will force the household to cut down on some
other income-producing (or expense-saving) activity. This may
be especially true of women's activities.

These changes in turn influence productivity, food supply,
income, and pattern of land use. There will be consequences
both at the household and the community level.






-30-


The impact evaluations did not look specifically at the
projects' impacts on consumers. However, the projects may have
influenced the food price structure through increased produc-
tion and also through changes in cropping systems. A shift in
land use toward a crop (e.g., rice) or a variety that is espe-
cialy in demand in urban areas is likely to benefi-t the urban
consumers, although not necessarily the poorer ones.


3. Conclusions


Lessons 1 to 6 describe the type of research which is
likely to be most effective in meeting farmers' needs and in
leading to increased production. One further lesson derives
from these: the research institution must be given the means
to implement a research strategy that focuses on the farming
system as a whole as well as addresses the technical problems
of commodity production. Institutional development is such a
crucial component of AID assistance to agricultural research
that it will be discussed in a separate section (Section D
below), but this section has already established the need for
interdisciplinary expertise and for the material and human
resources to establish on-farm testing and to gather baseline
data. This section has also established the importance of
close coordination between research, extension, and agencies
involved in support services, as well as training institutions
that take research requirements and findings into account in
their curricula.

This does not mean that farming system research is the
only effective type of research program and therefore the most
worthy of investment. The need for basic research programs and
commodity-oriented programs will remain, but such programs are
more likely to lead to useful results if they are planned in
conjunction with farmer oriented research.

The next section will focus on the impact of national
policies and economic environment on the programming of re-
search activities and on the utilization of research results.





-31-


C. The Utilization of Research Findings Is Dependent on a
Favorable Political and Economic Environment


Nothing in development occurs in a political and economic
vacuum, not even scientific research in a laboratory. This
basic fact pervades AID's experience with research projects, as
the mandate of research institutions changes over time, as
budgets and human resources ebb and flow, and as extraneous
constraints impede the utilization of research findings.


1. Technological Solutions Alone Cannot Solve Problems Which
Are Basically Economic in Nature


The successes--albeit mitigated--discussed in the previous
pages should not hide the fact that technological constraints
are but a few of the factors that influence food production,
and that technological solutions should not be expected to
solve economic problems. Examples of such factors are present
in all the impact evaluations, but they may be most clearly
stated in the West Africa Rice Development Association (WARDA)
evaluation. The team has shown that the original mandate of
WARDA emphasized solving the technical problems of rice varie-
ties suitable to the ecological conditions of West Africa, when
in fact indigenous rice production was discouraged not only by
the lack of such varieties, but by pricing and marketing regu-
lations.

Lesson 7. Government policies and infrastructure determine, in
part, whether farmers will adopt improved technology and prac-
tices.

Section B has shown that the farmers act as managers in
selecting production strategies and therefore take into account
the macropolicies which determine price, net return, and mar-
keting opportunities for their crops. The farmgate and consu-
mer price of food and other agricultural commodities; price,
quality, and availability of inputs; efficiency of marketing
systems; foreign trade regulations; and land tenure are all
potential constraints on farmers' decisions that are affected
by government policy.

This means that the researchers should be aware of exist-
ing policies and may eventually try to influence them. It does
not mean that research findings are doomed if policies are not
favorable or if the required support services are not avail-
able; in many countries, both developed and less developed, a
new technology can spread and stimulate the necessary changes
or additions to existing infrastructure and services. Thus, in
Europe, farmer cooperatives were created when the farmers





-31-


C. The Utilization of Research Findings Is Dependent on a
Favorable Political and Economic Environment


Nothing in development occurs in a political and economic
vacuum, not even scientific research in a laboratory. This
basic fact pervades AID's experience with research projects, as
the mandate of research institutions changes over time, as
budgets and human resources ebb and flow, and as extraneous
constraints impede the utilization of research findings.


1. Technological Solutions Alone Cannot Solve Problems Which
Are Basically Economic in Nature


The successes--albeit mitigated--discussed in the previous
pages should not hide the fact that technological constraints
are but a few of the factors that influence food production,
and that technological solutions should not be expected to
solve economic problems. Examples of such factors are present
in all the impact evaluations, but they may be most clearly
stated in the West Africa Rice Development Association (WARDA)
evaluation. The team has shown that the original mandate of
WARDA emphasized solving the technical problems of rice varie-
ties suitable to the ecological conditions of West Africa, when
in fact indigenous rice production was discouraged not only by
the lack of such varieties, but by pricing and marketing regu-
lations.

Lesson 7. Government policies and infrastructure determine, in
part, whether farmers will adopt improved technology and prac-
tices.

Section B has shown that the farmers act as managers in
selecting production strategies and therefore take into account
the macropolicies which determine price, net return, and mar-
keting opportunities for their crops. The farmgate and consu-
mer price of food and other agricultural commodities; price,
quality, and availability of inputs; efficiency of marketing
systems; foreign trade regulations; and land tenure are all
potential constraints on farmers' decisions that are affected
by government policy.

This means that the researchers should be aware of exist-
ing policies and may eventually try to influence them. It does
not mean that research findings are doomed if policies are not
favorable or if the required support services are not avail-
able; in many countries, both developed and less developed, a
new technology can spread and stimulate the necessary changes
or additions to existing infrastructure and services. Thus, in
Europe, farmer cooperatives were created when the farmers






-32-


became convinced of the advantages of using fertilizer but were
disappointed with the quality of the existing distribution
services. In India, the availability of new wheat strains
stimulated the development of a fertilizer industry and the
multiplication of irrigation systems.

The breeding of new maize varieties in Kenya led to the
development of seed multiplication and distribution by the
private sector. In Tunisia, the spread of improved wheat cul-
tivation was hampered at first by the Government policy of
cooperative cultivation of the estates previously controlled by
foreign colonists. Only when the Government backed off and
allowed private cultivation did modern technology spread.

In other projects, for example, in Nepal, the unreliabil-
ity of input supply has hampered the adoption of improved tech-
nology (see Section II.B).


2. Host Government Commitment Is Essential


Lesson 8. Real, long-term commitment to agricultural research
on the part of the host government determines the sustainabil-
ity of a research project and utilization of its findings.

No matter how productive a research station may have been
during the implementation of a project, and even within a
favorable policy environment, the ability of an institution to
sustain research activities on its own is a function of the
host government's commitment to research. This is basically
what determines whether the research institution will be given
the human, financial, and administrative means to pursue its
activities. The commitment of the host government also deter-
mines how research activities will be programmed and whether
related policies might be revised to facilitate the utilization
of research findings.

The research institutions in Kenya and Thailand suffered
from the lack of such support, expressed through insufficient
staff allocation in Kenya and through the uncertain legal sta-
tus and changes in mandate of the Thailand Center. In both
cases, the teams found that research activities could not con-
tinue at the same pace after the departure of the project's
technical assistants.

The very success of the Korea project is attributable in
large degree to the commitment of the Government, which gave
agricultural research and extension high priority. Research
stations existed and were already effective prior to the AID
project. Its program to increase the production of rice and





-33-


other crops was conducted with the full support of the Govern-
ment, which revised its pricing policy for rice to encourage
widespread use of improved varieties and to increase the farm-
ers' incomes.

Routine evaluations frequently mention inadequate host
government support for the project as resulting in implementa-
tion difficulties, while the impact evaluations have focused
more on the impact of host government commitment on the long-
term effectiveness of the research institution. Among the
routine evaluations reviewed for this study, there was a clear
correlation between inadequate support and a "less than satis-
factory project performance" rating (17 of the 23 projects with
inadequate support were found unsatisfactory). The effect of
inadequate support is immediately visible through the lack of
counterpart personnel, delays in procurement and management,
and delays in identification of candidates for training.

A consequence of this after the project has ended is the
inability of the research station to maintain an adequate staff
and sufficient equipment. (Lessons learned on this issue are
presented in Section D since they are pertinent to institu-
tional development.)

Lesson 9. Agricultural research programs should be planned
within the broader rural development planning.

This integration of rural development and research plan-
ning (but not necessarily implementation) will help ensure that
priorities are set up for research activities according to na-
tional goals and that there will be coordination among research
activities and other development activities that influence the
impact of research, such as extension, the provision of inputs
and credit, and marketing channels.

It also will facilitate the phasing of various actions,
including the effects of changes in policy and regulations on
prices of crops and inputs and on marketing. Development,
after all, is the systematic elimination of limiting factors.

Perhaps the effectiveness of an integrated approach to
research and agricultural policies is best demonstrated in the
Korea project. The Korean Government showed its commitment to
research by establishing a network of competent research sta-
tions, by assuring a productive collaboration between research
and extension, by assuring the availability of required support
services, and by revising its pricing policy for rice, as
needed, to encourage farmers' adoption of a new improved vari-
ety. This resulted in a rapid spread of the new technology.






-34-


Lesson 10. A dialogue among politicians, administrators, and
researchers will greatly increase the likelihood of adequate
support to research; the potential benefits of research for the
host government should be made clear.

No government should be expected to commit its limited
human and material resources to an activity for which it cannot
foresee a benefit for the country or for itself. In other
words, researchers should not expect a continuous flow of re-
sources if they do not show some results which the government
can understand as economically and politically beneficial to
itself, and this within a fairly short time. A problem is
likely to occur if a station expects many years of support
before it has anything to show for it.

The farming system approach or problem-oriented research
which has been found most effective in AID experience may re-
quire a larger staff than traditional on-station research did,
but it also is more likely to show some rapid results, as re-
searchers propose solutions to problems identified by farmers
and extension agents. It can speak for itself more quickly
than basic research does.

The workshop participants, especially the host government
officials and the AID field personnel, were keenly interested
in finding ways to demonstrate to the host governments the
potential benefits of such research programs, and they empha-
sized the need for a continuous dialogue among politicians,
administrators, and researchers during research programming and
resource allocation, as well as during project identification
and design.

However, this should not obscure the fact that few coun-
tries could possibly assume the recurrent costs for all of the
development activities that are currently under way with AID or
other donor-institution assistance; this is particularly true
in the African countries which are presently in most need of
developing their agricultural research capacity.


D. Characteristics of Effective Agricultural Research Systems


The importance of adapting the research program to farm-
ers' needs and constraints and of devising policies and support
services that facilitate the adoption of improved technologies
have been established. The next questions that need to be an-
swered are what kind of administrative structure is more likely
to achieve the desired results, and what kind of staffing pat-
tern is necessary? This section will discuss training and in-
stitutional issues, as well as the mechanisms for coordination
among research and development institutions.






-35-


All the projects selected for an impact evaluation in-
cluded a component for training and institution development at
either the regional or national level. Whether the research
institutions are functioning adequately after the project has
ended is a crucial element in determining the sustainability of
the project achievements. There are also sets of issues recur-
ring in the evaluations: the location of the institution with-
in a country's administrative system and within the research
community, and the staff and resources allocated to the insti-
tution.


1. The Effectiveness of an Institution Depends on Its Place in
the Administrative Structure and the Resources It Receives


A well-trained body of researchers will only be as effec-
tive as the institution they work for. Furthermore, the more
motivated and competent researchers will not remain in an
institution that does not allow for professional growth and
satisfaction.

Lesson 11. The mandate and authority of a research institution
must be clearly defined and agreed upon with the host govern-
ment.

Institutional issues were found to be a problem by most of
the impact evaluations; although the case of "bureaucratic im-
potence" described in the Thailand report may be extreme, it
does illustrate how the lack of administrative authority can
hamper the effectiveness of an otherwise competent and well-
equipped institution. Conflict between the Research Center and
the central Ministry of Agriculture "created an atmosphere in
which much research done at the Center is rejected out of hand
by the [Ministry] and often has to be redone in order to be
acceptable. Declining budgets, loss of coordinating authority,
frequent institutional redefinition, and loss of status and
professional autonomy have combined with previously mentioned
factors to defeat effort to build a major research capacity in
Northeastern Thailand."' One of the sources of the problem
became evident early during project implementation when the
Government postponed giving the Center the proper legal status.
The report points out that AID structures and procedures in
project design and implementation did not encourage a revision
of the project after the project had started; the negative ef-
fects became clear after the project itself had ended.


16Report No. 34, p. iv.






-36-


Similarly, the East African research institution faded
away when the technical assistants left because the host coun-
tries had not provided adequate staff and support to assure its
sustainability.

Implementing a project through an institution outside the
line ministry could be a temptation in many countries, in an
effort to assure more autonomy to the project staff. This was
the case in Tunisia, and it did in fact facilitate project im-
plementation. There is a danger of insufficient communication
between the research institution and entities involved in agri-
cultural development that could hamper the development of a
long-term, self-sustaining research capacity. This is not hap-
pening presently in Tunisia because of good personal contacts
and exchanges among scientists in the various entities involved
in research, partly because those individuals have been trained
together and know each other well. This is fine as long as it
lasts, but it does make the research institution vulnerable,
since these exchanges have never been formalized.


2. Institutional Development and the Concomitant Training of
Scientists Is a Long-Term, Complex Process of Critical
Importance to the Sustainability of a Research Project


The training component seems to have been achieved suc-
cessfully in all the projects evaluated except perhaps Kenya,
but keeping the trainees on the job after their return has been
more of a problem.

Training is considered a major benefit in many development
projects, especially for the attainment of scientific degrees
which could not be obtained locally in many countries. One
should not, however, think automatically of a Ph.D. from a U.S.
university when talking about training. In-service training,
short-term technical courses, and even observation tours and
participation in professional meetings can be of great advan-
tage to the trainees, as well as the formal M.S. and Ph.D.
degrees. The workshop participants recommended that AID be
particularly flexible in its approach to training. There are
great variations among host governments in their training
needs, which depend both on the planned research systems and
the number and level of training of existing researchers.

There are three key aspects in training: the level and
scope of the training program, its timing and scheduling, and
its location.






-37-


Lesson 12. Training should be adapted to needs.

Training is likely to be needed not only in the tradi-
tional areas of expertise related to technical aspects of
agricultural research, but also in disciplines of the social
sciences which are necessary if the socioeconomic factors that
influence development are taken into account in the research
program. In addition, training in management is often cited as
important, as many researchers are given extensive managerial
responsibilities.

M.S. and Ph.D. training is by definition a long-term pro-
cess. A Ph.D. may require that a candidate spend three to four
years out of the country, even if the thesis research is done
in-country. Remedial courses and language training can make
the process even longer. If one takes into account the time
necessary for identifying suitable candidates and getting them
accepted in a U.S. university, this becomes a major enterprise,
longer than the timeframe of a development project.

There is indeed a major timing difficulty in projects that
combine long-term training abroad and technical assistance.
More often than not, the technical assistant is not providing
in-service training to his counterpart, he is conducting re-
search until the counterpart comes back to "sink or swim."

Providing training assistance outside of a specific devel-
opment project would help solve this difficulty. AID has done
this in the past, and is currently funding such a program in
the Africa Bureau, albeit on a small scale.

Where training is provided is also an important factor.
AID restrictions against training in developed countries out-
side of the U.S. are a problem, especially for trainees from
non-English speaking countries. Workshop participants encour-
aged emphasis on arrangements through which trainees who have
completed course work can return home and do thesis work in the
environment and on the type of problem they will deal with in
professional life.

Lesson 13. There should be official linkages and feedback
mechanisms among institutions and government entities with
responsibilities in education, research, extension, and the
provision of services.

In Section B, it was shown that feedback mechanisms from
extension to research and among research, extension, and the
service providers are important to an effective research pro-
gram. This also means that feedback should exist to the educa-
tional institutions in agriculture and the social sciences,
which must adapt the curricula on research, extension, and
agricultural development courses to the expected needs of






-38-


dealing with these activities. This is essential as the coun-
try becomes less dependent on training opportunities offered
through donor institutions.

Lesson 14. The scheduling of training and that of technical
assistance should be complementary.

Routine evaluations, which focus on implementation prob-
lems more than impact evaluations do, frequently mention sched-
uling conflicts between the training and technical assistance
components of a project, where the technical assistant leaves
when his "counterpart" returns from training abroad. Preproj-
ect training was strongly recommended by the workshop partici-
pants and is now encouraged, at least in the Africa Bureau.

Lesson 15. Trainees should be assured of .satisfactory material
and professional awards.

Staff attrition has been found to be a problem, at least
in Kenya, Guatemala, and Nepal, in great part because re-
searchers are given a status and payscale different from that
of civil servants. If, in addition, professional rewards are
insufficient because the researchers have no say in the selec-
tion of research topics, or must work with unsatisfactory
equipment, the danger of staff attrition is indeed great.
Training abroad is considered a great reward, but the returned
trainees who face difficult working conditions and low pay may
soon be tempted to move on. Warnings are raised on this sub-
ject in most of the eight evaluations.


3. Linkages Among National and International Research
Institutions Are Essential


No research institution can be fully self-sufficient, nor
should it try to be so. This is especially true of national
research institutions which have limited human and material
resources at their disposal.

Lesson 16. National research institutions should not work in
isolation.

For reasons of research effectiveness and professional
satisfaction of the researchers, all the impact evaluations
(except Kenya) emphasized the absolute necessity of establish-
ing effective coordination mechanisms among the various govern-
ment institutions related to agricultural development, including
the research institutions, and more specifically between the
research institutions and the extension services. In addition,
the importance of coordination and exchange of information be-
tween research stations within the country as well as witn
regional and international institutions was emphasized.






-39-


The evaluation team in Guatemala found that "ICTA's links
to international agricultural research centers and to U.S. cen-
ters of technology expertise were highly productive. Technolo-
gies and concepts from these centers were applied in Guatemala,
and through these same centers the Guatemalan experience is
coming to the attention of other countries around the world.
Both AID as an Agency and its Missions within each country
should be aware of the capabilities of research centers and
consider ways to make uselof these resources in future research
and development efforts."

An effective means of coordination has been the creation
of working groups in which representatives from the various
agencies and institutions regularly exchange information on
achievements and future plans. For example, in Nepal, where
research stations are specialized by commodity, yearly work-
shops enable the researchers to present their findings to their
peers, discuss each station's future program, and coordinate
some common trials. The Cropping Systems Working Group in Asia
has become a much appreciated means of communication among
national scientists in the region.

Lesson 17. An international research entity can provide very
useful assistance to national research systems.

This is verified in the impact evaluations which assessed
the impact of international institutions (CATIE in Central
America, WARDA in West Africa), as well as the evaluation of
the Tunisian national system, which greatly benefited from
CIMMYT assistance, and that of the Guatemala institution.

The CATIE evaluation also found that exchanges of informa-
tion and coordination among institutions were useful, and it
calls for "maximum collaboration and information sharing .
among related projects and programs." It does point out, how-
ever, that such "collaboration and synergism" pose difficult
managerial problems.

The WARDA report raises some interesting issues. It
points out that a regional institution "should not be used as a
fallback resource when national systems prove administratively
inadequate for pursuing development of objectives, but
rather as a means to improve the scientific inadequacies of
these national systems." Donors should not use a regional
institution as a substitute manager for their national develop-
ment programs, thus preventing the regional institution from
assuming its own scientific role. The team did find WARDA to
be "a particularly effective quality control, advisory backstop


17Report No. 27, p. 13.






-40-


to the national research systems of the 15 countries it
serves."


E. Logistical Difficulties Should Not Be Underestimated


While logistical difficulties are to be expected in any
development activity, they appear with a vengeance in research
projects, which often include large training and commodity com-
ponents. The very thought of ordering one million dollars
worth of scientific equipment, bringing it into the country and
getting it out of customs, and respecting the regulations of
AID, the host government, and the contractor's institution
ought to give nightmares to even experienced project officers.
That task, however, is given not to an experienced procurement
officer but to the chief of party of the research project, who
is selected for the job on the basis of research experience and
accomplishments.

Many routine evaluations point out that the chief of party
is obliged to neglect his/her research role simply to keep up
with--or try not to get too far behind--the managerial tasks of
the project team. This can lead to much frustration and bad
feelings between the technical assistants and the host govern-
ment, as both sides are shortchanged in the process.

This situation is compounded in a loan, when logistical
support of the technical assistants is to be provided by the
host government. Some routine evaluations have recommended
that in such a case, AID should assure that adequate logistical
support will be available on time, either by budgeting for it
or through precedent conditions. This does not apply to normal
recurrent costs of the host institution, only to special ex-
penses for the direct benefit of the technical assistants
(e.g., housing, transportation, and secretarial services).


IV. UTILIZATION OF LESSONS LEARNED FOR FUTURE AID ACTIVITIES18


Lessons from past experience are worthless unless they are
incorporated into the planning and implementation of new activ-
ities. About half of the Workshop on Impact of Agricultural
Research was devoted to small group discussions of how the


18This section draws heavily from the Workshop Proceedings
presented in Appendix D. It is not an official statement of
AID policy or strategy, but the sum of the experience of the
workshop participants.






-40-


to the national research systems of the 15 countries it
serves."


E. Logistical Difficulties Should Not Be Underestimated


While logistical difficulties are to be expected in any
development activity, they appear with a vengeance in research
projects, which often include large training and commodity com-
ponents. The very thought of ordering one million dollars
worth of scientific equipment, bringing it into the country and
getting it out of customs, and respecting the regulations of
AID, the host government, and the contractor's institution
ought to give nightmares to even experienced project officers.
That task, however, is given not to an experienced procurement
officer but to the chief of party of the research project, who
is selected for the job on the basis of research experience and
accomplishments.

Many routine evaluations point out that the chief of party
is obliged to neglect his/her research role simply to keep up
with--or try not to get too far behind--the managerial tasks of
the project team. This can lead to much frustration and bad
feelings between the technical assistants and the host govern-
ment, as both sides are shortchanged in the process.

This situation is compounded in a loan, when logistical
support of the technical assistants is to be provided by the
host government. Some routine evaluations have recommended
that in such a case, AID should assure that adequate logistical
support will be available on time, either by budgeting for it
or through precedent conditions. This does not apply to normal
recurrent costs of the host institution, only to special ex-
penses for the direct benefit of the technical assistants
(e.g., housing, transportation, and secretarial services).


IV. UTILIZATION OF LESSONS LEARNED FOR FUTURE AID ACTIVITIES18


Lessons from past experience are worthless unless they are
incorporated into the planning and implementation of new activ-
ities. About half of the Workshop on Impact of Agricultural
Research was devoted to small group discussions of how the


18This section draws heavily from the Workshop Proceedings
presented in Appendix D. It is not an official statement of
AID policy or strategy, but the sum of the experience of the
workshop participants.






-41-


lessons learned could be incorporated into the design and im-
plementation of future AID activities. The result of these
discussions is summarized in Section B below. The discussions
were not limited to the design of "good projects." A good
project, however one defines it, does not fulfill its develop-
ment goal if it does not establish a sustainable and effective
indigenous research capacity.


A. The Changing Relations Between Host Governments and Donor
Institutions


Several formal presentations during the workshop discussed
the changing relations between host governments and donor in-
stitutions. Mr. Curt Farrar, then Deputy Assistant Administra-
tor for Research with the Science and Technology Bureau, USAID,
emphasized that many dimensions of current assistance to re-
search are changing, among which are a decelerating growth of
investments in international research centers, increased donor
collaboration, a stronger focus on understanding the farmers,
an awareness of needed changes in training programs and time-
frames for research assistance, and finally a greater interest
in assisting national research systems and institutions. More
attention is also being given to involving the private sector
in technology innovation and support services.

At a time when development concerns are becoming more
complex, the mechanisms that provided assistance in the past
are becoming less effective. The private foundations whose
leadership was at the origin of the international centers net-
work have much reduced their activities in research. The
international research centers have accomplished dramatic
breakthroughs but must now handle more diversified local needs
under less favorable agroecological conditions. The technical
expertise of USAID has greatly decreased because of a shift
toward managerial staff and an increased reliance on contrac-
tors for technical assistance. The international development
banks are emphasizing resource transfer rather than development
programs.

Professor Vernon Ruttan (University of Minnesota) pointed
out that while successful research projects can be found, suc-
cessful research programs and national systems are rare, and in
many cases the development of physical facilities is outstrip-
ping the growth of a country's capacity to use the facilities.
A disturbing phenomenon is the cycle of rising national re-
search capacity resulting from donor activity, followed by
relative deterioration, as may have been the case in Thailand
and Kenya.






-42-


Donors need to ask if this problem is related to the way
they do business or if the donor project system provides per-
verse incentives to the leaders of national systems. The po-
litical systems of most countries cannot be relied on to turn
out "good" people. They can be relied on to turn out ambitious
individuals, and ambitious individuals respond to organized
pressure. Research managers have to learn to marshall politi-
cal support, and a few national managers have done so. For
many, however, donors are easier to deal with than national
financial sources, and this discourages research leaders from
building the political support essential for a sustained pro-
gram.

Ruttan pointed out that decisions related to project
assistance should be made by criteria of the national system,
not by those of the donor system. This is true also of project
evaluations. Professor Ruttan proposes a formula by which
donor support would be based on increments of national support
and so would give the correct incentives. The formula would
vary from country to country as a function of both fiscal
strength and political will. Under this system, decisions
would be left to the host country, the learning process would
be rapid, and self-interest would bring increasing productiv-
ity.

A second-best alternative would be planning between donors
and the host country following the Joint Commission on Rural
Reconstruction (JCRR) model in Taiwan. The process of learning
and internalizing the management process would be slower under
this alternative.

There would be opposition to this strategy, flowing chiefly
from the loss of identity of donor contributions. However, many
countries would support Ruttan's ideas. Participants from the
Philippines pointed out that researchers are grateful when
donors negotiate with their Government to increase commitment.
Once there is an international contract, it tends to maintain
the stability of the research program even through changes in
government.

The CGIAR experience has provided some lessons regarding
the value of continuity and maintenance of funding, the value
of periodic replanning, and the utility of external, formalized
reviews. The donors who make up the CGIAR treat their national
efforts differently, however. They expect too much, too soon.
They need to apply to national efforts what they have learned
through the CGIAR.

Professor Ruttan suggested that a Consultative Group for
National Agricultural Research (CGNAR) could have an impact on
national systems comparable to that of the CGIAR on the inter-
iational centers. With a five-year planning horizon and a two-






-43-


year plan of work that is continually rolled forward, all
actors would have a basis for commitment. Donors could set
some minimum requirements regarding linkages among research
institutions. The CGNAR would consist of two national leaders
(one from research and .one from planning) and one representa-
tive per donor.

The CGNAR may need a group, probably internal to the re-
search system, to provide information and analysis. Donors
would need to indicate their intended level of support far
enough into the future to allow the national government time to
adjust to changes and to provide for security of expectations.

Some regional research has produced good results, but it
is often beset with political problems and may have no institu-
tionalizing mechanism. An institution like WARDA, which is
independent of national mechanisms, has been found to be espe-
cially helpful for training and, surprisingly, for identifying
and coordinating micro-level research. Networks of researchers
from developing countries could be useful when the country pro-
grams really are interdependent. The success of the Cropping
Systems Working Group in Asia is encouraging.

While not all workshop participants agreed with Professor
Ruttan's proposal, it was generally felt that the role of in-
ternational centers and regional institutions will change as
the capacity of national systems improves. Indeed, the mandate
of the International Service for National Agricultural Research
(ISNAR), the youngest of the international research centers in
the CGIAR, is to provide assistance to host governments in
strengthening their own research system, rather than to organ-
ize research programs directly. Donor countries are also in-
creasing coordination of their activities, for example, through
the Cooperation for Development in Africa (CDA). The United
States has taken primary responsibility for coordination of
assistance to agricultural research under the CDA.


B. Suggestions for AID Assistance


1. Planning Assistance to Agricultural Research Activities


Throughout the impact evaluations and the workshop, the
importance of adequate macropolicies and of government support
to the research system was emphasized. Thus, the current em-
phasis in AID on facilitating policy changes that will encour-
age food production is supported by past experience. A project
must be designed to fit national objectives. This means that
there may be country-specific answers to specific issues and
situations and that the strategy selected for assistance must






-43-


year plan of work that is continually rolled forward, all
actors would have a basis for commitment. Donors could set
some minimum requirements regarding linkages among research
institutions. The CGNAR would consist of two national leaders
(one from research and .one from planning) and one representa-
tive per donor.

The CGNAR may need a group, probably internal to the re-
search system, to provide information and analysis. Donors
would need to indicate their intended level of support far
enough into the future to allow the national government time to
adjust to changes and to provide for security of expectations.

Some regional research has produced good results, but it
is often beset with political problems and may have no institu-
tionalizing mechanism. An institution like WARDA, which is
independent of national mechanisms, has been found to be espe-
cially helpful for training and, surprisingly, for identifying
and coordinating micro-level research. Networks of researchers
from developing countries could be useful when the country pro-
grams really are interdependent. The success of the Cropping
Systems Working Group in Asia is encouraging.

While not all workshop participants agreed with Professor
Ruttan's proposal, it was generally felt that the role of in-
ternational centers and regional institutions will change as
the capacity of national systems improves. Indeed, the mandate
of the International Service for National Agricultural Research
(ISNAR), the youngest of the international research centers in
the CGIAR, is to provide assistance to host governments in
strengthening their own research system, rather than to organ-
ize research programs directly. Donor countries are also in-
creasing coordination of their activities, for example, through
the Cooperation for Development in Africa (CDA). The United
States has taken primary responsibility for coordination of
assistance to agricultural research under the CDA.


B. Suggestions for AID Assistance


1. Planning Assistance to Agricultural Research Activities


Throughout the impact evaluations and the workshop, the
importance of adequate macropolicies and of government support
to the research system was emphasized. Thus, the current em-
phasis in AID on facilitating policy changes that will encour-
age food production is supported by past experience. A project
must be designed to fit national objectives. This means that
there may be country-specific answers to specific issues and
situations and that the strategy selected for assistance must






-44-


fit the host government's political set-up. The total environ-
ment, farm-level constraints, economic policies, and institu-
tional capabilities should be taken into account. A project,
or even the AID program of assistance, does not necessarily
address all of the constraints identified, but it better be
aware of them.

Coordination between the government, AID, and other donor
institutions is essential, at this early stage, to determine
government commitment and priorities as well as to assess the
constraints and resources at hand. The host government should
be actively involved in the preparation of assistance programs,
project identification, and project design. The issues of
availability of counterparts and potential trainees, the ca-
pacity of the host government to assure its contribution to
projects and recurrent costs, the potential conflicts between a
project's timeframe and a realistic schedule and phasing of
activities should be discussed with the donor institution very
early in the process.

In some countries, this may mean that assistance at the
policy and program level will be required first, and that a
"critical mass" of personnel, facilities, and management cap-
ability (both at senior and junior levels) must be assured
before a full research program can be established.

Workshop participants recommended that donor institutions
resist the temptation of pushing a research program through by
temporarily duplicating insufficient local institutions.
Short-term projects run entirely by expatriates make a limited
contribution to the national research capacity.

Institution-building and the concomitant training of sci-
entists is an especially long-term, complex process of critical
importance to the sustainability of a research project. If an
existing institution is to be strengthened, it must be care-
fully selected and treated as part of the overall administra-
tive system of the country and not as an isolated entity.
Training of counterparts for both scientific and managerial
tasks is an integral part of institution-building.

Bilateral agreements with developed countries and inter-
national organizations are not the only sources of assistance;
technical cooperation and exchange of trainees among developing
countries should also be encouraged.

However, both the institutions and the host governments
need visible results on a rather frequent basis as a justifica-
tion for continuing assistance and as an enticement to policy-
makers to reinforce their commitment to the research program
and to continue funding. This can be achieved if it is in-
cluded in the program planning and if the project scientists
and managers are committed to it.






-45-


A well-run research system can give the government a pow-
erful tool for development if it is used both for technology
generation and for problem solving at.the level of micro-
agroecological regions. Used in this way, research investments
can give short-term as well as long-term payoffs.

However, a major difficulty for many donors, and certainly
for USAID, is the fragmentation of assistance into relatively
short projects. This does not allow adequate planning for most
research programs and unduly taxes the host government with re-
quirements for counterpart, support staff, and recurring costs.
It is likely that long-term commitments, if only in principle,
to agricultural research programs will be more acceptable to
host countries at the political and technical level. Mr. Joseph
Wheeler, then Assistant Administrator of USAID, was sympathetic
during the workshop to the suggestion that AID make a commitment
to long-term projects or programs. With long-term approval,
funding could still be handled on a project basis. This, how-
ever, requires from both AID and the host government a long-term
research program with assigned priorities and definite goals
clearly tied to national development goals. Such an exercise,
by itself, would be extremely beneficial to the research system
and to the government, as was shown in Section III.

Since, for the foreseeable future, AID will provide as-
sistance in the form of projects, further recommendations in
this report are made within that framework.


2. Project Design


The preparation of project documents is a complex and
lengthy process. Negotiations will have to take place between
agricultural research institutions and various sections of
government, between donor country mission and home office, and
between country and donor. The host country may spend six
months to a year before a proposal is ready for AID review. It
is essential that the project design be as collaborative an
effort as possible to attain the support of all parties within
both the government and the mission.

Since agricultural research is a long-term endeavor re-
quiring a steady support of funds, donors should consider
whether to include funds for operating expenses in the project,
and how incentives can be built in for national governments to
find sources of long-term support for these increments to the
agricultural research system. A realistic assessment of the
resources the host country can provide, especially human re-
sources and operating funds, should be made during the project
design.






-46-


Project design should be influenced more by the implemen-
tation of the.host country than by the theoretical considera-
tions of the AID administration. As projects go through the
various clearance processes in AID/Washington, and each office
looks at them from its particular viewpoint, they tend to ac-
quire appendages that may inhibit their implementation. Bang-
ladesh has developed a project implementation document that
responds to the project document, but that is related to host
government procedures and uses government vocabulary. It may
be a useful model for other missions.

Project targets need to be realistic, attainable, and re-
lated to the real world and specific country conditions. Per-
haps this needs to be reiterated more often in Washington than
at the missions. The project designers need to have available
an appraisal of the farming technology used in the area and an
assessment of the policy and institutional framework of the
country. Documents such as the Country Development Strategy
Statement and the Social and Institutional Profile, when avail-
able, should be complemented with special assessments as nec-
essary.

Indicators of progress at various phases during project
implementation should correspond to the target projects. The
preparation of good baseline data and a regular monitoring of
project progress make it possible to assess progress toward
institutional and research goals and to revise these activities
during project implementation when inadequacies in planning or
unforeseen difficulties are encountered. This requires that
the project paper maintain some flexibility in the implementa-
tion schedule and program.

Scheduling of project activities as listed in the project
paper is often a cause for difficulty, especially those involv-
ing training and technical assistance. Training may need to be
started well before other project activities if trainees are
important to project implementation. Having available a prese-
lected pool of persons who have been cleared by their govern-
ment to receive training may speed the training process.


3. AID Management of Research Projects


AID's resources (particularly in-house talent and operat-
ing expenses) must be marshalled to support project managers in
the field. Often managers for country-level research projects
have insufficient technical experience and require backstopping
to do an effective job. They should have access to training,
technical assistance (including consultants), and research net-
works that permit them to draw on top expertise, both within
the country and externally. In regard to technical assistance,






-47-


closer relations should be developed between the international
agricultural research centers and the missions--perhaps on a
more formal basis.

The workshop participants, however, believed that reliance
by the project manager on technical backstopping should be only
a temporary stopgap. Better research-oriented training of man-
agement professionals should be the rule: generalists may not
know how to handle difficulties and crises in research imple-
mentation. Assignment of AID agricultural professionals should
be based on the appropriateness of their language skills,
training, technical specialty, and geographic experience.

Ideally, the AID manager should be assigned through the
life of a project. The mission participants to the workshop
also recommended that the AID manager spend more time on the
project site(s) rather than at the mission, and even live in
the project area, as should the host country manager.

Flexibility is essential in managing a research project;
however, this does not mean disorganization: an appropriate
management plan should be agreed upon with the host country and
enforced. The workshop participants emphasized that the
project manager must clearly and cogently communicate AID
regulations to the host country and to the AID contractor.
Difficulties too often arise because of lack of information and
communication among the host government, contractor, and AID
staff, yet it is essential, for the rules and regulations of
each institution involved must be respected and eventually
reconciled.

Host country managers and/or project leaders and donor
counterparts should meet periodically to take stock of imple-
mentation. The AID administrators and the AID agricultural
professional (project manager) should participate in these
periodic monitoring reviews along with their host country coun-
terparts. Efforts should be made to arrange these reviews so
as not to duplicate those already scheduled by host country
governments. Host country scientists and administrators should
make sure that reports of monitoring reviews reach the levels
of the research institution and government where plans are made
and funds are allocated.


4. AID Evaluation of its Assistance to Research


Research projects usually have a dual goal: they seek to
produce specific technological outputs as well as to develop
the institutions involved. Both are long-term goals and their
accomplishments cannot always be measured adequately within the
scope of the project.






-48-


Project design is the most critical factor in achieving an
effective evaluation program. The design of the project itself
is more important to evaluation than the design of the evalua-
tion per se. Project targets should be realistic. Overopti-
mistic targets make useful evaluation more difficult and
exacerbate the antagonisms inherent in evaluation. Without
flexibility in the project design, evaluation is much less ef-
fective: there is little point in recommending changes in a
research program if the project design does not have sufficient
flexibility to permit such mid-course corrections without a
major redesign effort. The project's institutional placement
affects the willingness of host country officials to partici-
pate actively in evaluations and in the project itself.

The evaluation design should consider not only the type
and scheduling of evaluations, but also the methodology to be
used, the composition of the team, and the necessary pre-
evaluation preparation. Project information systems must be
established from the beginning of the project in order to pro-
vide the raw materials needed for evaluation and project man-
agement. Data cannot be gathered by the team unless adequate
preparation is made.

Finally, evaluation is not an unmitigated good. Evalua-
tions can be disruptive and divisive as well as constructive.
This is particularly true when evaluation staff members do not
have a technical background sufficient to judge project
achievements. The workshop participants made four recommenda-
tions to increase the effectiveness of evaluation:

1. Participation by host country representatives, AID
mission personnel, AID/Washington managers, and outside experts
is critical to evaluation success, if they have the necessary
language skills and country experience. Host country partici-
pation is essential for meaningful evaluation, despite the po-
litical and technical difficulties that this may pose. Effec-
tive AID/Washington participation is hampered by the conflict
between its personnel's technical skills and administrative
duties.

2. Project design should establish a mechanism for sus-
tained evaluation attention. This may take the form of a peer
review committee drawn from host country, AID, and other
sources. It may also take the form of a contracted, informal
arrangement permitting a core group of individuals to be in-
volved in several evaluations over time (regardless of their
institutional location). This continued overview would in-
crease both the value of the recommendations made and their
acceptability to project staff.

3. Research projects should be flexible enough to allow
for changes during project implementation. Project control






-49-



must therefore be decentralized to allow the AID mission and
host country managers to respond constructively to evaluation
recommendations. The research process does not permit complete
planning, but requires a flexible response to opportunities as
they are identified.

4. Research evaluation requires an explicit methodology
and a carefully developed plan to guide team performance. The
overall guidelines for such evaluations should be revised and
made more available, but this does not obviate the necessity
for tailoring this design to specific needs and fully briefing
teams on the job they are expected to perform before they go
out.
































APPENDIX A

LITERATURE CITED






A-i



LITERATURE CITED


Bachman, Kenneth and Leonardo A. Paulino. Rapid Food Produc-
tion Growth in Selected Developing Countries. IFPRI No. 11,
1979.

Food and Agriculture Organization. Yearbook 1980.

Oram, Peter A. and Vishva Bindlish. Resource Allocations to
National Agricultural Research: Trends in the 1970's. (A
review of Third World systems.) IFPRI and ISNAR, November
1981.

Oram, Peter, et al. "Investment and Input Requirements for
Accelerating.Food Production in Low Income Countries by
1990," IFPRI No. 10, 1979.

Ruttan, Vernon W. Agricultural Research Policy. University of
Minnesota Press, 1982.

Ruttan, Vernon W. Bureaucratic Productivity: The Case of
Agricultural Research. UMEDC 82-3, 1980.

University of Minnesota, Economic Development Center. Bulletin
No. 81-2, March 1981.

University of Minnesota, Economic Development Center. Bulletin
No. 82-5, April 1982.

U.S. Agency for International Development. AID Experience in
Agricultural Research: A Review of Project Evaluations.
AID Discussion Paper No. 13, 1982.

U.S. Agency for International Development. AID Food and Agri-
cultural Development. Bureau for Program and Policy Coordi-
nation, May 1982.

U.S. Agency for International Development. Agricultural
Research in Northeastern Thailand. AID Project Impact
Evaluation Report No. 34, May 1982.

U.S. Agency for International Development. Central America:
Small-Farmer Cropping Systems. AID Project Impact Evalua-
tion Report No. 14, 1981.

U.S. Agency for International Development. Food Grain Tech-
nology: Agricultural Research in Napal. AID Project Impact
Evaluation Report No. 33, 1982






A-2


U.S. Agency for International Development. Guatemala: Devel-
opment of the Institute of Agricultural Science and Technol-
ogy (ICTA) and Its Impact on Agricultural Research and Farm
Productivity. AID Project Impact Evaluation Report No. 30,
May 1982.

U.S. Agency for International Development. Kitale Maize: The
Limits of Success. AID Project Impact Evaluation Report
No. 2, 1980.

U.S. Agency for International Development. Korean Agricultural
Research: The Integration of Research and Extension. AID.
Project Impact Evaluation Report No. 27, 1982.

U.S. Agency for International Development. Tunisia: The Wheat
Development Program (in preparation).

U.S. Agency for International Development. West Africa Rice
Research and Development. AID Project Impact Evaluation
Report No. 44, 1983.
































APPENDIX B

IMPACT EVALUATION TEAM MEMBERS AND WORKING GROUP MEMBERS






B-1


IMPACT EVALUATION TEAM MEMBERS


Kitale Maize: The Limits of Success
(Report No. 2)


Central America:
Cropping Systems
(Report No. 14)


Small-Farmer


Guatemala: Development of the
Institute of Agricultural Science
and Technology (ICTA) and Its
Impact on Agricultural Research
and Farm Productivity
(Report No. 30)

Korean Agricultural Research:
The Integration of Research
and Extension
(Report No. 27)


Charles W. Johnson, Team Leader
(Bureau for Asia)
Keith M. Byergo, Agronomist
(Bureau for Development Support)
Patrick Fleuret, Anthropologist
(Bureau for Program and Policy
Coordination)
Emmy B. Simmons, Agricultural Economist
(Bureau for Program and Policy
Coordination)
Gary Wasserman, Political Scientist
(Administrator's Office)

Harlan H. Hobgood, Team Leader
(Bureau for Development Support)
Rufo Bazan, Agricultrual Scientist
(InterAmerican Institute of
Agricultural Sciences)
Rollo Ehrich, Agricultural Economist
(Bureau for Development Support)
Francisco Escobar, Rural Sociologist
(University of Costa Rica)
Twig Johnson, Development Anthropologist
(Bureau for Program and Policy
Coordination)
Marc Lindenberg, Political/Institutional
Analyst (Development Studies Program)

Kenneth McDermott, Team Leader
(Bureau for Science and Technology,
Office of Agriculture)
David Bathrick
(USAID/Thailand)


David I. Steinberg, Team Leader
(Bureau for Program and Policy
Coordination)
Robert I. Jackson
(Bureau for Development Support)
Kwan S. Kim
(Bureau for Program and Policy
Coordination)
Song, Hae-Kyun
(Seoul National University)


Title


Members






B-2


Title

Food Grain Technology:
Agricultural Research in Nepal
(Report No. 33)


Agricultural Research in
Northeastern Thailand
(Report No. 34)










Tunisia: The Wheat Development
Program



West Africa Rice Research and
Development
(Report No. 44)


Members

Emmy B. Simmons, Team Leader
(Bureau for Program and Policy
Coordination)
Joseph Beausoleil
(Bureau for Science and Technology)
Gary Ender
(U.S. Department of Agriculture)
Gregory Heist
(Cornell University)
Josette Murphy
(Bureau for Program and Policy
Coordination)

Michael M. Calavan, Team Leader and
Anthropologist (Development Studies
Program)
James D. Wilson
(Development Studies Program)
John DeBoer, Agricultural Economist
(Winrock International)
Isara Sooksathan, Agronomist
(Kasetsart University)
Paitoon Rodwinij, Agricultural Economist
(Kasetsart University)

William F. Johnson, Team Leader and
Agricultural Economist (BIFAD)
Carl Ferguson, Consultant Agronomist
Mona Fikry, Consultant Anthropologist

James Van Dusen Lewis, Team Leader
(Bureau for the Near East)
Robert I. Jackson, Agronomist
(Bureau for Science and Technology)
Sidney F. Bowers, Agronomist
(Bureau for Africa)
Elon Gilbert, Agricultural Economist
(University of Colorado)
William Scott, Agricultural Economist
(Consultant)






B-3


AGRICULTURAL RESEARCH WORKING GROUP


The Agricultural Research working group was created in December 1981
to ensure participation of all AID bureaus in the preparation of the work-
shop. The participation of the following individuals in all or some of the
monthly meetings is gratefully acknowledged.


S&T/AG, Mr. Ken McDermott
S&T/AG, Mr. James Walker
S&T/RAD, Mr. Joseph Beausoleil
S&T/HR, Mr. Norm Nicholson
AFR/DR/ARD, Mr. William H. Judy
AFR/DR/ARD, Mr. Joe Hartman
AFR/DP, Mr. Hank Miles
ASIA/TR, Mr. Allen Hankins
ASIA/TR, Mr. Thomas Arndt
ASIA/TR/ARD, Mr. Richard Hughes
ASIA/DP/E, Mr. David Robinson
LA/DR/RD, Mr. Wayne Nilsestuen
LA/DR/RD, Mr. William P. Warren
NE/TECH, Mr. Robert Morrow
BIFAD, Mr. James Nielson
PPC/PDPR, Mr. Douglas Caton
PPC/PDPR, Ms. Emmy Simmons
PPC/PDPR, Mr. Gary Hansen
PPC/E, Mr. Robert Berg
PPC/E, Mr. Richard Blue
PPC/E, Mr. David Steinberg
PPC/E/S, Ms. Josette Murphy, chairperson
PPC/E/S, Mr. Twig Johnson
PPC/E/S, Ms. Charlotte Suggs

































APPENDIX C

Executive Summaries of Impact Evaluation Reports








KITALE MAIZE: THE LIMITS OF SUCCESS


AID first became involved with hybrid maize research in
Kenya in 1963, through the Organization of African Unity and
the East African Community. By 1970, the yield of the original
hybrids had been successfully improved by 25 percent under
research station conditions. The breeding program was
continuously followed with similarly positive results until the
EAC broke up in 1977. Other aspects of the A.I.D. program were
less rewarding. Research to improve maize protein quality and
to develop varieties for low rainfall areas did not succeed.
Nor did the attempt to train Kenyans and integrate them into
the research operation succeed. When the last American
scientist left almost 15 years after the first A.I.D. project
began, the effort was not sustained by Kenya.

In 1964, the first hybrid maize seeds were released for
commercial production. Hybrids produced a remarkable 40
percent increase in yield over local seed and proved
appropriate to the environment of the high potential areas of
Kenya, with their fertile soils, abundant rainfall, and
moderate temperatures. At the time, it was assumed that
African farmers would continue to use the local improved
variety rather than the new hybrid--it was less prone to crop
failure and it could be re-used year after year whereas hybrid
seed had to be re-purchased each year. But the hybrid was
clearly superior in yield, enjoyed the status of a crop used by
large farmers, and small farmers soon demanded it. By 1977,
the majority of smallholders in high potential Central, Rift
Valley and Western Provinces grew hybrid maize and their
production far surpassed large farmer output.

An aggressive private firm, the Kenya Seed Company,
reproduced the seed, distributed it, and promoted it throughout
the country via a network of private shopkeepers. Extension
agents demonstrated the use of improved cultivation
techniques. The government-supported official prices and
marketing system provided incentives, particularly for large
farmers, to adopt and profit by the hybrid technology.

Innovations are usually unfair in the sense they reward
those who have the means to benefit from them. Consequently,
it is not surprising that hybrid maize was of greater value to
those farmers with sufficient land, labor and capital to fully
utilize the innovation. More surprising is the large number of
smallholders who did gain access to the hybrid maize technology
and who have improved their food security as a result. The
overall impact of the increased maize production attributable
to the use of hybrid seed is that Kenya has continued to be
more or less self-sufficient in maize, the country's staple
food. As a result, Kenya, despite a very high rate of
population growth, has not had to face some food policy
problems which have confronted other developing countries.
Without hybrid maize, population pressure would likely have led






C-2


to a demand for more land for food crops and a reduction in
less essential export crops. Hybrid maize helped to keep the
price of food down in the cities, thus muting the pay demands
of urban workers and keeping Kenya attractive for foreign
investments.

There is a question, however, whether the government saw
the increased production of maize as more of a problem than an
opportunity. The government continued a pricing and marketing
system more suited to dealing with the problems of scarcity
than those of abundance. The Maize and Produce Marketing Board
responded to an obvious need for increased storage capacity,
for example, with too little, too late. Nor did the government
take adequate measures to ensure the continued success of
hybrids by: guarding the flow of critical inputs, including
sufficient credit and chemical fertilizers; and being
supportive of the research facilities which made the hybrids
possible. The loss of the incremental benefits which the
A.I.D. project demonstrated were possible by improving hybrid
seed year to year, cannot be calculated--but based upon the
benefits derived from the program in early years, the loss is
substantial.

Smallholders have not yet exerted policy influence on the
government (as did the European-dominated large farm sector
prior to Independence) by forming effective organizations of
their own. If government policy toward maize is to become more
effective, it will require not only better long range planning
but wider popular participation, especially among smallholders,
in its formulation.

From the experience of hybrid maize in Kenya and from the
observations of Kenyan maize growers and consumers, an A.I.D.
evaluation team drew seven key lessons:

1. Simplicity and viability were the decisive factors in
the success of hybrid maize.

2. The private sector was crucial in the rapid diffusion
of hybrid maize.

3. Perfect equity cannot be expected even from the most
successful technology.

4. The long-term continuity of foreign experts was basic
to the success of the breeding program.

5. Foreign advisors and finance do not automatically create
institutional capacity to perform agricultural research.






C-3


6. Pragmatism and skepticism should surround A.I.D. support
for regionalism.

7. Too many lessons should not be drawn from a unique
experience in one African country.






C-4


Central America: Small-Farmer Cropping Systems


The small-farmer cropping systems research project in
Central America was selected for evaluation as part of A.I.D.'s
effort to assess the impact of its activities in several
development sectors. Field work for the evaluation was done in
Costa Rica, Guatemala, Honduras and Nicaragua by a six-person
team in February 1980. The findings and interpretations are
those of the team and pertain only to this project. However,
they will contribute to a forthcoming analytical report for the
agricultural research sector as a whole.

In 1975, AID's Regional Office for Central American
Programs (ROCAP) began support to the Center for Tropical
Agricultural Research and Training (CATIE), located in
Turrialba, Costa Rica, to develop and test "a coordinated
regional research approach for improving the cropping systems
of small farmers in Central America." CATIE agreed to
negotiate working arrangements with the principal agricultural
research institutions of the five Central American republics.
These arrangements were to provide for CATIE and national
scientists to collect survey data on the cropping practices and
crop yields of the peasant farmers as well as data on their
socio-economic environments. Then the scientists were to work
with representative farmers by setting up experimental plots
designed to test and evaluate alternative crop combinations for
their potential in increasing production and income.

ROCAP undertook this project with the expectation that
CATIE would develop and demonstrate an innovative
multidisciplinary methodology for doing research on the
cropping systems of the small farmers of Central America. It
hoped to mobilize a permanent regional institutional capacity
and commitment for on-farm research and training addressed to
the needs of this vital sector of rural society. It also
expected to see CATIE produce, through the project, improved
cropping systems alternatives for different ecological zones of
the region that might be suitable to rapid verification and
dissemination by the national institutions. Its longer-term
goal was that as farmers adopted these proven, improved systems
the total yields from small farms would significantly increase
and family incomes would rise.

By the end of the project in 1979, CATIE had made working
arrangements and had carried them out in varied ecological
zones of all five of the Central American republics. Twelve
agricultural scientists from CATIE had been engaged full-time
in on-the-farm research. They had developed and demonstrated a
cropping systems research methodology working on the farms of
seventy-five smallholders. Impressive production gains and
potential economic benefits had been documented for the ten






C-5


major cropping systems alternatives elaborated by the project
staff. But these alternatives were yet to be verified through
extensive field trials in the region. However, one highly
promising alternative crop mix of sorghum and beans, which did
undergo limited verification, had been adopted by Nicaraguan
agricultural officials for widespread dissemination among
peasant farmers.

During this five year period, CATIE increased its graduate
training on small-farm systems and generated a five-fold
increase in its budget, largely from international donors and
almost exclusively for smallfarmer oriented agricultural
research activities using the "systems" approach. CATIE's
institutional commitment to improving small farmer production
had become well established as had its ability to work with
national institutions in the region.

Although the project had achieved most of its stated
objectives, the beneficial impact of the emergent research
methodology and of the expanded institutional capacity at CATIE
on large numbers of small farmers was yet to be demonstrated.
There was no wide-scale adoption of the newly tested cropping
systems alternatives developed from the on-farm experiments.
In spite of this and partly because of it, some lessons were
learned from the project evaluation.

Doing agricultural research on the farms of smallholders,
as opposed to research done on far-removed experimental
stations, holds much promise for the development of truly
appropriate production technologies and their more rapid
adoption and dissemination. But for that potential to be
realized, the projects should be designed to include the full
cycle of research through both verification and dissemination.
Donors sponsoring such research should provide the time and
resources necessary, perhaps eight- to ten-year authorizations,
to allow for validated technologies to reach numbers of small
producers. International or regional research institutions,
like CATIE, must be prepared to maintain their collaboration
with the national agencies, not only to support the
verification and dissemination phases as they come on line, but
to capture important findings during these phases for improving
subsequent research work.

Agricultural institutions undertaking on-farm systems
research must give adequate attention to non-agronomic
issues--such as input constraints, market analysis, and
household and area labor availabilities by season--in the
planning of the research, the analysis of constraints to
production, and the implementation of research, verification,
and dissemination programs. To do so requires that the
institution have adequate staff skills in the social sciences
and in farm management within the multidisciplinary teams
undertaKing each phase of the research effort.






C-6

Scientists need to be aware of the difference between doing
research on small farms and doing research with the active
interest and participation of small farmers. The former may
well inform the agricultural scientist about agronomic issues,
but only the latter is likely to educate both the scientist
about how the small-farmer household economy works and the
farmer about new agricultural options that will fit with the
economy. Several of CATIE's field staff demonstrated that
being a scientist and an involved participant, or even change
agent, are not mutually exclusive roles.






C-7


Korean Agricultural Research: The Integration
of Research and Extension


A profound change occurred in the early 1970s that
transformed the Korean Government's rural development
strategy. From one emphasizing industrial exports, the costs
of which were largely borne by the Korean farmers, the strategy
evolved into one devoted to improving rural Korean life. The
genesis of this approach was both political and economic: a
hardening of PL 480 terms and the results of the 1971 election
that amply demonstrated that government support had eroded in
the countryside. The Korean Government responded with a rice
pricing policy advantageous to the farmers, the strengthening
of the extension service, the formulation of the Sae-maul ("New
Village") Movement, and a rapid increase in rural
infrastructure.

The origins of AID's support to agricultural research are
found in the Korean Agricultural Sector Survey (1972) and
succeeding documents that advocated a strengthening of research
as a primary need. The project, proposed in 1973 and
implemented in 1974, provided $5 million for a tripartite
program to strengthen the capacity of the Office of Rural
Development of the Ministry of Agriculture and Fisheries. It
included training of Korean researchers overseas, equipment
(including a computer and library materials), and both resident
and short-term expatraite advisory services. At the close of
the project in 1980, 21 Ph.D. students and 17 M.S. students
were trained overseas, while an additional 94 received
short-term training and 106 participated in observation tours.

Although there were problems with the English language
competence of prospective students, the training aspects of the
project were universally regarded as the most successful part
of the program. Of notable, but secondary, importance was the
provision of equipment and supplies, especially the computer
and the library materials. Lagging far behind was the value of
resident expatriate assistance, which was of marginal use to
the project but was more significant in terms of relieving the
AID Mission from continuous monitoring of the project than in
providing help to the Koreans. Of greater importance was
shorter-term foreign technical advice.

The inchoate goal, from a Korean perspective, was probably
rice self-sufficiency--a strategic, political, and economic
objective. The project purposes, however, were specified in
considerable detail outlining exact yield increases on
agricultural experimental stations over a ten-year period in
the areas of rice, barley, wheat, and soybeans as well as
generalized improvement in potato production and in the
cropping systems. Specific increases were also proposed for






C-8


farm fields for the same time. Since the decade of crop
improvement is to end in 1984, this evaluation must be
somewhatcircumscribed.

The project paper suffered from spurious specificity
regarding experimental station crop increases. Before the
project began, experimental yields were higher than those
indicated in the paper, often by considerable amounts. The
research breakthroughs that the project anticipated were
generally made prior to the project. Farmer yields may well
reach their objectives by 1984, but the AID project was only a
beneficial increment to Korean agricultural research. It
supplemented an existing, competent system, but offered little
that was innovative.

The concentration on rice led to a lack of emphasis on
other crops, an inattention caused by national concerns as well
as social and economic factors the project ignored. Although
there have been increases in crop yields, hectarage of the
other crops has consistently been falling, even before the
project began. Thus, national targets will not be met even if
a relatively few farmers benefit. The choice of some of the
crops covered by the project such as wheat, soybeans and
potatoes seems questionable, as does the emphasis on increased
fertilizer responsiveness.

Critical to a developmentally effective agricultural
research program is the transference of experimental results to
the farmers. Through a widespread extension service, a farmer
training program that includes almost all families annually,
demonstration plots, and the Sae-maul Movement, Korea has
developed an authoritarian but effective means of disseminating
research results.

Thus, beginning in 1972 the spread of the high-yielding
varieties of rice was pushed with alacrity by the Korean
bureaucracy in response to a national command structure. The
effort was effective, making Korea self-sufficient in rice by
1975. Yet there were two inherent problems in this
comprehensive effort: these varieties were sensitive to cold,
and new races of the fungal disease called blast normally
develop after a few years if large areas are planted to a
single variety.

The crisis developed first in 1979 with a drop in
production caused by blast followed by a disastrous 1980 crop
due to cold temperatures. The rice crop fell by one-third,
creating a crisis of confidence in the government and in the
guidance service.

Ironically, the failures of 1979 and 1980 can be attributed
to the strengths of the Korean guidance service. Thus its






C-9


weakness is based on the omnipresent bureaucratic hierarchy
that, in contrast to most developing societies, can transform
research into production. In singleminded pursuit of its
political goals, it neglected elemental precautions that might
have avoided the problems of the last two years.

Agricultural research was an appropriate intervention for
AID at the time. It assisted a well-established, agricultural
research network, but did not materially transform it. It
created no new institutions.

Agricultural research will continue in Korea but
replication abroad will be difficult. Any successful adaptive
agricultural research project will be dependent upon a positive
pricing policy, an effective extension service, rural
infrastructure, and continuous contact with international
research centers, among other factors. Political will is
required for its success, but too strong an emphasis on
political objectives can undercut its effectiveness.






C-10


Guatemala: Development of the Institute of Agricultural
Science and Technology (ICTA) and its Impact on
Agricultural Research and Farm Productivity


During the decade of the sixties, food production in
Guatemala barely kept pace with the demands of a growing
population. In 1970, the Government of Guatemala initiated a
restructuring of public agencies to provide coordinated service
to small food-producing farms. An innovative organization, the
Institute of Agricultural Science and Technology (ICTA),
emerged from this restructuring with responsibilities for
generating and promoting the use of improved technologies in
basic food crops. AID supported this restructuring with a
series of loan and grant projects beginning in 1970.

In 1975, AID approved the Food Productivity and Nutrition
Project. Its purpose was to increase the production and
nutritive quality of basic food crops in Guatemala and to
strengthen and develop ICTA as an institution. Of $1.73
million allocated for the project, $1.2 million was for
expatriate technical assistance, including plant breeding
experts and other technicians who staffed ICTA while
project-sponsored Guatemalans were being trained to assume
positions within the new Institute.

Three crops, maize, beans, and sorghum, were targeted for
increased production. Working with experts from international
agricultural research centers, ICTA personnel developed new
varieties and tested them under small farm conditions by
collaborating with farmers. With the assistance of the
Inter-American Development Bank, a seed service was organized
to process seed and help maintain genetic quality.

New varieties of both maize and beans were introduced and
increased yields have been recorded. Using improved seed and
other technologies recommended by ICTA, collaborators have
obtained increased yields. Gains in maize have been primarily
in lowland varieties, but one new highland variety is
promising. The impact of new seed on maize production is
expected to increase as the amount of seed produced increases.

New varieties of beans may reduce or eliminate the need for
costly programs to control Golden Mosaic. New varieties of
sorghum were not released until 1980 and thus could not be
evaluated. However, they appear markedly superior to
previously available varieties.

In addition to developing and recommending improved seed,
ICTA developed and recommended other farming practices related
to increased yields, such as planting distances, seed
densities, fertilizer applications, and weed and insect






C-ll


control. Indices of acceptance developed by ICTA indicate that
increasing numbers of farmers who have collaborated in the
fieldtesting of such new technologies are adopting ICTA
recommendations. Interviews with ICTA personnel and with
individual farmers support this impression.

The AID project facilitated and hastened the strengthening
of ICTA as an institution. The number of ICTA staff increased
and staff qualifications improved. Expatriates facilitated the
research work of ICTA and its growth as an organization. With
project support, 10 Guatemalans received advanced training and
by 1979 and 1980, they were returning to ICTA to replace
expatriates.

However, high attrition rates among personnel with advanced
degrees are a serious problem for ICTA. Rigid salary schedules
are apparently responsible, but ICTA managers have been
unsuccessful in efforts to obtain the authority to revise these
schedules. With the departure of expatriate advisors, these
high attrition rates may make sustaining and expanding the
present ICTA system more difficult.

Some confusion remains regarding the respective role of
ICTA and DIGESA, the extension service of the Ministry of
Agriculture, particularly as ICTA's approach to research draws
on some techniques of traditional extension methodology. ICTA
and DIGESA are working on this problem, and it seems likely
that new patterns of relationships will develop.

ICTA has come to represent a new model for agricultural
research that planners and researchers in other countries are
studying and attempting to replicate. If there is continued
and increased support from the Government of Guatemala, it will
be able to sustain and expand its present activities.






C-12


Food Grain Technology: Agricultural Research in Nepal


In 1957, the U.S. Operations Mission initiated support for
a broad-ranging agricultural development effort in Nepal. This
project continued without pause for 17 years, largely in
pursuit of the objective of increasing Nepal's food grain
production capacity by enabling and encouraging Nepali farmers
to apply the techniques of scientific agriculture. While the
U.S. financial and technical assistance was continuous, the
emphasis, the pace, and the amount of Nepali involvement were
altered considerably during the course of project
implementation. The project began as a "general agriculture"
initiative and gradually evolved to its concluding emphasis on
the development and dissemination of "food grain technology."

The project successfully contributed to the establishment
of agricultural research and extension systems by training
almost 600 Nepalis to the B.S., M.S., and Ph.D. levels and by
constructing facilities for research at five stations in the
Tarai--at Nepalganj, Bhairawa, Parwanipur, Janakpur, and
Rampur. With the assistance of the extension service, improved
wheat, rice, and maize varieties that were tested on the
research stations were spread to farmers across the Tarai.
Some of the selected improved varieties proved widely adapted
to Nepal's enormous range of agroecological conditions and
spread into the Hill and Mountain farms as well. Other parts
of the "technology packages"--which included recommendations
for fertilizer, time of planting, spacing, and irrigation--were
not so widely adopted.

In trying to assess more precisely the differences that
could be attributed to the implementation of the Food Grain
Technology project, we first examined statistical fact sheets
and research reports. We then talked with agricultural leaders
(many of whom had apparently taken advantage of training
opportunities offered under the project) and with agricultural
producers. We took a long view in these dialogues, trying to
comprehend the pattern of changes which had occurred in the
agricultural sector over the past two decades. While looking
at reports of experimental trials and at growing fields of
wheat and mustard, we discussed not only what had happened, but
what might not have occurred had the project never been
implemented.

Our examination provides both a sense of solid
accomplishment and a basis for some disquieting fears. On the
positive side, we found the following:






C-13


A functioning research system has been developed.

Farmers are immensely aware of the need for and the
problems related to Krishi bikash (agricultural
development).

Extension and research services can, at times, work
together in complementary, mutually reinforcing
activities which result in new plant varieties and
increased knowledge in the countryside.

On the negative side were the following factors:

Researchers and farmers are not in complete agreement
on which agricultural problems need to be addressed,
nor are the channels for communication as open as they
could be.

The "green revolution" as it has occurred in Nepal has
not yet resulted in long-term security and economic
independence as expected but has contributed to
economic and environmental destabilization.

The productivity of farmers, extension workers,
researchers, and those agencies charged with input
supply distribution is far from optimal.

Thus, researchers articulate the need to continue the
search for new varieties which are higher yielding, more
disease resistant, and produce grain with acceptable qualities
of taste. Farmers agree that variety development is important,
although they emphasize other criteria for variety selection as
well. Farmers also recommend that increasing reliability of
water and fertilizer supplies is more important for handling
their problems of deteriorating soil fertility, declining farm
sizes, with low yields, and high risks. The role of
agricultural research and extension is not in question; at
stake are the issues of research priorities and their relevance
to farmers' resources and constraints.

The fact that farmers have adopted components of technology
packages at all may reflect less the persuasive rhetoric of
research and extension than the farmers' response to the
increasing pressure of population and to their families'
requirements for food and cash. Nevertheless, without the
technology packages, it is unlikely that Nepal's farmers would
be as productive as they are today.






C-14


Agricultural Research in Northeastern Thailand


In 1962, the Ministry of Agriculture and Cooperatives in
Thailand officially established an agricultural research center
at Tha Phra near Khon Kaen, located 400 kilometers from
Bangkok. The center was to be a multidisciplinary research
facility focusing on the Northeastern region and responsive to
the needs of the farmers. In addition, it was to support and
coordinate the work of the Ministry's 112 small research
centers and stations in Northeastern Thailand.

USAID/Bangkok first assisted this project in the mid-1960s
by providing graduate training to 24 Ministry employees who
were to staff the center. In 1966, a multifaceted project was
launched for institution-building at the center. A contract
was signed with the University of Kentucky, Lexington Kentucky,
and from 1966 to 1975 Kentucky Project officials were
responsible for (1) advising center administrators; (2)
arranging for training employees in the United States; (3)
assisting in the establishment of research laboratories,
research programs, and extension activities; and (4)
coordinating functions at the center.

An excellent physical facility was constructed which has
been carefully maintained. Since 1966, a total of 118 Ministry
employees have received U.S. training in agricultural
disciplines mostly at the University of Kentucky. By 1975,
laboratories were well established and substantial research
work was underway. However, since 1975, research programs have
been reduced and the professional staff of the center is far
below projected numbers. The research carried out is
essentially conventional and laboratory- or station-focused;
there is little evidence that it is responsive to the needs of
small farmers in Northeastern Thailand.

Kentucky Project extension and training activities started
slowly, but since 1975 several initiatives have been launched.
These include a series of television and radio programs, a
mobile extension unit, and an agricultural information
network. These initiatives were not planned at the beginning
of the project. However, at the time of review, these
activities and their support units were the most dynamic at the
center. Modest USAID support to these programs could do much
to enhance the quality and quantity of agricultural information
available to Northeastern farmers.






C-15


Scientists at the center need to familiarize themselves
with the complexities of agricultural production and
decision-making in the Northeast. This could contribute to
future research activities and outreach programs which are more
relevant to the needs of a greater variety of farmers.
Furthermore, bureaucratic conflict has created an atmosphere in
which much research done at the center is rejected out of hand
by the central Ministry of Agriculture and often has to be
redone in order to be acceptable. Declining budgets, loss of
coordinating authority, frequent institutional redefinition,
and loss of status and professional autonomy have combined with
previously mentioned factors to defeat efforts to build a major
research capacity in Northeastern Thailand.

Ministry, USAID, and University of Kentucky Project
officials chose not to reexamine and reformulate the project,
inspite of ample, early evidence that the center lacked
sufficient bureaucratic potency to accomplish its long-range
goals. It seems unlikely that more detailed planning could
have pinpointed and overcome this problem. However, AID
officials should have recognized the problem by the late 1960s
and done something about it. They could have
(1) pulled out, (2) decided to support only the most promising
portions of the project (e.g., the training component), or (3)
worked with the Ministry to strengthen the bureaucratic
position of the center. That none of these things happened
reflects negatively on responsible USAID officials, but perhaps
more so on AID structures and procedures. These may have
discouraged Mission officials from reexamining projects and
making mid-course corrections 10 years ago. Whether or not
there have been sufficient changes in incentive structures to
encourage them to do so today remains to be seen.






C-16


West Africa Rice Research and Development


The West Africa Rice Development Association (WARDA) was
created in 1970 to increase rice production in the 15 member
countries through research and training. Importation of the
rice necessary to satisfy an increasing demand for what is
becoming the food staple in urban areas is a drain on foreign
exchange, yet the climate and ecology of West Africa are suited
to rice production.

A decade after its creation, one cannot hold WARDA
responsible for the fact that West Africa is importing more
rice than ever. WARDA was encouraged to look for technological
solutions to this deficit, not for economic policy solutions.
But a technical solution cannot be divorced from its economic
environment. One of the greatest weaknesses of WARDA's
research design is its tendency to separate these two. Some of
WARDA's research results demonstrate the disadvantages of this
tactical separation, laid on the association by its founding
charter and by the orientation of the donor and member state
support it has received. Nevertheless, because of its
scientific professionalism, WARDA, through its development
department, has discovered a politically acceptable way of
targeting project identification research design on specific
situations that are not only ecologically but also economically
conducive to expanded rice production.

Much of the more recent, second phase of AID support to
WARDA (project 698-0429) is built upon WARDA's evolving skill
in contextualizing rice research and development inputs such
that, for specific contexts, their outputs are not hindered by
the widespread economic constraints on rice production in West
Africa. Therefore, with the advantages of hindsight,
therefore, we are evaluating the first-phase AID/WARDA project
(698-0382), not only in terms of its own stated goals, but also
to identify the part it played, if any, in helping WARDA define
this more successful, interdisciplinary role for itself.

Under the first-phase project, AID supported (1) two
special research projects--one for mangrove rice at Rokupr,
Sierra Leone, and one for deepwater/floating rice at Mopti,
Mali; (2) a training center adjacent to Liberia's Agricultural
University at Fendell just outside of Monrovia; (3) participant
training in the United States for key WARDA researchers; and
(4) a rice economics study undertaken in conjunction with the
Food Research Institute at Stanford University.






C-17


Tunisian Wheat Development Program


The Tunisian Wheat Development Program (Project Ble) was
designed and implemented from 1965 to 1977 by AID, the Ford and
Rockefeller Foundations, the International Maize and Wheat
Improvement Center (CIMMYT) in Mexico, and the Government of
Tunisia. It was conceived in 1965 at a time when the economic
chaos following independence from the French prompted the
Government of Tunisia to explore every avenue to reverse the
decline in agricultural production, particularly of food.
Development of Tunisian institutions and training of Tunisian
staff were priority goals to fill the gap created by the exodus
of the French civil servants and other European farmers and
entrepreneurs in 1964. The ultimate goal of the Government was
and remains "self-sufficiency in food production."

The purpose of the program was to introduce and adapt to
the Tunisian environment and climate the new semi-dwarf
high-yielding wheat varieties that had been developed at CIMMYT
in Mexico. The other important purpose of the program was to
train Tunisians in agricultural research and extension methods
as a means of developing institutional capabilities for Tunisia
to carry out research and extension activities alone.

The impact of the program has been slow but positive. Much
of the impact is being felt now, some five years after the
program was phased out and 17 years after its conception. If
one single factor had to be identified as the program's most
important contribution, it would be the development of the
program for advanced degree training, particularly to the Ph.D.
level. The research capability developed by this advanced
training has become most effective in the past three years.
The impact is being demonstrated in research results; in an
effective extension program; in improvements in institutional
capabilities in research, extension, and education; and in
farmers' increased acceptance of new varieties and improved
technology, resulting in increased yields and production.

Training has enabled Tunisians to successfully continue
research and extension activities without assistance after the
program was phased out. Nineteen Tunisians were trained in the
United States to the level of M.S. and Ph.D. degrees in
agricultural sciences. This was supplemented by practical
training of 55 other Tunisians at CIMMYT in Mexico, in
Australia, and in Tunisia. Of the 19 who received advanced
training, 13 are working directly or indirectly in the cereals






C-18


program in Tunisia; 11 of these are directly involved. Of the
19 Tunisians, one is continuing advanced studies in the United
States and five are working abroad with international or other
organizations. Four of the Tunisians who received Ph.D.
degrees are involved in research at the National Agricultural
Research Institute of Tunisia (INRAT) while teaching at the
National Agricultural Institute of Tunisia (INAT, the national
agricultural university). Two Tunisians trained to the M.S.
level are participating in the research program at INRAT.

During the life of the program, five new bread wheat and
five new durum wheat varieties were developed and introduced to
farmers with varying degrees of success. After the program was
terminated in 1977, Tunisians had been trained under the
program continued to develop varieties with characteristics that
improved on those developed earlier. In 1980 and 1981, two
improved varieties of bread wheat and two improved varieties of
durum wheat were developed and put into use. Some of these
later varieties were more resistant to diseases and drought
than earlier varieties, and consequently were more acceptable
to farmers.

An extension and farm demonstration system and program were
developed in the beginning of the Wheat Development Program to
work closely with the research activities to extend results to
farmers and to feed back problems to research scientists. The
Technical Division, established in the Office of Cereals,
successfully carried out its functions during the life of the
program. It is now staffed with trained Tunisians and is still
operating a successful program.

As a result of the program, Tunisia's cereal production
(wheat and barley) was greater during the 11-year period 1971
through 1981 by 5.302 million metric tons than during the
previous 11-year period. Despite population growth, annual per
capital production of cereals increased from 104 kilograms in
1970 to 160 kilograms in 1980, using average annual production
figures for the two periods and the population levels of 1970
and 1980, respectively. Furthermore, the increased production
was achieved on an area of land less (by over 200,000 hectares
in each year, 1980 and 1981) than in the previous four years.
The increased production of cereals saved the Government of
Tunisia the foreign exchange costs of annually importing
299,000 metric tons of durum wheat, 77,000 metric tons of bread
wheat, and 106,000 metric tons of barley that would have been
required otherwise during each year 1971 through 1981. The
value of this amount of annual imports at 1981 prices would
have been $125,944,000 (cost, insurance, and freight in
Rotterdam, imported from the United States). This was made
possible at a total cost to the U.S. Government, Rockefeller






C-19


and Ford Foundations, and less than $3.5 million in technical
assistance.

The program has resulted in other benefits to Tunisia. It
contributed to increased per capital consumption of cereals,
mostly in the form of increased use of commercial bread and
noodles. While no national data were available to confirm the
fact, there was evidence that farmers had been integrated into
the money economy. Cereal farming had become mechanized and
farm families were purchasing prepared foods such as commercial
noodles and bread.

The positive impact was not without some negative effects.
Rural migration of men had led to a change in the role of rural
women, with an increase in their participation in farming and
rural industries, and a decrease in their role in home
preparation of food. While this may be viewed as a positive
gain for women, it has had one negative result. Increased use
of purchased, prepared foods (principally noodles and bread)
instead of home-prepared food has decreased the nutritional
levels of farm family daily diets.

Not all the institutional goals have been achieved.
Integration of research and extension has not been acted on.
The planners had sought flexibility in management, financing,
decision-making, and action by establishing the program under
the parastatal, semi-autonomous Office of Cereals, a commercial
organization concerned with the purchase and sale of cereals.
This office, which is outside the Agricultural Services of the
Ministry of Agriculture, was not impeded by the bureaucratic
constraints of other agencies. At the same time, it did not
play a role in providing technology to farmers. During the
life of the program, activities were integrated through
personal cooperation of scientists who cut across institutional
lines. This system continues today.

Despite these weaknesses, the institutions in research,
education, and extension have developed basic capabilities,
resulting directly and indirectly from the program, which
permit them to continue successful activities. However, the
goal of self-sufficiency in food production has not been
achieved. This goal is illusory and has tended to overshadow
the progress that has been made, as continued growth of
population and increased per capital consumption of cereals have
widened the food gap, requiring increases in imports.
Tunisia's overall goals of using its resources to comparative
advantage, and of producing higher valued crops on the better
land (under irrigation where feasible) for export and to supply
the thriving hotel-tourist industry are both aimed at achieving
a balance in international trade of agricultural products,
which makes good economic sense. Achievements in cereal






C-20


production are due not only to the scientific progress achieved
under this program, but also to improvements in institutions,
economic conditions, and policies in the agricultural sector.






C-21


AID EXPERIENCE IN AGRICULTURAL RESEARCH
A Review of Project Evaluations


This study reviews the experience of the U.S. Agency for
International Development (AID) in the area of agricultural
research. It was completed by Development Alternatives, Inc.
(DAI) at the request of AID's Office of Evaluation, Bureau for
Program and Policy Coordination (PPC/E). The study's
objectives were:

To review historical trends in agricultural research,
especially of AID's expenditure in that sector;

To identify the set of projects comprising AID's
agricultural research portfolio; and

To identify major issues affecting the design and
implementation of agricultural research projects by
reviewing evaluations of a sample of those projects.

A review of the literature and interviews with various
professionals identified several recent trends in agricultural
research. These included an increasing attempt by researchers
to develop technology applicable to the needs of farmers under
adverse envrionmental conditions and in resource poor regions
of the world. Moreover, in an attempt to better align research
with farmer needs, a broader array of production constraints
(both agronomic and socioeconomic) is now being examined in the
technology generation process than in the past. This has
entailed more emphasis on on-farm research, the use of
multidisciplinary teams and a more holistic approach to
research, as well as greater participation by the farmers
themselves in the technology generation process. Additional
issues receiving increased attention are the importance of
strong national research systems and the amount of time
necessary for agricultural research projects to produce useful
results.

AID support to agricultural research has been increasing in
recent years. Historically, however, the sector has received
relatively little attention from the Agency. According to the
interviews and literature review conducted during this study,
one reason for this lack of attention was the belief, prevalent
in the early 1950s, that the technology necessary to improve
agricultural productivity in the developing countries already
existed. Limitations during the 1960s included Congressionally
imposed restrictions on the amount and type of research that
AID could undertake together with decreases in the Agency's
in-house technical expertise in agriculture. Finally, the New
Directions legislation passed in the early 1970s, while
contributing to important changes in the nature and focus of
AID's agricultural research, emphasized other development
strategies such as rural development and food production






C-22


projects, or the delivery of services to meet basic human needs.

AID's increasing interest in agricultural research in
recent years has partly resulted from a realization that a lack
of appropriate agricultural technology is a serious constraint
to food production increases. Moreover, the success of the
green revolution technology developed by the international
agricultural research centers (IARCs) in increasing production
levels of selected crops in certain regions of the world has
furthered this realization.

Between 1978 and 1981 AID funds going to agricultural
research increased by almost 70 percent, from $84.7 million to
$143.7 million. In relative terms, AID's investment in this
sector rose from 12.8 to 19.5 percent of the agriculture, rural
development and nutrition appropriation (excluding economic
support funded appropriations). Most of this increase came
from projects funded by AID field missions. On the other hand,
the proportion of AID support going to the IARCs and centrally
funded bilateral research has increased only slightly.
However, the passage of Title XII and the creation of the Board
for International Food and Agriculture Development (BIFAD) may
provide a basis for greater activity in this area.

Aside from reviewing historical trends in agricultural
research, the study examined issues affecting projects in the
sector based on a review of 131 evaluations of 48 agricultural
research projects (39 regionally and mission-funded and 9
centrally funded). It found that the evaluation documentation
provides only an imperfect picture of any project's overall
performance. The evaluations were most often focused on the
provision inputs and the achievement of outputs. Attempts to
measure project impact (to determine the effect of project
activities on the beneficiaries welfare) were limited to the
four Impact Evaluations included in the sample (part of a
series of indepth, ex post evaluations currently being
undertaken by AID). The standard evaluations did not provide
the basic information (such as project characteristics and
standardized performance indicators) necessary to permit a
comparative analysis of the projects in this sample.

Using the evaluation documents, it was possible to identify
several recurrent issues common to projects in the agricultural
research sector. For regionally and mission-funded projects
these included:

Operational problems entailed in doing on-farm,
farming systems-type research, and involving farmers
in the research process;

The quality of the research conducted and the setting
of research priorities;






C-23


The phasing of activities, especially construction
delays which impeded planned research, as well as the
amount of time allowed to achieve the research
objectives;

The adequacy of AID's research project supervision,
given a lack of technical expertise and high staff
turnover in the missions;

Weaknesses in the links between research and extension,
as well as inadequacies in complementary services
(inputs, credit, marketing, and so forth);

Host government support for the projects;

The lack of qualified counterpart personnel to work
with expatriate technicians, together with low
salaries for host country researchers which makes it
difficult to maintain competent staff;

Inadequate participant training programs;

Delays in procurement; and

The delays or inability of AID and its contractors to
provide qualified technical assistance.

For the nine centrally funded projects in the sample (each
of which involved overseas research), the issues discussed in
the evaluations included: the creation of linkages with host
country institutions; the performance of long-term staff; the
project's scope and funding; and the quality of the research
conducted. Issues not fully treated by the evaluations of
these projects included: the problems entailed in simply
conducting research within developing countries and in
conjunction with local institutions and researchers; the
feasibility or necessity of conducting more research away from
the research station; and the dissemination of the research
findings.

In conclusion, this review of past AID evaluations
identified and documented a set of issues or problems that were
more or less familiar to development professionals
knowledgeable about the sector. The study also identified
significant gaps in the evaluation data base that was
analyzed. In terms of producing information that might
influence overall policy within the sector and feed into the
design of future projects, this study highlighted the need for
investigations outside the Agency's system of regularly
scheduled evaluations in assessing its project implementation
experience.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs