January 21, 1981
Here, at long last, are the collected papers from the December
8 and 9 Farming Systems Research symposium. The comments we
received from many of you indicate that the symposium achieved
the goals we had set for it. We thank all participants for the
contributions they made to its success,
The hope was expressed during the meetings that the information
exchange which the symposium initiated could be continued, We
will be considering ways in which that can be accomplished, and
we will be contacting you for your input of suggestions,
ohn D. Hyslop
technical Assistance Officer
Technical Assistance Division
FARMING SYSTEMS RESEARCH SYMPOSIUM
December 8 and 9, 1980
List of Papers
Introduction and Objectives, Quentin M. West
Overview of Farming Systems Research, Donald L. Plucknett
An Overview of Farming Systems Research Methodology, Richard R. Harwood
Overview of the Potential Applicability of Farming Systems Research to U.S.
Small Farms and U.S. Research and Extension, Art Hansen
The Allegheny Highlands Project, Barton Baker
Evaluation of Farm Systems Research in Relation to the South Central Small
Farm Research and Extension Center Program, J.R. Gifford
Applicability of the Farming Systems Research Approach to Less Developed and
Developed Countries Linkages and Constraints, Michael S. Joshua
Research-Extension Interface, Howard W. Kerr, Jr.
Initiating Applied Farming Systems Research in Developing Countries,
Problems of Interdisciplinarity in Farming Systems Research, Randolph
Farming Systems Research Program Structure, Staffing and Funding,
FSR and National Agricultural Development, Peter E. Hildebrand
Issues of FSR Evaluation, Willis W. Shaner
The ICTA Case; Overview of ICTA Program, Astolfo Fumagalli C.
Farming Systems Research in Guatemala, Ramiro Ortiz Dardon
Staffing and Funding in the Instituto De Cienciay Tecnologia Agricola (ICTA)
Robert K. Waugh
The ICTA Case; The Linkages with Other National, International and
Regional Institutions, Porfirio Masaya
Discussion Group Reports: Summary
Lessons for the Future
Summary of Tony Babb's Closing Remarks
Farming Systems Research Symposium Participants
December 8 & 9, 1980
SYMPOSIUM ON FARMING SYSTEMS RESEARCH:
INTRODUCTION AND OBJECTIVES
Quentin M. West*
INTRODUCTION AND OBJECTIVES
U.S. overseas technical assistance programs are grounded in methods and
expertise developed and applied here at home. This is an assumption underlying
all such work. It was explicit in the Point-Four days of the 1940's and 1950's
and, though not emphasized, remains as part of our philosophy today.
The notion is particularly strong in technical assistance in agriculture.
Variants of the U.S. model for agricultural research and extension have been
applied in dozens of developing countries during the past 30 years. Though
this application has contributed to significant increases in agricultural
productivity and rural incomes in a number of countries, the benefits have
often been unequally distributed toward larger farmers having access to
required services and inputs. For small farms in most countries the applica-
tion has failed. Research results have become laboratory shelf items, never
reaching the farmer or bearing little relevance to the constraints under which
The applications have failed because they have been largely based on the
notion that technological progress in agriculture is a monotonic sequence
from the laboratory to the test plot to the extension agent to the farmer.
It is a highly over-simplified version of our research and extension system.
It does not take into account the economic, social, and institutional environ-
ment in which U.S. farmers operate.
The farming systems approach to agricultural research and extension has
arisen out of the realization that too many of the necessary elements of
the U.S. model, of which public research and extension services are only a
part, are missing in developing countries. Among the missing elements are
efficient factor and product markets in which the effect of resource constraints
is made explicit through movements in relative prices.
The farming systems approach is a product of the imagination and professional
skills of researchers and extension workers in these countries. It is an
attempt to compensate for the lack of an environment supporting technological
progress in agriculture. Though it is also identified with the work of a
number of the international research centers and is closely related to farm
management work as practiced in the United States and elsewhere, its full
development and application is thus far clearly and uniquely a developing
*Director, Office of International Cooperation and Development, U.S. Department
We obviously have here an activity in which U.S. technical assistance would
not be grounded in methods and expertise developed and applied here at home.
The interest shown in this symposium and the agenda's list of participants
suggests that the flow of technical assistance should perhaps be in the
This came home to us in OICD just over a year ago when we were asked to
field a USDA team to explore possible technical assistance and cooperative
relationships in agricultural research with one of the AID "graduate"
countries. The request specified that a USDA expert in farming systems research
be included in the team, and that he be prepared to identify U.S. universities
having a strong farming systems research program.
It didn't take too many phone calls to determine that there was no such
expert in USDA (certainly not in OICD) and there appeared to be no strong
farming systems research program in any U.S. university.
I know that the truth of that assertion depends partly upon one's definition
of farming systems work. I will not offer a definition now, since we are
going to be working together on one over the next two days.
With that caveat, it still seems that the situation remains true today,
though a number of the universities are beginning to establish farming systems
programs or are strengthening programs which were just beginning last year. In
USDA the seminal programs in small farm research in SEA's Northeast and
Southern Regions cannot yet be called an expertise in farming systems research.
They are a nucleus around which a USDA expertise may develop. This depends
largely on the extent to which the methods of the farming systems approach are
perceived to be applicable to the small farm programs.
Here we have both our initial motivation for, and a major objective of,
this symposium: The increasing interest in the farming systems approach in
developing countries makes it imperative that our USDA technicians on over-
seas assignment be knowledgeable about the concept. Further, this symposium
is a good starting point for USDA's exploring the applicability of the farming
systems approach to U.S. small farm research and extension. These programs
were started out of the belief that the main thrust of our research and extension
work is not adequately serving the small farm segment of our rural population.
There are, of course, other objectives. The Office of Agriculture in AID's
Development Support Bureau, our third co-sponsor, is planning a project of
support to farming systems research and extension programs in AID-assisted
There are still a number of gaps in the plan for such a project which this
symposium will help to fill:
1. Is the farming systems approach cost effective?
2. Are there general methodological or implementation issues
in the farming systems approach which must be resolved before a
broad-based assistance project is feasible?
3. What are the important resource constraints developing
countries are facing in attempting to implement farming
4. Does U.S. expertise to address these constraints exist
in sufficient quantity to give an assistance project
5. What form of U.S. assistance is most appropriate to the
development of farming systems programs in developing
There are some of the questions you will be considering during these two
Equally important, the Office of Agriculture is concerned that AID's
agriculturists be familiar with the concept of farming systems research and
extension. Though the general level of familiarity with the concept probably
is higher in AID than in USDA, due to contact with programs in the field, the
immediate need-to-know level is also higher. It is AID that is the point of
contact with the developing country programs.
Finally, USDA has the objective of looking to the future, so far as our
agricultural research and extension agencies are concerned. Requests for
assistance to farming systems research and extension programs are certain to
increase. What is the response the universities and USDA will and should
make? Should we be thinking in terms of some sort of co-ordination? What
sort? We'll have the opportunity later to discuss these questions directly.
I agree with the assertion that we do best in technical assistance overseas
those things which we do well at home. However, we must always remember
that the overseas conditions are not those existing here.
The farming systems approach may be the very device for making that assertion
true in research and extension. This symposium will help us make that
Overview of Farming Systems Research (FSR)
Donald L. Plucknett
Modern farming systems research (FSR) is becoming an accepted approach
to applied agricultural research. FSR was developed because of concerns
that farmers were not adopting research innovations being developed by ex-
periment stations. Also, in recent attempts to improve the lot of small
farmers, an awareness has developed that there is inadequate understanding
of small farmers and their problems.
In a real sense, FSR is not new. Some of the farm management research
in agricultural economics in the USA in the past did involve some aspects
of FSR. However, a truly multidisciplinary effort involving experimentation
in a systems mode on the farm and in focused research on systems problems
on experimental stations did not result until recently with the rise of
Modern FSR can be traced to a group of individuals or institutions,
mostly working in isolation, who began to try to understand several things:
(1) how relevant information could be generated for small farm systems,
(2) multiple cropping principles and potentials, and (3) the potential pro-
ductivity of tropical areas. The following examples cited are meant to be
illustrative of the range of efforts that have led to modern FSR. Space does
not allow me to cover even briefly, all of the important programs. For
information on these, I recommend the FSR literature to the reader.
Most famous of the pioneers was Dr. Richard Bradfield at IRRI, who was
interested in devising rice-based systems to maximize year-round production
in the tropics. He popularized the study of multiple cropping and demon-
strated the enormous productive capability of the tropics with careful
management. Dr. Bradfield was succeeded by Dr. Dick Harwood and a group
of young co-workers who moved away from the intensive research station-
based productivity studies of Dr. Bradfield and began to examine more
closely the small farm itself. Their work led to innovative on-farm trials
and studies of the constraints and potentials of existing farming systems.
Later, as they began to become familiar with farm problems, they felt a need
to understand more of the climatic and land resources of their target farmers.
Dr. Hubert Zandstra succeeded Dr. Harwood about this time and has led a young
and vigorous team, including a very well organized and coordinated Asian
Cropping Systems Network, in designing and testing methodologies, training
and on-farm research.
Another pioneer was Dr. David Norman and a group of associates based
in northern Nigeria at Ahmadu Bello University. Their work centered on resource-
poor farmers growing millet and other dryland crops. This work was important
in dramatizing the benefits of FSR in Africa, in multidisciplinary research
team efforts, and in establishing a strong role for social scientists in FSR.
ICTA (The Instituto de Ciencia y Tecnologia Agricoles) in Guatemala, has
made a solid contribution to the problem of data collection and analysis on
the farm. Closely involved in the work was Dr. Peter Hildebrand who, along
with his ICTA co-workers, was responsible for devising and improving the
"Sondeo" or rapid analytical survey of the small farm and its problems. Many
other FSR programs are beginning to use or modify the Sondeo approach for
their own research.
A home-grown regional program that has generated much new thinking
on how FSR can be used to improve small farm systems is that of CATIE in
Central America, which is headquartered at Turrialba, Costa Rica. This
program has emphasized research and training approaches that are suitable
for the varied ecological situations and small.farm systems of Central
America. Most FSR programs vary to some degree, because they must adjust
to suit the local physical, social or political environment. For that
reason each program has or can contribute to FSR knowledge. Programs that
should or could have been listed and discussed in a history of FSR include
the ISRA program in Senegal, ICRISAT, IITA, and CIMMYT.
"Upstream" and "Downstream" FSR
FSR is generally seen as being important in helping to improve research
effectiveness and relevance. Often this is referred to as "upstream" or
"downstream" FSR. The most common definition of these terms is that upstream
FSR is seen as "FSR in the large", i.e., that generalized prototype solutions
are being sought that may have longer-term impact, while "downstream FSR"
is seen as being "FSR in the small" or research that is focused on more problems
that appear to have practical, immediate results and benefits.
There is another use for the terms upstream and downstream in FSR, re-
lating to research planning and the role of the farmer in that process. Here
downstream activities are seen to be a process of researcher decisions and
activities using professional training, skills, intuition and so on, but
without involvement of the farmer, in which research products are passed
"downstream" to an extension service that is charged with responsibility for
marketing the innovation at farm level. By contrast, upstream activities
are those where maximum intelligence, including the wishes, decisions and
concerns of the farmer, is used to focus research on real farm problems.
In this sense upstream FSR can be used to focus on problems facing farmers
now, or those which are most likely to face farmers in the future.
Characterizing different activities of FSR
Many people have had difficulty in understanding what FSR is and how
it differs or coincides with "conventional" agricultural research. In part,
this is because the terminology of FSR is not fixed nor agreed upon, and
new terms keep being coined. However, a greater problem, in my opinion, is
that there are different areas of activity in FSR that need explanation and
specification, if FSR is to be understood. The TAC Review of FSR at the
international agricultural research centers (IARCs) recognized this problem
and proposed a conceptual framework for understanding and conducting FSR.
The TAC Review suggested that there are three "Activity Areas" of FSRi Base
Data Analysis, On-Farm Studies and Research Station Studies. These activity
areas can be defined in part by where and for what purpose the research is
to be conducted. Also, the balance between the Activity Areas is usually
determined by the stage of development and needs of the individual FSR programs.
1. Base Data Analysis. This involves the collection,
collation and understanding of the many factors characterizing
the environment of a region. Much such analysis will entail
exercises in land resource mapping and evaluation, and in large
part can be done at research stations or in head offices, mostly
relying on secondary data. In addition to physical resource in-
formation, there is also a need for socioeconomic data on popula-
tion, farming systems used, production and income levels, and
various aspects of the infrastructure. The purpose of Base Data
- 5 -
Analysis is to learn as much as possible about the land and
water resources of a region, as well as socioeconomic factors,
and how these physical and social factors influence agricultural
production. The normal end product of the physical resource
analysis is a series of maps depicting agro-climatic zones,
land/soil units, and land use (farming/commodity systems).
Such information can then be used to assist identification of
potential target zones for on-site study and to determine the
Best locations for experimental stations or benchmark sites, as
well as providing a basis for later studies on research impact.
In general, Base Data Analysis will seldom involve detailed on-
site investigations except where larger non-farm units (for
example, villages) are the object of study.
On-Farm Studies can be used both to improve research planning and focus,
as well as assist in finding uses for improved technology at farm level. Thus,
On-Farm Studies may be used to gather information on systems as they are; to
conduct research on new innovations on the farm, either under researcher con-
trol, joint researcher/farmer control, or farmer control; evaluate adoption
of new technology; monitor changes in farming systems; and assess impact of
new technology. It should be pointed out that On-Farm Studies includes ex-
perimentation on the farm, in addition to more conventional surveys; such
experimentation does raise some methodological problems. On-Farm Studies offer
a great opportunity for cooperation with local extension services or institu-
Research Station Studies are seen to involve a focused research program
to generate new technology, design components for new systems, or modify
existing systems. Such research differs from conventional, on-going dis-
ciplinary research in that it is designed to fulfill a need in the context
of a given farming system. Sometimes it may be useful to distinguish dif-
ferent classes of Research Station Studies, for example: (1) exploratory,
developmental research aimed at solving specific problems. Once the problem
is defined in an FSR context, its solution may have a largely disciplinary-
oriented basis (i.e., a reductionist framework); (2) integrative studies,
where component parts are assembled and tested in a holistic framework, i.e.,
the synthesis of research results into applicable systems and management
Subdivision of FSR into three activity areas can be useful in helping
persons to understand just where particular FSR programs are focused or
December 29, 1980
AN OVERVIEW OF FARMING SYSTEMS RESEARCH METHODOLOGY I/
Richard R. Harwood
Rodale Press, Inc.
The person being newly exposed to the broad spectrum of farming
systems research (FSR) approaches used in third world agriculture will
immediately face a new vocabulary and a bewildering assortment of
approaches. Worse yet, in reading the literature (mostly in mimeo-
graph or annual report form) of the last five years, one finds rapid
changes or evolution in approaches if not philosophy. Because of
the newness of the approach, that change rate may even make what is
seen in a field visit seem at variance with the latest available reports
from those trials.
Modern day third-world FSR really began with the work of Dr.
Richard Bradfield in the late 1960's as related by Don Plucknett in
the preceding paper. During the early 1970's this approach was
"institutionalized" and adapted to the Asian network during the time
when I coordinated IRRI's program. Since 1975, there has been wide-
spread adoption and refinement of the basic principles of FSR through-
out much of the third world. It is fitting that this approach has now
"come home" in our attempt to address today's problems in the U.S.
The relationship of t6day's methods and approaches to the work of the
1930's and 1940's in the U.S. is in name and concept, but the methods
are not really very similar. I can say with certainty that the
methods used In FSR for third world systems, being very much adapted
to third world institutional, social and technological environments
cannot be adapted without major change to the U.S., but American
scientistsplanning to work in the third world or even those planning
FSR work in America can well profit from exposure to those methods.
Types of Farms where a Systems Approach has Benefit
Not all farming systems types afford equal payoff to holistic
studies. In single-enterprise agriculture such as in dryland, con-
tinuous wheat or in monoculture corn production, the single enterprise
has a particular fit to its environment. There are relatively few
variables which interact. Those interactions are continuous and can
be approximated nicely with relatively simple linear models. The
introduction of new technologies in production methods can be done
1/ Presented at the Symposium on Farming Systems Research,
Jefferson Auditorium, USDA South Building, Washington,
DC, December 8-9, 1980. Sponsored by the U. S. Department
of Agriculture, Office of International Cooperation and
empirically, as few options exist, and'changes are quantitative rather
than qualitative. Traditional, experiment station-based trials of
component technologies can meet many of the needs of such systems. But
where environmental 'gradients are sharp and where farming systems have
many component enterprises with strong interdependencies, the im-
provement of productivity requires not only better production com-
ponents, but an understanding of "what fits where." Impirical testing
here translates into a highly inefficient "hit and miss." Effective
technology development and targeting in such a complex environment where
integrated systems must optimize use of scarce production resources
requires either a method for progressive understanding of the systems
by a broad segment of development workers,or the availability of that
rare, widely-experienced and gifted scientist with keen insight and
intuitive understanding. We must therefore make use of a procedure
which will empower we lesser mortals to acquire and use, hopefully
early in our careers, similar insight in dealing with complex systems
and their environmental interactions.
The Spectrum of FSR Approaches
Modern day FSR for third world countries began under the sponsor-
ship of the larger, centralized research institutions. Their roles and
structure, being somewhat different from those of national programs,
encouraged an FSR approach peculiar to their own mission and capabilities.
In studying the various so-called FSR programs of today, one finds.varia-
tion due to many other factors as well:
1. The stage of development of the particular program. A
complex program does not all of a sudden appear in complete
2. The degree to which the program has been thought through
conceptually. Some programs may be merely copied from a
different setting and environment.
3. The resources that have been committed (or not committed).
4. The degree to which an ongoing program has been renamed or
slightly modified and the FSR terminologies applied because
of "fashion,' "trend" or availability of funds.
Variation across programs generally can be traced to three categories
of differences in objectives or "mission" of the FSR program.
1. The overall purpose of the,research:
Is the program a study-and-learn program or is the program
ultimately to result in change through an extension process for a
sizeable geographical impact area? In the first instance a cross-
section or sample of "target" farms will be chosen based primarily
on type of farm. There will not be undue concern over the extent
or size of the geographical area that they represent. If an even-
tual extension program is the ultimate goal, the program will be
structured first according to parameters which define the target
area and secondly, according to representative farm types within
that area. Early FSR programs focused on the first approach, but
by the mid-1970's the programs with a front-line development mission
and thorough conceptual base began to focus on definition of target
areas. The concepts of farming systems "determinants" and agro-
production complex came into being. Today's better programs first
identify target areas across which one or more identifiable farming
systems types are present, and then proceed to study and change
those characteristic types.
2. The change which is envisioned as resulting from a program.
Most programs are carried out on the premise that FSR studies
will encourage an understanding of the farming system, but that under-
standing will not, by itself, lead to improvement in the systems.
New and appropriate technology will have to be introduced,possibly
with some change in structure in the system, in order for impact
to be made. Where a single commodity is predominant such as in
many maize, wheat or rice-growing regions, the FSR may be truly
systems-oriented but will focus on change of the predominant enter-
prise in the systems of its target area. The change focus will
depend largely on the technologies available to the FSR institutions.
Those may be a single agronomic crop, a group of such crops, agro-
nomic crops associated with a major crop, horticultural crops, tree
crops or one or more animal components. Rarely will an organiza-
tion have sufficient expertise (or in fact mandate) to try to
change all of these components. The IRRI program, for instance,
worked on about one-half dozen agronomic crops which normally fit
around upland and-lowland rice. It occasionally worked with one
or more horticultural crops. No work was done to add animal tech-
nology. Indications were given when changes were needed in animal
types, numbers or feed availability. Some of these differences in
scope have been somewhat inappropriately labeled FSR "in the small"
vs. "in the large."
3. The extent to which the FSR is used as an overall institution-
building tool as opposed to a mere development tool.
On-farm FSR studies can be extremely useful in staff training,
in serving as a vehicle for cross-disciplinary studies, in bridging
the gaps between national institutions, in linking national and
regional programs, in fostering problem identification and feedback,
and in strengthening the research-extension linkages. With these
more broad goals, the methods and approach will be somewhat
A fourth difference in methods can be traced to the level of
support a program received. The more complex methodologies and
the detail of systems documentation possible for large, well-staffed
programs is clearly not possible or even necessary in national pro-
grams whose success is measured in terms of development rather than
in terms of "learning" or publications.
The Goals of a Regionally-focused National Program
I will concentrate my discussion on the national FSR program and
in particular that part of the national program having area-specific
responsibility, as I assume that most interest and participation of
American scientists will be in these programs. The goals for FSR in
such programs are broad 1. The primary end goal, linked closely to
the research itself, is agricultural development of a particular geo-
graphical region or district. Success or failure of the program is
evaluated by various development indices. While the program may be
called a research program, it has a responsibility for extension link-
ages, outreach testing and verification,and successful implementation
of the recommended technology. The primary goal includes an element
of successful extension. Research responsibility is thus expanded
to include the region-wide verification testing or so-called "applied
A second goal is to arrive at an understanding of the composition
and function of key types of farming systems in the target area. With
present status of the better national programs, this goal is reached by
the immediate study team but there is a noted lack of effective con-
ceptualization and articulation of that overview of the systems'
structure and function. Transfer of systems information and of the
often intuitive understanding of the production systems is at present
very limited. The effectiveness of the present programs depends upon
the pinpointing by the immediate study team of specific technological
needs which can then be-delivered by traditional methods. The ultimate
methodology would be the effective relaying of an understanding of the
systems to enable a broader participation of the national development
team in analysis and decision-making.
Other goals have been previously mentioned. The need for prac-
tical, on-farm training of scientists at all levels has been overlooked
for the past decade or longer. Many if not most of today's scientists,
either in developed and third world countries, have had little or no
professional experience in a farm setting with immersion in its complex
patterns of management and environment. It seems incredible that we
scientists profess to employ the scientific method without a first-
hand knowledge of what we are supposed to be researching. I know few
crop production specialists (other than those of the present FSR gen-
eration) who have ever participated in farmer-collaborative trials.
FSR provides excellent staff training opportunities.
Linkages between development institutions, departments and even
ministries,or between national and district or regional divisions is
facilitated in on-farm research. The farmer's field is the common
ground that unites most development agencies. It is "neutral turf,"
belonging to no agency.
Finally, the research-extension linkages can be maximized in on-
farm research. Extension specialists should participate in the on-farm
research. Verification trials are then often used for extension demon-
stration purposes, with the research and extension functions being
The Regional FSR Organization
Others have concentrated on this subject in the workshop, but
I would like to make clear the type of program I am referring to.
The methods that I will describe are used in FSR for a particular
regional target area. They may be for an area development program,
a district or provincial level program, or for any program having
specific area development responsibilities preferably within a single
political or administrative development unit. The FSR would be co-
ordinated by the agency with lead responsibility for providing
technology for the area. In other words, the FSR should not be con-
ducted by a separate team, with information relayed to a different
organization or ministry for the eventual providing of technology.
With information transfer currently being the weakest link, the need
for such transfer should be minimized. This regional, area-specific
orientation must be distinguished from the centralized, national pro-
gram which may have as its first goal the understanding of the major
systems and their technology needs in order to properly orient
centralized development of technology. Area-wide extension and
development is thus secondary in purpose.
Methods for Area Targeting
Target development areas are normally chosen from considerations
other than those based on agro-production complex definition. Effective
FSR depends on the ability to identify particular farming system types
which respond in similar fashion to their environment, which have rela-
tively similar internal interactions and which represent relatively
large production areas. Improved technology to fit those systems is
then verified across the target areas in systems of similar type. Such
types are limited in geographical area by combinations of environmental
determinants which may be socio-economic, physical or political. Com-
binations of determinates which give rise to similar types of farming
systems define the agro-production complexes across which FS informa-
tion can then be extrapolated. The two approaches to defining such
complexes are first, to simply survey the farming systems to identify
similar types. The geographical extent of those types then is assumed
to be the production complex. A second approach is to identify the
major determinants and to then define the production complex in terms
My own observations have found different methods to apply depending
on the nature of variation in the particular area. In upland, gently
rolling, subtropical areasthe physical parameters of rainfall and
elevation combined with distance from market are predominant. With
such patterns of variability on a macro-scale, simple identification
of farm type can serve as an indicator of production complex. In the
humid tropics with systems involving lowland rice, production complexes are
determined by similar macro-determinants, but also by "micro-determinants"
which change even within farms. Paddy size, shapeand elevation determine
the eventual cropping pattern. Technology is targeted on a district basis
by macro-determinant and then specified at the local level according to
type of paddy (which is easily classified by the research or extension
worker and farmer).
A third and somewhat more complex method is to define environmental
gradients and to determine how systems (or technology) change across
thosegradients. In actual field practice, however, the boundaries tend
to be rather discrete. Treating them as such, at least, simplifies the
While precise determination of agro-production complexes is im-
possible and to the newcomer seems highly subjective at best, their
effective determination is crucial to FSR and to the extension phase
which must follow. FSR without prior targeting and identification of
the production complex is purely impirical and is neither efficient
The Farmer Participant Approach
Much has been written about farmer collaborative research as it has
become popular in the past few years2. The key element is the farmer's
participation in the entire on-farm research process. The basic approach
was designed in the early 1970's as a combination of the Chinese "learn
from the masses" method and the western, top-down method that most of us
were trained in. The theory is that both farmer and scientist have
particular skills to contribute to the learning (research) process.
All too often, however, in actual practice the farmer is "used" in FSR
as a mere laborer rather than a full participant. The truly collaborative
relationship is an extremely sensitive one depending on human relation-
ships. It requires a very particular attitude, personality and training
on the part of both researcher and farmer. Effective farmer-researcher
interaction is both highly satisfying and profitable, but to those trained
exclusively in the top-down philosophy such relationships are difficult
to achieve. One may even hear the comment: ..."'we are wasting time
attempting to learn from farmers." Others may go overboard, losing
confidence in their ability to teach farmers anything. The proper
relationship and balance is essential.
Systems Description and Problem Identification
Once the target areas have been chosen, and the FSR team selected,
the initial research phase is that of description. The most effective
method to date for starting that process has been the rapid survey on
"Sondeo" method It has many advantages which will be covered in
later papers. Basically it consists of a rapid, reconnaissance-type
survey by an interdisciplinary team which will be eventually working in
the area. At the end of the 5-to 7-day survey, a final descriptive
report is written on the farming systems types of the area. The inter-
disciplinary makeup of the team, their commitment to the area and the
requirement for rapid conceptual analysis and summary are crucial to the
effectiveness of the process. All too often the surveys of different methods
are conducted by separate groups of specialists, with analysis being time-
consuming and eventually failing to provide a clear, conceptual understand-
ing of the systems the communications problem. The Hildebrand method
represents a tradeoff between objective measurement, quantification, and
intuitive understanding and perception. It is by far the most effective
approach. It may even be desirable to do the survey in two stages,
with the first stage being a reconnaissance survey across several villages
of the target area. This rapid once-over would give indication of the
variation in type of farm across the target area and may take as much
as two to three weeks. From that variation the specific types of farms
to be targeted by the FSR would be chosen. Farm size, types of enter-
prises and development level for FSR focus would be decided. A decision
would be made on villages for the FSR effort, based on predominance of
the desired farm type as well as cooperation from local village heads,
availability of housing, access and other criteria. The detailed, one-
week follow-up Sondeo would then be conducted. This study constitutes
the "benchmark" and focuses on systems description, productivity, problem
areas, and potential for change.
In a regionally-focused development program, the survey team should
be led by a member of the lead development agency who should probably be
a field-experienced social scientist. The team should include crop
production specialists, a plant protection person, a soils person, and
a member of the regional extension service. About six team members usual-
ly seems adequate. Many of these people should be destined for assignment
to the FSR teams that will work in the area. The other survey members can
be drawn from the central research organization and stations that will be
backstopping the FSR team. This experience is invaluable for staff train-
ing and for forming the personal relationships vital to communications
between locations and agencies. It is during these surveys also that the
tone is set for the "joint learning experience" between farmers and re-
searchers that characterizes FSR.
The detailed surveys should uncover serious problem areas. Many of
these are identified by farmers themselves, but the survey team, since it
includes production-trained scientists who are familiar with the best
available production technologies, has some knowledge of what might be
possible for the survey environment. Hopefully those standards for com-
parison would come from physical environments which at least are not
vastly different from that of the study area. If one of the team members
has international experience from similar areas in other countries, the
problem identification might be especially keen. At this point, however,
the process is highly subjective and dependent upon the skills and
training of the team, and on their ability to communicate effectively
The immediate product or output at the end of this 3- or 4-week
survey process should be a series of conceptual descriptions of farming
systems types and their prevalence for the agro-production complex. The
selected target types should be described in detail, with some quantifi-
cation of production levels, income, farm structure and other relevant
factors. Detailed lists of the more important production problems and
best estimates for improvement potential will be made.
Locating the Study Team
Once the surveys are completed, the study team is located in the
area. In most cultures it seems best if the team members are of the
same tribe or ethnic grouping, but whose homes are beyond easy travel
" 8 -
distance from the research site. It seems best if the study team can
be located within one of the study villages, removed from location in
a sponsoring institution. This maximizes contact with farmers. The
farmer linkage is usually more difficult than linkage back to the
research station or parent organization. A small, rented house in the
village is ideal,where team members can live and where a small office,
supply storage, and work area can be provided.
In many countries the two to five team members are supplemented by
"village assistants" who are usual lyteen-agers from the villages hired
to help in the project. They become team members and identify with the
research team. They are valuable in village level-research linkage and
with farmer-research linkages, since they are village members and are
well-known in the villages.
This is an often misunderstood phase. The "design phase" simply
is the overview and analysis of the systems, and the projection of possible
changes in the system based on available technology. Higher level re-
search scientists may well be used during this phase, but care should
be taken that they do not completely overshadow the study team. During
the design phase, the team never "redesigns" the entire system. They
simply program stepwise or single element changes in the system that
should lead to their improvement. Participant farmers should play a
major role in this process, and have a degree of (but not absolute) veto
power over it. Some.FSR people have been hesitant to introduce change
and may wait for a year or more before doing anything to alter the sys-
tem. The a a. veAy ,eAZou4 mistake. First of all, one learns much
faster by doing. Secondly, the farming system is much better understood
when you see how it can or cannot be changed. Most importantly, the FSR
team should neveA be allowed to remain in the village as merely
observers. They will develop a reputation for laziness or as observers,
not doers. The initial changes should be small, but they should be tried.
A variety trial is an ideal icebreaker and is noncommittal. Simple
changes in row spacing, weed control or a few rows of a well-tested new
variety across a field are all "safe" options for initial design changes.
An experienced team with good backstopping and technology from similar
environments may attempt changes in the crop rotation on small areas in
the first year.
The design should be upgraded year by year as test results come in.
As one or more adventuresome, high-management farmers are identified,
more adventuresome technologies can be tried. It is very important for
the FSR team to make an early contribution, to have "a winner." That
can be small, but it should come early. This lays the groundwork for
more adverturesome types of trials in subsequent years.
4 5 6
Much as been written about on-farm study methods '5 I will only
briefly summarize them. As in all FSR, field testing goals must be clear-
ly defined. The approach to be used as well as the field design depend
upon those objectives.
There are three basic field methodologies: farmer-managed trial.
the "superimposed" trial, and the research-managed trial. In farmer-
managed trials, the farmer conducts all of the operations in crop or
animal production. The treatment units are usually large enough to
make them economic units. Researcher participation in such a trial
is limited to design, monitoring, perhaps some management advice and
at the end, recording of yield. This trial is used where the farmer's
management skill is either needed for successful completion of the trial
or is actually under test as a part of the new technology or where
economic data are needed on labor use.
The superimposed trial is one where the farmer manages the field
but the researcher imposes a particular treatment on subplots. Fertil-
izer, weed control, insect control or other variables may be superimposed,
with the farmer managing all other factors across the plots and the
entire field. This is done with minimal disruption of the farmer's system.
It measures the effect of incremental changes in technology as applied to
The final type, the research-managed trial is used where a completely
different management is needed, where small plots are required, where many
variables are involved or where high-risk or low-yield treatments are used.
Most statisticians favor this approach as it gives maximum statistical
precision. It has the disadvantages of requiring large amounts of staff
time and doesn't take advantage of farmer-management expertise.
In all arguments over method, the FSR person should always clearly
state his research goals for a particular trial. The goal determines
the method to be used, with the requirements for statistical precision
being secondary. The needed precision can usually be achieved with
any of the methods if proper design is used.
It is important that the field study team do the initial data
summaries, using simple hand calculators if necessary. It is crucial
that they have a feeling for research results soon after the data are
collected. At the beginning of the trials the summary methods should
be determined, with forms printed and available for all data summaries.
The staff must be trained and disciplined to accomplish this quickly.
If needed, a person from the headquarters station can come down to the
field to help with these summaries, but they should be done in the. ield.
Further, higher level analysis can then be accomplished by other staff
at a central location after initial summary reports are done by the staff.
It is important also that all staff in the field be familiar with results
from all trials. With division of responsibility among team members and
heavy work loads, there is a tendency for breakdown in communications.
Following two or more years of successful research testing of new
technologies, (with participation of extension staff in the research)
the trials are then carried to similar farms in representative areas of
- 10 -
the agro-production complex. The role of extension becomes progres-
sively more dominant as this phase is entered, as these trials usually
will serve as demonstration trials prior to an all-out extension effort.
A regional extension person will usually have been assigned as a member
of the FSR team. He will have trained other extension workers in the
technology and its testing in the research villages. The extended veri-
fication testing may be done either under extension or research direction.
There will almost always be research participation by member of the FSR
team, whpse role has begun to change. In the verification stage, the
options are limited in number and the trials simple but in many locations.
Linkages Across Agencies
With village-level research,linkages are almost always easier.
Different research divisions representing different disciplines, or
even different ministries can be broughttogether on common ground.
While problems can still exist, my own experience has indicated that
communication is almost at&wct easier in the village or in the farmer's
field than it is at the institution level.
I find also that the discussions of problems of interdisciplary research
are more relevant to the U.S. than they are to third world situations.
Those difficulties are in direct proportion to "academic" level, being
greatest in an American university setting, intermediate in third world
academia where promotion is seldom based strongly on quality of published
documents, and minimal in the line development agencies with field re-
sponsibility. At the field level promotions and pay scale in a third
world situation almost never are influenced by publication. They are,
however, often influenced by distance from headquarters, so just being
assigned to a remote location may be detrimental. I find much of the
American-based discussion about interdiciplinary problems and the
problems of publication for promotion nonrelevant to regional field pro-
grams in the third world. The scientists assigned to village-level
programs are not highly specialized. They do "some of everything" as
is needed in the village which may include helping to fix the roof of
their participant farmer or to buy and administer vaccine for his sick
Linkage and communication problems are minimized within the village-
level team. Communications back up the institutional or bureaucratic
chain are more difficult, requiring good program leadership.
Tne ensuing region-wide production programs from such an approach
will have had broadly-based research input. Most of the concerned
agencies should have been involved during the various phases of the FSR
and verification testing program. All should have a conceptual under-
standing of the farming systems to be targeted.
The role of the FSR team has now changed to that of trouble-shooting
on the technology. They should be free to travel and to monior the new
technology as it is applied. The greatest danger in a production effort
is the attempt to apply new technologies beyond their area of adaptation.
Production programs must be first targeted to the specific production
complex defined by macro-determinants. The adjustment to micro-deter-
minants is dependent upon the clear identification of thosedeterminants
in the field by both farmers and extension workers. For example, in
the terrace agriculture of Nepal's hill rice areas, the extension of
the new wheat varieties for wheat-rice rotations is highly dependent on
proper identification of field or "paddy" types. The identification is
simple, but requires training of extension workers and farmers alike,
and labeling of the package of practices according to needed paddy con-
ditions. The FSR team should trouble-shoot this application process
during the production effort.
The Need for Strong Leadership
FSR programs have greater leadership requirements than do programs for
testing of simple component pieces. The demands for conceptual under-
standing and clear establishment of goals and priorities require a much
stronger and intensive management function of the project. Without this
leadership the field-level,middle-skill scientists will be lost in com-
plexity. I feel that this management area is a perfect application for
the talents of expatriate scientists iL they aAe highly t'wained and ex-
peA.enced in FSR methods.
Following goal-setting and conceptualization in importance to
the FSR team is the necessity of judging the scope and amount of work
to be attempted by the FSR team in the village. The tendency is aeoWy6
to attempt more than they can do well. Village-level work, if it is
effective, has a dramatic effect on team members. Morale, dedication
and empathy for their farmers become extremely high in a good team. It
requires a high level of management skill by leadership from above the
village team level to limit and shape the activities of the team to
coincide with their skills and training.
FSR methodologies are newly emerging, conceptually complex and
quite different from mostapproaches used in the past. The conceptual
differences are crucial to their success. The scientist being exposed
to them for the first time should seriously attempt to sense and under-
stand those concepts and differences. The application of FSR in its
broadest sense as related here, has received wide acceptance in third
world agriculture. That acceptance by third world development leaders,
concerned with the broad aspects of both institution-building and area-
development programs, has preceded widespread acceptance by the more
tradition-minded scientific community.
I do not see FSR as "the" answer to agricultural development. I
do see it as a very important tool, among others, which can play a
central role in development efforts.
1. Harwood, Richard R. 1980. Farming Systems Research and its
role in the national agricultural research system.
Paper presented at the workshop: "Increasing the productivity
and impact of agricultural research." Sponsored by the
International Agriculture Development Service and the Indonesian
Agency for Agricultural Research and Development, Yogyakarta,
November 9-14, 1980.
2. Harwood, Richard R. 1979. Small Farm Development: Understanding
and Improving Farming Systems in the Humid Tropics. Westview
Press, Boulder, Colorado.
3. Hildebrand, Peter. 1979. Summary of the Sondeo Methodology
used by ICTA. Institute de Ciencia y Tecnologia Agricolas,
4. Gomez, A.A. and K. A. Gomez (in press). Multiple Cropping in
the Humid Tropics. IDRC, Ottowa, Canada.
OTHER GENERAL REFERENCES ON FSR METHODOLOGIES
The following are excellent comprehensive summaries:
5. Consortium for International Development (in press). Guidelines
to assist nationaF government in the implementation of Farming
Systems Research and Development Programs aimed at farmers with
limited resources. Boulder, Colorado
6. IRRI (in press). Guide to on-farm cropping systems research.
Los Banos, Philippines.
7. Norman, David W., E. H. Gilbert and Fred Winch, 1980. "Farming
Systems Research in the Third World: a Critical Appraisal."
Michigan State University Rural Development Paper.
8. Rohrbach, David D. 1981. A discussion of issues relevant to the
development and implementation of a farming systems research
approach. Concept paper for this workshop.
OVERVIEW OF THE POTENTIAL APPLICABILITY OF FARMING SYSTEMS
RESEARCH TO U.S. SMALL FARMS
AND U.S. RESEARCH AND EXTENSION
Department of Anthropology
University of Florida
Paper presented at the Symposium on Farming Systems
Research (December 8-9, 1980) in Washington, D. C.,
co-sponsored by the Office of International Coopera-
tion and Development and the Science and Education
Administration of the U.S. Department of Agriculture
and the Development Support Bureau, Office of Agri-
culture of the U.S. Agency for International Develop-
We are examining farming systems research (FSR) from two perspectives:
(1) as experts in one or more aspects of agricultural change, and (2) as em-
ployees and officials of different bureaucratic organizations. The examina-
tion is both theoretical and practical. Theoretically, what is the logic
behind FSR, and what should it do? In practice, how does it really work?
Those of us with actual field experience with FSR must describe how it helps
us, and what are its costs as well as its benefits? In the end, as experts
and bureaucrats, do we want to use an FSR philosophy and methodology?
In examining the applicability of FSR to the U.S. and to research
and extension programs oriented toward small farms, we must start by being
very clear about our evaluation of existing and past small farms programs.
Are they working as we want or not? Where do they work, and with whom?
What kinds of farmers and farm problems are handled the best and the worst?
Do we think that FSR could theoretically do a better job? Would we ourselves
work better, and would small farmers be better off, if we worked through the
FSR process? In performing this evaluation, we must remain constantly aware
that we (the expert evaluators) owe our jobs, our professional careers, and
a great deal of loyalty to our governmental agency or land grant university.
These are the selfsame agencies and universitiesthat have masterminded or
ignored American small farmer programs in the past. This paper is not de-
signed to judge these bureaucracies and programs nor to sell FSR. Each of
us, and the group that is assembled, must wrestle with our own evaluations
FSR Surveys in Alachua County,Florida:
Most of my own experience since the mid 1960s is international:
Bolivia, the Dominican Republic, Zambia, and Malawi. Only since 1978 have
I begun to work in Florida with small farmers. This last spring a multi-
disciplinary group of Florida faculty, led by Pete Hildebrand, taught a
course in FSR methodology to a multidisciplinary group of graduate students.
We took them out in our county to conduct a sondeo or rapid reconnaissance,
then we analyzed together the small farmer situations they encountered,
and finally planned possible interventions. At the end of the class, its
findings and recommendations were presented to an audience that included
county small farmers and an extension agent.
This past summer Elon Gilbert, an agricultural economist, and I ex-
tended our knowledge of the county by supervising a set of FSR surveys
(Table I). The core survey was a random sample of operators of agricultural-
ly assessed land. For the single county we were trying to work out what
kinds of systems exist and their frequency, while at the same time testing
FSR survey methodology.
In addition, we supervised three surveys of production and marketing
practices for specific commodities: beef cattle, watermelons, and squash.
These FSR studies "in the small" were designed to illuminate differences
between low and high resource farmers and to point out potential areas of
under-utilized or mismanaged resources for future research and extension.
Reports of these four surveys will be published in January 1981 (Hansen,
Griffith, et. al. 1981; Gilbert, et. al. 1981).
ALACHUA COUNTY FSR SURVEYS
-describe existing systems/practices
-general verification of county farming
-specific study of production and marketing
for different commodities
-faculty and graduate students from:
agricultural economics, agronomy, animal
science, anthropology, education, entomology,
history, and vegetable crops
-county extension agent
within a commodity
farm and off-farm
-research flows into extension
Faculty and graduate students from eight departments, as well as a
local extension agent, were involved in planning, carrying out, and analyz-
ing the four surveys. This is really the major reason why I personally be-
came involved in a domestic program. These domestic, in-state operations
are the best and cheapest way to train people in FSR--our graduate students,
fellow faculty and agency colleagues, and visitors from other nations may
receive hands-on, do-it-yourself experience and skills right in our own
backyards. At the same time, we may more easily and quickly test and evalu-
ate FSR methodologies in these in-county programs, as well as call upon the
accumulated expertise of all our agencies and universities.
We who participated in the spring and summer projects are now con-
vinced that the multidisciplinary FSR method is very rewarding. Without
tackling the cost-benefit analysis of these rewards, the interplay of
agronomists, animal scientists, anthropologists, entomologists, economists,
etc. kept all of us constantly aware of the complexity of the conditions
that farmers face and the complexity of the farming systems that people
create to cope with their environmental conditions. Other advantages of
FSR were the constant testing of our ideas with county realities--including
extension agent and farmer input throughout our research kept us from be-
coming too theoretical--and the flowing together of research and extension
as extension agents helped plan and question research, and as researchers
got out onto farmers' fields.
County production is quite diverse, and there is a lot of hetero-
geneity in farming systems. We made sense of this by first dividing farms
into three major production strategies: (1) livestock-centered farms in
which crops (other than pasture and feed) are absent or minimal elements;
(2) crop-centered farms in which animals are absent or minimal; and (3)
mixed or balanced farms in which animals and crops are about equally impor-
tant. Within these major strategies we differentiated systems and recom-
mendation domains by separating a few key enterprises and splitting low
and higher resource farmers. At the end we had the nine systems shown in
COUNTY FARMING SYSTEMS AND RECOMMENDATION DOMAINS
1. Beef-centered low resource farmers
2. Beef-centered higher resource farmers
3. Non-beef livestock-centered low resource
4. Horticultural crop-centered low resource
5. Specialty crop-centered low resource farmers
6. Agronomic crop-centered higher resource
7. Agronomic and horticultural crop-centered
higher resource farmers
Mixed or Balanced Farms
8. With tobacco higher resource farmers
9. Without tobacco low and higher resource
In our surveys we found that farmers (like many other people else-
where) did not like to tell their income. This research problem is confirmed
for the U.S. by Lola Smith, a colleague with extensive research experi-
ence with family farmers in this country. Since income data was not readily
available, we used acreage and herd size as the criteria for defining small
or low resource farmers. We categorize people who operate less than 101
acres or, if livestock-centered, fewer than 50 head of livestock as
Low Resource Farmers (LRF).
Fifty per cent (54%) of our random sample of county farmers are low
resource, a percentage that agrees fairly closely with U.S. agricultural
census statistics. It must be clearly pointed out that these people who
practice a low capital agriculture are not necessarily poor people. In fact,
the relative absence of obvious poverty among the surveyed low resource
farmers is a surprising finding of our survey.
In many other ways, however, the Alachua County low resource farmers
do fit the common assumptions made about American small farmers (Table III).
Their land resources, although small, are under-utilized: only 55% of the
land they control is in pasture and crops, while much of the rest is still
in woodland. Our three commodity-specific surveys also show that these far-
mers are producing much lower yields, in general, then higher resource far-
Low resource farmers in our survey are part-time farmers: off-farm
income and commitments are very important, perhaps more important for most
of them than on-farm income. Those who are not working off the farms are
often old and retired. Time and energy, therefore, should be recognized as
limiting factors. Most of them are involved in farming systems that require
By far the most popular strategy for low resource farmers is being
livestock-centered, probably because of its low management needs. This
means that county research and extension programs that want to work with the
LOW RESOURCE FARMERS IN COUNTY
(<101 acres, or <50 cattle)
54% of all farmers in survey
55% in pasture and crops
Time and Energy
61% are part-time farmers
39% are 60 years of age or
mixed or balanced
interested in growth of farm
receive both income and
17% only income
24% only subsistence
majority of the present low resource farmers will need to concern themselves
with livestock and forage rather than agronomic or horticultural crops.
Less than half of the low resource farmers in our survey are growth-
oriented, i.e., oriented toward increasing their production and profits.
Almost as many are more interested in stability, neither risking, gaining,
nor committing more to their farms. About one in every ten are reducing
their involvement or getting out of farming completely (in decline). More
than half receive both subsistence (largely from gardens and livestock) and
cash income from their farms. Another one fourth only get subsistence,
while only 17% receive only cash income.
Putting these together, it means that many of the sampled low re-
source farmers are not really profit-oriented in terms of their farms. In
general, they want to live on a farm. Subsistence is important because it
cuts their living costs, underwriting the desirable rural life. They may
have inherited or purchased the land but that is not their real income-
earner. Many put something in the ground or run a few animals because they
like to do that and because they need to amortise their farm life by getting
an agricultural assessment to lower their taxes. Only some are really
interested in working harder on their farms for immediate income growth,
and others want long-term systems to set up their farms as self-sufficient,
often looking ahead to retirement.
FSR at the University of Florida
The surveys we have described are part of the first phase of estab-
lishing through the University of Florida an in-state FSR program to work
with Florida's low resource farmers. Before elaborating on this specific
program, it is important to note the broader involvement of people within
the university community (faculty, students, and administrators).
FSR AT THE UNIVERSITY OF FLORIDA
-faculty and graduate student core
-sack lunch and seminars
-courses (especially methodology courses)
-minor in FSR
-some outside funding
-FSR/E small farmer in Florida program
Core people in FSR activities at the university are faculty and
graduate students who are dissatisfied with current disciplinary--specific
and commodity--specific approaches. There is an informal sack lunch for
the Social, Agricultural, and Food Scientists Study Group every week that
provides a way to communicate with each other about our interests and up-
coming research, courses, conferences, etc. Occasional seminars are
sponsored by the FSR group, and several courses are taught, such as the
field methods, hands-on course mentioned earlier. Multidisciplinary teach-
ing and research projects, such as the low resource farmer surveys mentioned
above, are set up and staffed by the faculty and graduate students from this
core, occasionally supplemented by otherswhen other disciplines are needed.
Recently, to formally recognize the existence of relevant courses
and strong student interest, we have established a minor in FSR. Administrative
support and some outside funding have been secondary in the past to the
faculty and graduate student interest, although that support and funding
are now increasing.
Right now, spearheaded by Pete Hildebrand, we are setting up an
FSR/E (research and extension) program to work with Florida's low re-
source farmers. Although focused on domestic farmers, the staff and pro-
grams will be available for use in training, planning, and evaluation
of international FSR projects as well.
The program structure is outlined in Table V. A coordinator
will report to the Office of International Programs, Institute of Food
and Agricultural Sciences (IFAS), and the Deans for agricultural re-
search, extension, and instruction. An advisory council of department
chairs and a technical committee of relevant faculty will work with this
coordinator on general policy and specific projects, respectively. In
the Florida counties targeted for work, an advisory committee of county
extension agents will be involved in planning and implementation, and
in those counties an FSR/E field team of scientists will identify and
work on tests and trials for specific systems and commodities in collabora-
tion with local agents and farmers. For more information on this, please
contact Pete Hildebrand, the acting coordinator.
UNIVERSITY OF FLORIDA DOMESTIC FSR/E PROGRAM
International Programs and R/E/I Deans
Advisory Council --- FSR/E Coordinator
of Dept. Chairs
Technical Committee -
Advisory Committee -------------
of County Agents
FSR/E Field Team
International and Domestic Comparisons:
In working with low resource farmers in Florida, it became rapidly
apparent that there are major similarities with Third World situations. But
rather than dwell on the similarities, let me point out some equally impor-
tant differences (Table VI).
INTERNATIONAL DOMESTIC CONTRASTS
Domestic Small Farmers
-heterogeneous (in survey)
-not as committed nor as dependent
-not as poor nor as isolated
-more mechanical skills
-not as significant to country
Research and Extension
-extensive fund of knowledge
-insiders not outsiders
-we have met the enemy ...
U.S. low resource farmers are very diverse, as our sample shows.
There is no widespread homogeneity to improve the cost-benefit performance
of FSR work. They are part-time, part-committed, and part-dependent on
agriculture. We cannot automatically anticipate that these farmers will
commit more time to farming even if we show them improved technologies.
Also, at least in our county, low resource farmers are not primarily the
poor and isolated people that are the stereotype. In Alachua County, small
farmers include airline pilots, shop owners, university staff, etc. Less
than ten per cent are obviously poor. What this means is that low resource
farming does not necessarily mean that the farmers are completely low re-
source people, only that they have committed few of their resources to agri-
Another significant difference is the importance of low resource far-
mers to the country as a whole. The U.S. farming population is only 3.5 to
4% of the national population. If we assume that 60% of them are low re-
source, that is only 2.5% of the country. In contrast, in other countries
this population may be so large that small farm research and extension are
needed to vitalize national agricultural production or prevent massive dis-
location, unemployment, and welfare problems.
There are other critically important differences between the research
and extension environments in which farmers operate. We have in this country
a lot of agricultural, socioeconomic, and political data to buttress our
domestic programs. Institutions for communication, marketing, research and
extension are well developed here and we know these institutions as insiders.
We are the native informants, the host country nationals of the agencies
and universities that are responsible. Here again we must remember that
we are bureaucrats--the officials and employees of the research and exten-
sion institutions that are responsible for much of the available data and
many of the programs. If they are insufficient or misguided, we cannot
shift the blame to "the natives", for we are they.
1. The surveys were funded by grants from the Center for Community and
Rural Development and from Sponsored Research, University of Florida.
Gilbert, E., et. al. 1981. Production and Marketing Practices for Water-
melon and Squash in Alachua County: A Comparison of the Farming
Systems of Small and Larger Scale Producers. Center for Community
and Rural Development, University of Florida. Gainesville, Florida.
Hansen, A., D. Griffith, et. al. 1981. Farming Systems of Alachua County
Florida: An Overview with Special Attention to Low Resource Farmers.
Center for Community and Rural Development, University of Florida,
Farming Systems Research Symposium
Dec 8-9, 1980
Barton Baker, W. Va. Univ.
THE ALLEGHENY HIGHLANDS PROJECT
The Allegheny Highlands Project was an experiment designed to test the
effectiveness of a method for delivering a package of technological infor-
mation to farmers as a means of promoting rural development. The Project
began in 1970 and terminated in December 1979. The Project was supported
by West Virginia University and by two grants from the Rockefeller Foundation.
University personnel assigned to the Project consisted of an agronomist,
animal scientist, farm economist and veterinarian who worked together as a
team to provide individual recommendations to cooperating farmers located in
a 9-county area of central West Virginia. In addition to the 4 staff members
listed above, there were also assistants in each of the discipline areas, a
resident panel consisting of 5 faculty members located on the campus of West
Virginia University who provided support and liason between field and campus
personnel, and part-time help to assist with record keeping and routine
activities of the Project.
Field Staff worked directly with some farmers and indirectly with others in
order to compare two methods of technology transfer. Approximately 80 farmers
were involved as official cooperators in the project. Many others received in-
formation but no data were maintained on their operations.
The major objective of the Project was to determine what would happen to farmers
and communities over 10 years if existing technology was presented in a new
approach. This objective originated from the question: "Is present technology
not adopted by farmers because it is not meaningful to their situations or has
it never been presented to them in a meaningful fashion?"
Due to time limitations, I will not dwell in great detail with our methodology.
The following are some of the things done in our Project that are similar to
those presented in the Farms Systems Research concept:
(1) Scientists were located in the community with people they were
(2) Extension (technology transfer) and research (demonstrational and
applied) were combined into a single functional unit.
(3) Individual farms were viewed as a total system and not a group of
(4) Farmers had a role in determining research needs and participated
in experiments conducted on their farms.
(5) A team or package approach was used.
Some things learned relative to organization and procedures that I think are
applicable and needed in the Farming Systems Research concept are:
(1) Goals and procedures
(2) Competent staff
(3) Appropriate disciplines
(4) Community involvement
(5) Mutual confidence
(6) Adequate funds
(7) Administrative support
(8) Package approach
Some further conclusions based on our experiences are:
(1) Existing technology needs modification
(2) One-to-one contacts are most effective
(3) Changes occur slowly
(4) Continuity of staff is important
(5) Staff communications are essential
(6) Support staff can play a major role
(7) Office location is important for farmer initiated contacts
We have found the approach used in AHP to be very successful. Farmers
benefitted from increased production and increased income per farm. Their
attitudes became more positive and cooperation increased. The University
benefitted from increased contacts with clientele which resulted in more
meaningful research and teaching and better public relations within the
scientists (Animal Scientist, Agronomist, Horticulturist and Entomologist)
presently assigned or now on board at the Center. Logistics of developing
the 1500 acres of demonstration/validation experimental lands are in dif-
ferent stages of both planning and actual development. Hopefully, the 1981
growing season will find some "gap-filling" field experimentation underway
for all initial thrusts of the Center.
Currently, while logistically putting together the Center, detailed research
and extension programs with both short and long range goals of the Center
are being worked on by the Core Staff in consultation with many small farmer
research users, other researchers and extension specialists of the
University/SEA/and private foundations, program planning administrators,
other federal, state and county agencies and in some cases businessmen that
understand to some extent the plight of small farmers, are sympathetic to
their needs and can favorably impact success of the small farmers operations.
This approach, through the use of AD HOC committees as well as other meetings,
scheduled with the help and inputs of the state cooperative extension service
and county agents is bringing forth what might be called "grass roots" or
"users" inputs for the consideration and benefit of we in research, extension
and education. Additionally, complimentary efforts are now underway to use
intensive subject matter statistical surveys to characterize small farms of
the "hill" region and determine their "real world" researchable needs and
problems. The Center's "gap filling" research will address high priority
needs or problems surfacing through all of these inputs. Thus, the Center
will initially involve some of the same small farm people that are served
by the Center's research and extension in planning the program of the Center.
Hopefully, as we progress in our Center program either in-house or extra-
mural procedures will be devised to include the expressed felt needs of
a greater number of small farmers throughout the 10 state area. Hopefully,
we will be able to establish mechanisms that will allow these people to
present and receive response to their problems from the Center in a timely
manner. Identification of genuine technological needs and problems of the
small farmer is a time consuming process. Technological needs of small
farmers are in many cases location specific. Problems or needs of small
farmers are continuously shifting. These problem shifts can result from
biological or physical changes (such as diseases, pests, long term weather
conditions, etc). Among other things they can result from sociological
considerations including such things as consumer/market demand often resulting
from "easy living/or hard" times. Whatever the cause, they will require
constant reassessment and incorporation of new ideas into experimental design
to meet and overcome enough of these problems to make a real and favorable
impact toward helping the small farmer be more productive and become more
on-farm self sufficient.
Only the research component of the Center is being realized initially and
then only partially. Four, of what hopefully in time will be an inter-
disciplinary small farm research team of 9 scientists, are now assigned
to the Center. Three of these scientists are SEA-AR employees and the
fourth is on the Staff of the University of Arkansas supported through
a SEA Cooperative Agreement. Pressures of existing organizational
scientist evaluation systems, pressures to produce quick tangible results
in order to justify the program of the Center by any fashion or form will
require thorough considerations and determinations -- "in-truth" -- whether
the Centers goals are principally to generate, validate, and demonstrate
to the point of adoption technology that will help the small farmer become
more productive and better off or whether our goals are principally directed
toward satisfying requirements in order to document something on paper
therefore survive to investigate and serve the actual research users another
day. Hopefully, small farms researchers, extension specialists and educators
will find the kind of support from Agricultural Research and Extension
administrators that will develop alternate ways to help researchers and
specialists withstand these pressures that are often only vaguely if at all
productively oriented toward the end-product of materially helping the small
It has been projected that state cooperative extension specialists and
perhaps federal extension will become part of the Center team within a
couple of years after a Center research program is established. These
specialists, an integral part of the Center team, will collect and disseminate
information on small farms throughout the region to be served. They will serve
as team liaison people to state and county staffs in the region, packaging
educational material by modules for initiating or expanding work with small
farmer clientele in a more timely manner. They will also be complimentary to
the research staff of the Center toward greater interfacing of researcher's
with user's needs and problems when planning the Center's research program
and providing timely responses for these small farmer needs. With many
questions yet unresolved where the Center may be headed fund-wise, personnel-
wise, facilities-wise, and even program-wise, I would like to share these
plans that are subject to change with you. Much of what is planned for the
Center falls within a Farming System's research approach. (a) For instance,
although team scientists have been disciplinarily trained and are generally
commodity oriented they will be concentrating their efforts toward an
interdisciplinary approach of small farm problems while taking into account
and orienting the Center program toward the total small farming operation.
(b) There will be a blend of both that referred to as "upstream" and "downstream"
research. Research conducted on the Center will be directed toward solving some
of the major problems of small farmers that are shared commonly throughout a
given region by all or at least a homogenous group but only when answers to
these problems are not available elsewhere. By the same token "location
specific" problems may be addressed both on the Center or on farmer cooperators.
Regardless, the planning of all research will have the inputs of small farmers
and a considerable degree of the implementation and evaluation using farmers
fields and resources will in time occur.
As the Center program is initiated conventional but "gap filling" research on
the four thrust areas previously spoken to begins. This research will soon
be advanced to sub-enterprise module experimentation on site. These modules
can be evaluated individually or collectively for alternative enterprise com-
positing for total farming systems analysis. They will also serve the Center
for demonstration and validation purposes. In time enterprise interaction
studies will be conducted by physically developing small farm demonstration
models on the Center. These small demonstration farms perhaps developed with
the benefits of simulation modeling and total systems analysis will serve
APPLICABILITY OF THE FARMING SYSTEMS RESEARCH APPROACH TO
LESS DEVELOPED AND DEVELOPED COUNTRIES LINKAGES AND CONSTRAINTS
MICHAEL S. JOSHUA
BUREAU OF ECONOMIC RESEARCH AND DEVELOPMENT
VIRGINIA STATE UNIVERSITY
SYMPOSIUM ON FARMING SYSTEMS RESEARCH
WASHINGTON, D.C., DECEMBER 8-9, 1980
A. VIRGINIA STATE UNIVERSITY AND RURAL DEVELOPMENT
(1) Domestic (USA) Program
(2) International (Ghana) Program
B. FARMING SYSTEMS RESEARCH AND DOMESTIC-INTERNATIONAL
C. CONSTRAINTS ON THE FARMING SYSTEMS RESEARCH APPROACH
(1) Management of the Knowledge System
(2) Applicability of Farming Systems Research Approach
A. VIRGINIA STATE UNIVERSITY AND RURAL DEVELOPMENT
Virginia State University (VSU) through its Bureau of Economic
Research and Development (BERD) conducts rural development research in
seven areas: Manpower, Management, Production Analysis, Marketing,
Agricultural Finance/Rural Financial Markets, Farming Systems, and Urban
and Ecological Studies. Its approach is interdisciplinary utilizing the
disciplines of Economics, Business Administration, Sociology, Statistics,
Psychology, International Studies, Education and Agricultural Economics.
Aware of the unique perspective which the 1890 land grant institutions
can bring to the study of rural development problems, BERD has articulated
a program of domestic and international research aimed at solving rural
development problems. This program gives evidence to those involved in
rural development decision-making of the expertise available at VSU.
(1) Domestic focus
Research is carried out in a 16 county (and 5 city) area of South
Central Virginia which BERD has defined as its study area for domestic
rural development. These 16 counties include: Amelia, Brunswick, Charlotte,
Cumberland, Dinwiddie, Greensville, Halifax, Isle of Wight, Lunenburg,
Mecklenburg, Nottoway, Pittsylvania, Prince Edward, Southampton, Surry and
Sussex. The five cities of Danville, Emporia, Franklin, South Boston and,
Suffolk are included.
The research activities (completed and ongoing) in South Central
Virginia have beep organized into parts of a rural sector study focusing
on the socioeconomic aspects of the low income small farm and non-farm
units. An illustrative cross section of the research projects include:
(1) MARKETING STUDIES
(a) Market Organization and Activities by County in South
(b) The Small Farmer and Agricultural Marketing in South
(c)- Market Structure of Selected Agricultural Products and
Services Industries in South Central Virginia
(2) AGRICULTURAL CREDIT AND RURAL FINANCIAL INSTITUTIONS
(a) An Analysis of Commercial Bank Credit to Farmers in
Pittsylvania County, Virginia
(b) Demand for Agricultural Credit
(c) Credit to Small Farmers in Selected Counties of
South Central Virginia
(3) PRODUCTION STUDIES
(a) Economic Analysis of Small Farm Production in Swine
Operations of South Central Virginia
(b) The Effects of Socioeconomic Characteristics Adop-
tion of Recommended Beef Farm Practices
(4) RURAL DEVELOPMENT STUDIES
(a) Agricultural Resources and the Development of the
Small Farmer in South Central Virginia
(b) Sociological and Economic Factors Affecting the Labor
Supply and Earnings of Male and Female Heads of House-
holds in South Central Virginia
(c) A Comparative Socioeconomic Profile of South Central
(d) Socioeconomic Impact of Intrastate Motor Carrier Regu-
lations on Rural Development
(e) Occupational Mobility of Rural Labor
(f) Labor Turnover in Small Enterprises in South Central
(5) WOMEN AND RURAL DEVELOPMENT
(a) Social and Economic Characteristics of Women in Virginia-
the Labor Status of Rural Women
(b) The Value of Farm-Wives' Time in Agricultural Production
in South Central Virginia
(6) HUMAN RESOURCES AND RURAL DEVELOPMENT
(a) An Assessment of Human Resources in South Central Virginia
(b) Human Resources in South Central Virginia Social and
Research projects conducted in the South Central Virginia laboratory
area develop and utilize research methodologies that are transferable to
LDCs. One critical methodological area relates to the unit and framework
of analysis. VSU's experience in studying constraints on rural develop-
ment points to the need to focus on the farm-firm-household (FFH) as the
unit of analysis. The interaction between consumption and production
activities play an important role in resource allocation decision making
in the rural sector. This means that the appropriate model is the FFH.
Further, since the the performance of the FFH is affected by forces ex-
ternal to it (marketing facilities, input delivery systems, extension,
agricultural knowledge systems, and value system) the most appropriate
framework for analysis is the systems approach. This approach permits
an examinatation of the internal and external factors affecting consump-
tion/production activities and so permits better targeting of constraints
and more accurate and effective intervention strategies.
Another critical methodological area relates to information mobiliza-
tion strategies. Focusing on a rural area in the U. S. which has little
secondary data available for assessing its basic socioeconomic constraints
and needs is analogous to the conditions prevailing in some LDCs. Experience
in determining factors affecting socioeconomic conditions in this country's
rural areas provides the researcher with important insights into the pro-
blems of limited resource rural people in developing countries. Methodo-
logy has focused on household and farm survey techniques and extensive in-
terviews, as well as careful quantitative and non-quantitative analysis of
data. Since primary data collection expertise is essential for socioecono-
mic research in developing countries, the field experience and the research
methodologies developed in the laboratory area encourage the attainment of
the skills needed to gather information in developing countries.
(2) International focus
VSU, utilizing its experience in field work and the development of
methodology for resolving socioeconomic problems in rural South Central
Virginia, developed an international program with similar focus for imple-
menting in LDCs.
The program was initiated with a bilateral linkage agreement with the
University of Science and Technology (UST) in Kumasi, Ghana. This UST/VSU
linkage program has enabled staff members at VSU to increase significantly
their capability to identify, design, analyze and evaluate rural development
projects in LDCs.
Research carried out under the UST/VSU linkage program includes:
(1) CAPITAL, CREDIT AND RURAL FINANCIAL INSTITUTIONS
(a) Enterprise Combination and Capital Requirements in Northern
Ghanaian Agriculture: Case Study of the Kumbungu Area
(b) An Evaluation of Institutional Credit and its Role in Pro-
duction in Ghana
(c) The Role of Credit in the Adoption of New Technology. A
Case Study of Small Rice Farmers in Northern Ghana
(2) LAND TENURE AND AGRICULTURAL PRODUCTION
(a) Customary Land Practices in Relation to Agricultural Pro-
(b) Land Tenure Systems in Relation to Agricultural Production
The program afforded VSUstaff to spend approximately five person years
in Ghana (at UST) in research, teaching and the organization of conferences.
UST staff spent approximately two person years at VSU on similar activities.
(3) APPROPRIATE TECHNOLOGY
(a) Production of Agricultural Implements for Small-Scale Rice
Farmers: Socio-Economic Aspects
(4) SMALL FARMERS AND RURAL DEVELOPMENT
(a) Supply Response of Cotton Farmers in Northern Ghana
(b) Economic Viability of Small Scale Farmers in the Forest
Zone of Ashanti
(c) The Economics of Small Farm Systems and Socioeconomic Con-
ditions in the Atebubu District Ghana
(d) Marketing and Infrastructure-Food Marketing, Transportation,
and Small Farm Production in Ghana- A Case Study of the Ate-
For the most part the above research projects cover a range of issues
that reflect an attempt to understand the farm, farmer and the farm environ-
ment as a complex system of interdependent parts. Secondly, they reflect
an attempt to determine the intervention strategy by delineating the con-
straints posed by the internal and external farming environment. And
thirdly, they reflect an attempt to utilize an interdisciplinary/multi-
disciplinary framework of analysis.1
B. FARMING SYSTEM RESEARCH AND DOMESTIC INTERNATIONAL LINKAGES
While VSU might not be following the letter of FSR approach, its
rural development strategy does reflect the spirit of FSR. It is farmer
focused; it is interdisciplinary and it is systems approach oriented.
Further, the target group studied- small limited resources farmers- is
the same in both Virginia and West Africa (Ghana). The structure and
capacity for growth differ only in degree and not in kind. Our research
in South Central Virginia indicates that the limited resource farmers
1D. Rohrbach, (9)
D. Rohrbach, (9).
(1) are shifting from farm to non-farm activities (off-farm employment),
(2) Do not necessarily belong to a.homogeneous group, (3) are under-
utilizing land resources because of a lack of appropriate equipment and a
lack of access to capital, (4) can, with given resources, increase their
net income through better planning and better utilization of available re-
sources, (5) have limited access to appropriate technology, (6) interact
with support systems (marketing facilities, transportation, agricultural
knowledge system) that can be improved. As a result, they make less than
optimal contribution to socioeconomic development. With respect to the
limited resource farm families in Ghana all of the above problems affect
their operations albeit to a greater degree. The important point here is
that the nature qf the problem, being the same, there seems to exist real
opportunities to develop lessons of experience that serve both rural
American and rural Ghana. In short, there are links that can be forged
between these two environments as regards rural development problems and
solutions. The linkages are not sp much concerned with technological
practices only a relatively small part of which can be transferred any way,
but more with the management and administration of agricultural and rural
change, including production, output and input systems and the associated
questions of prices, capital, and credit; articulation of official ser-
vices with one another and with services provided by the private sector;
groupings of producers; and management at the subdistrict level, rural
level where the real producers are. Developed countries have evolved
1See Sammly L. Comer and R. C. Woodworth, Improving Incomes on Limited
Resource Farms in South Central Tennessee, Tennessee State University.
Nashville, Gennessee Bulletin #36 Oct. 1976.
Quanda L. Cooper- An Economic Analysis of Small Farm Production in
Swine Operations of South Central Virginia MA thesis, Virginia State Uni-
versity, Petersburg, VA 23803.
mechanisms for dealing with the whole process of managing rural develop-
ment. LDCs can adapt these mechanisms to their particular locale avoiding
the inconsistences that accompained their evolution. Plans for increasing
incomes of limited resources farmers in Virginia and Ghana deal with a very
complex undertaking. The complexity is not due to a technology problem in
the narrow sense, but to the administration and management of rural develop-
ment, because whether incomes can be increased will depend on the extent
to which the appropriate administrative mechanisms can be developed to
manage the cooperation among researchers/research institutions, extension
services, the delivery system of support systems and the limited resources
BERD/VSU has recently established a Small Farm Development Center,
whose major task will be the development and testing of a small farm model
that can be used as a framework to mobilize the potentials of limited re-
source farm families and enhance their productive capacity. Given the
similarity of the nature of the management problems in rural sectors of
Virginia and Ghana, it is expected that such a model with appropriate
modifications could be used effectively in less developed countries
C. CONSTRAINTS ON THE FARMING SYSTEMS RESEARCH APPROACH
Among the many major institutional, technical and managerial con-
straints that affect the implementation of the FSR appraoch the severest,
as far as limited resource small farmers are concerned, are those gene-
rated by the external environment. Of particular importance is the
agricultural knowledge system (AKS) and its management.
The ideal AKS consists of:
(1) Human capital (existing stock of knowledge)
(2) Means of expanding the stock of knowledge-research
(3) The means of preparing accumulated knowledge for use on
a practical level development
(4) The practical means of applying the knowledge or technology
to particular purposes- extension subsystem
(5) Education and training subsystem
(6) Farmers and farmers' organization
An effectively managed national AKS must
(1) Coordinate all of the following structures/functions:
(a) Information subsystem
(b) Education subsystem (primary, secondary, tertiary)
(c) Research institutions
(d) Extension services
(e) Education and training of rural people
(f) Farmers and farmer's organizations
(2) Produce trained personnel at all levels including technicians,
and in all relevant subject areas
(3) Provide the stability for monitoring and evaluation
(4) Provide for field service- extension services- to be a part
of the system as an essential partner in helping to study the
existing system and the multiple purposes and real difficul-
ties of producers, to help select objectives for research and
to monitor changes
(5) Involve producers especially small farmers
Experience in Ghana indicates that the components of the knowledge
system (the necessary conditions) are for the most part in place; however,
their coordination (sufficient condition) is either non-existent or mal-
function. Despite the existence of universities with agricultural facul-
ties, a structure of research institutes covering major areas of agricul-
ture, agricultural training programs and an extension service,1 national
coordination is weak and there is significant inbalance in emphasis on the
technical as opposed to the social science aspect, too little emphasis on
the development of appropriate technology, and inadequate linkages between
national researchers and national research users. It is possible to per-
ceive a real linkage between the South Central Virginia and Ghana in the
sense that in both areas the nature of the small farm problems is basical-
ly the same: the small farm sector is a relatively neglected sector,
'technology' needs are urgent, comprehensive information on the small farm
problem is non-existent, although from a technical standpoint, some of the
modern technology is scale neutral, operationally, small farmers are unable
to adopt because of the effectiveness of the constraints imposed by the
knowledge system as reflected in the form of inefficient dissemination of
technology and for inadequate extension services and system.
(2) Further Constraints: Applicability of FSR Appraoch.
The constraint on FSR by the management of the knowledge system of
LDCs can be traced to the evolution of the system. In many LDCs agri-
cultural research developed in the preindependence period much earlier
than agricultural education at the primary, secondary and tertiary levels.
Moreover, such research as developed was concerned with development of
export crops from the plantation sector, each of which had its own re-
search institution. Research was done by expatiates for the technical
staff of the plantation. Extension services were needed only where cash
crops were grown by small-scale producers (Cocoa-Ghana). The legacy of
this development is that in many LDCs the academic components of the know-
1Two major universities with schools of Agriculture and one concerned
with the training of agricultural (vocational) teachers; under the Council
for Scientific and Industrial Research (CSIR), of the nine research insti-
tutues, five are concerned with agriculture related research- soil, crops,
food, forest products, livestick, etc. Four institions for training exten-
sion workers- See Barron and Mensah (1), National Extension Services.
ledge system is neither organizationally or operationally linked to the re-
search component. Where this is the case, there is a tendency for separa-
tion between agricultural research and the human sciences. This operational
style puts severe constraints on attempts to develop multidisciplinary under-
standing and cooperation needed to describe, analyze and improve the existing
resource- use systems. More importantly it constrains the academic institu-
tions' ability to associate with ongoing efforts of the government in agri-
cultural research or in agriculture and rural development. Even where
attempts have been made for closer linkages, it has not been easy to build
strong associations between the universities which are either autonomous
or associated with the ministry of Education. Knowledge systems which re-
flect such weak linkages fail to provide the professional personnel through-
out the system with the professional growth that is expected to occur within
It seems useful to distinguish between knowledge systems with weak
coordination and those with strong coordination. The former will be re-
ferred as centrifugal systems reflecting a lack of consensus among the
disparate groups composing the system, and the latter centripetal system
reflect the existence of such consensus. It makes quite a difference to
the applicability of FSR if one or the other system is dominant in a given
country. In countries where the centrifugal system predominates (Ghana)
FSR practioners will have to consider using an interventionist, as opposed
to the submissive approach.
(1) Barron J. and Mensah I. Administrative and Organizational Issues in
Agricultural Research, Department of Agricultural Economics,
University of Ghana, August 1977.
(2) Bunting, A. H. Science and Technology for Human Needs, Rural Develop-
ment, and the Relief of Poverty, Occasional Paper, International
Agricultural Development Service, 1979.
(3) Cox, M. L. A Symplied Approach to Agricultural Systems, Occasional
Paper, International Agricultural Development Service, 1979.
(4) Comer, Sammy L. and Woodworth, R. C. Improving Incomes on Limited
Resource Farms in South Central Tennessee, Tennessee State Uni-
versity, Nashville, Tennessee. Bulletin No. 36, October 1976.
(5) Cooper, Quanda L. An Economic Analysis of Small Farm Production in
Swine Operation in South Central Virginia MA Thesis, Virginia
State University, 1979.
(6) Hinman, H. R. Management Training for Agricultural Extension Officers
Within Ghana- Successes and Shortcoming (Mimeo).
(7) Lionberger H. F. and Chang, H. C. Development and Delivery of Scien-
tific Farm Information: The Taiwan System as An Organizational
Alternative to Land Grant Universities- U. S. Style (Mimeo).
(8) Norman, D. W. General Overview of Farming Systems Research (Mimeo,
(9) Rohrbach, D. D. A Discussion of Issues Relevant to Development and
Implementation of a Farming Systems Research Program USDA/OICD
October 3, 1980.
Howard W. Kerr, Jr.
U.S. Department of Agriculturel
To prepare for the Farming Systems Research symposium, I read the paper of
Robert Chambers, University of Sussex, entitled "Understanding Professionals:
Small Farms and Scientists". Of particular interest was Mr. Chamber's
statement "The challenge is to listen to and learn from farmers, encouraging
them to express their categories, meanings, and priorities, and treating them
not just as professional colleagues and collaborators, but as teachers."
Viewing farmers as teachers is an approach, I believe, that is relevant to
domestic small farms research and specifically to the Northeastern Region (NER)
small farms research program. To appreciate this idea, one must first under-
stand the framework within which the NER small farms research program operates.
Agricultural Research is one of the three major subunits that comprise the
Science and Education Administration (SEA). Cooperative Research and the
Extension Service complete the structure of SEA. Each agency has specific
responsibilities and,in the most general terms, Cooperative Research and
Agricultural Research are, respectively, the State and Federal research arms
while Extension Service is the educator and disseminator of information. Prior
to the formation of SEA, those agencies had all functioned as independent agencies
within the Department. United under SEA, they were consolidated to form the
Department's science (research) and education (dissemination) thrusts.
Small farms research efforts began in FY 1979. In February 1979, the NER
Administrator named a coordinator for the program, charging him to "make the
approximate $1.5 million small farms research funds accountable, tractable,
visible, and yield tangible results." Seventy-five percent of these funds were
designated for in-house research and 25 percent for extramural research to be
conducted by State Agricultural Experiment Stations and institutions located
in the Northeastern Region. The target areas of research were selected and
identified by the SEA National Program Staff small farms committee.
Quickly, it became evident that a plan was needed to identify specific research
needs of small-scale farm operators so that limited AR small farm funds could
better be targeted to developing appropriate technologies. To develop a plan,
a survey questionnaire was sent to Extension Service county agents, requesting
information relative to small-scale agriculture in the Northeast. A definition
of small farms operations was provided. Questions were limited to 2 pages and
were designed to provide a profile of small-scale agriculture in that county,
particularly the immediate research needs, specific directions for future
research, and technology needed in 1984. The questionnaire also asked how
small farm operators would be affected in the next 5 years, as the economy
rapidly changed due to energy shortages, tight money, and expected higher food
Coordinator, Small Farms Research, Northeastern Region, Science and Education
Administration, U.S. Department of Agriculture, Beltsville Agricultural Research
Center, Eeltsville, Nd. 20705. Paper presented at the Symposium on Farming
Systems Research on December 8, 1980, Jefferson Auditorium, USDA, Washington, D.C.
SEA's Agricultural Research (AR) is a storehouse of knowledge, however, most
of the NER scientists are located in centers such as the Beltsville Agricultural
Research Center or Eastern Regional Research Center. These citadels tend to
restrict the scientists' exposure to the full breadth of real world problems,
particularly those of small-scale farms. Further, because the scientists are
experts in a particular agricultural science, their interest and expertise
will necessarily be focused on the scientific discipline in which they are
academically trained. Thus, AR scientists were not the best respondents to
On the other hand, Extension Service (ES) county agents are deployed in every
county of the United States, where they can better understand and further
educate the agricultural sector. This ubiquitous force of county agents havi
daily contact with the farming community. Thus, a selected group of 70 ES
county agents in the NER were identified to respond to the mailed questionnaire.
An analysis of their responses provided the major thrust for a relevant small
farms research plan.
The hallmark of the NER small farms research program is the communication
between AR and ES. This approach focuses on rectifying small-scale farm
operators' needs by targeting the limited resources of each agency to yield
maximum returns. ES has grass root contacts with small farmers and thus can best
identify the needs of small farmers. This information is needed by AR scient-.sts
when deciding the appropriate technology to develop within the constraints of
what is feasible considering the current capabilities of science-at-large for
small farms. AR scientists will publish their findings or results; however,
to more quickly reach small farmers, the new knowledge should be presented in
a SEA AR/ES communication. Disseminating information to the small-scale farm
operator, remains the responsibility and work function of ES. With proper
communication between research and extension,each can perform its role in a
team effort. The team is perhaps more like a track team than a football teaw.
but, nevertheless, each can take complete pride in their achievements.
-Kerr, Howard W. 1980. A Survey of Current and Expected Research Needs of Sm. 11
Farms in the Northeastern Region. ARR-NE-9, USDA/SEA, Beltsville, Md 2;5705
December 4, 1980.
INITIATING APPLIED FARMING SYSTEMS RESEARCH
IN DEVELOPING COUNTRIES*
Farming systems research (FSR), in its various manifestations, is re-
ceiving increased attention as a means to stimulate agricultural development.
During the last several years, donors, international agricultural research
centers (IARC's) and a number of national research programs have developed
alternative methods for conducting FSR. There has recently been a substantial
effort on the part of these "practitioners" to pool their knowledge and share
their experiences. The result has been a remarkable degree of consensus on
procedures for FSR, at least when the scope and purpose of research are carefully
specified, e.g., applied FSR conducted by national programs versus basic FSR
conducted by IARC's (Harrington, 1980).
Although research on FSR procedures continues, a related issue is gaining
prominence: implementation of applied FSR in national research programs. The
issue is clearly one of feasibility. How feasible are FSR procedures for na-
tional programs, given the various constraints under which they operate? What
are the basic decisions that must be taken to initiate FSR? What steps are nec-
essary to make FSR operational? The purpose of the present paper is to address
some of these questions.
* Paper presented at the AID-USDA Symposium on Farming Systems Research,
Washington, D.C., 8-9 December, 1980.
** Economist at CIMMYT, Mexico. The opinions expressed are not necessarily
those of CIMMYT.
In order to comment on the feasibility of implementing applied FSR in
national programs, however, a characterization of FSR is needed. Applied FSR
is viewed as research that is:
1) problem-oriented -- research focuses on the solution to production
problems that promise to have a considerable effect on farmer goals (e.g.,
increased income at reasonable levels of risk).
2) on-farm -- experimentation is planned and conducted in light of farmer
circumstances and under farmers' conditions.
3) multi-disciplinary -- the effective collaboration between biological
and social scientists is needed in the design and testing of new technology.
The approach to applied FSR used at CIMMYT may b.. seen in Figure 1.
On-farm experiments are planned in light of farmer problems and circumstances.
Experimental results are analyzed in order to formulate recommendations. Those
recommendations are subjected to farmer assessment before being promoted among
target farmers (Byerlee,.Collinson et al, 1980).
The very characteristics of applied FSR (farmer-oriented, on-farm, multi-
disciplinary), of course, lead to many of the issues and difficulties in imple-
mentation that will be discussed in subsequent sections.
2.0 National Research Programs: Constraints and Circumstances
There exists a great diversity among developing countries with regard to
the structure and operation of their agricultural research programs. Nonetheless,
the majority of national programs share at least two salient characteristics:
(1) an institutional environment that is not entirely sympathetic to the intro-
duction of FSR, (2) a scarcity of research resources.
SFigure 1. Over.icw of an Integrated Research Program
Choice of Target Farmers
and Research Priorities
National goals, input sup-
ply credit, markets, etc.
of Policy Issues
Obtain a knowledge and un-
derstanding of farmer cir-
cumstances and problems
to plan experiments.
Conduct experiments in
farmers' fields to formulate
Improved technologies un- a
der farmers' conditions. o
3. Recommend ,u
Analyze experimental re-
suits in light of farmer cir-
cumstances to formulate
Determine farmers' experi-
ence with technologies.
technologies to farmers.
New Components Incorpo-
rated into OrnFarm Research
Developing and screening
new technological com-
ponents (e.g., varieties,
new herbicides, pesticides)
Identification of Problems
for Station Research
The institutional environment often constitutes a major obstable for
the initiation of FSR. The agricultural research system itself, usually orga-
nized along disciplinary or commodity lines, often resists the introduction of
FSR because of the consequent need to re-define institutional jurisdiction.
Individual researchers, having been trained as disciplinarians, frequently see
FSR as a threat to traditional research activities. Even researchers who see
a need for FSR have little incentive to push for its introduction. The current
structure of incentives (salary increases, promotions, peer approval) normally
favors those who excel as disciplinarians.
Other institutions, outside of the established agricultural research
system, may also prove to be obstacles. On-going rural development programs
may frown on the introduction of FSR into their areas of influence, foreseeing
that FSR may discredit some of their past efforts. Even extension services may
see FSR as an usurpation of their traditional prerogative of linking with farmers.
Institutional difficulties are made more acute by financial constraints.
FSR can be expected to produce superior results when the research teams are
composed of highly-skilled personnel who control a budget adequate for survey
and experimental field work. Mobility is especially important: researchers must
be able to visit experiments and farmer-collaborators frequently.
Skilled personnel and funds for recurrent expenses are, however, espe-
cially scarce in LDC agricultural research programs. In many countries there are
few social scientists available for work on FSR. There are even fewer (and in
many cases, none) who have either training or experience in collaborating with
agronomists in problem-solving research. Similarly, resource limitations can
strongly affect the conduct of FSR. In Ghana, fuel for staff vehicles was ex-
tremely scarce during the initiation of an FSR project. Researchers had to use
considerable ingenuity to obtain a modest fuel allowance for on-site visits.
In the face of resource limitations, then, the demands of an FSR program must
either be met by diverting resources from on-going research activities or by
obtaining special-project funding.
Researchers engaged in traditional, on-station research can point, in
defense of their budgets, to recent studies indicating a high rate of return to
their past activities (e.g., Arndt et al, 1977). However this may be, FSR practi-
tioners believe that their approach is far more cost-effective in terms of "tech-
nology adopted per monetary unit spent" and that FSR deserves a chance in most
3.0 Initiating Applied FSR
A decision by a national agricultural research program (or by individual
researchers in such a program) to "give FSR a chance" leads directly to other
decisions on organization. These decisions focus on the institutional character
of FSR, the proper scope of research and the choice of target area for the initial
FSR enterprise. As the organization of FSR takes shape (at least for initial
efforts), increased attention may be paid to operational questions, such as
training researchers in FSR procedures, insuring effective multi-disciplinarity
and getting research onto farmers' fields.
3.1 Institutional Character of FSR
An important decision that must be addressed early is that of "program
versus project". As described in the CID Draft Guidelines (Consortium for In-
ternational Development, 1980), the "program versus project" choice is one of
scale. Initiating FSR as a project merely entails the assignment of a team of
researchers to conduct FSR in a given target area. It is not necessary that
the supporting agency make any large changes in research philosophy or organi-
zation. The project approach to FSR can be especially painless (in a budgetary
sense) because special-project funding may be available for initial FSR efforts.
The program approach is more ambitious, requiring a major commitment to
FSR from the beginning. An FSR program may be carried out through the re-orien-
tation and co-ordination of the activities of existing agencies or through the
creation of a new agency with a mandate for FSR.
The decision made on the "program versus project" issue will depend on
the specific circumstances of the case. However, the project approach to FSR
is more attractive under conditions of institutional neutrality or hostility,
and scarcity of research resources. Given that these are the very conditions
that most national programs must face, the project approach to FSR has proven
to be relatively popular. Within national programs, it is possible to "start
small", without major institutional rearrangements.
3.2 Scope of Research
Whatever the scale of FSR organization that may be adopted, another
issue must immediately be faced: the appropriate scope of research. It is
patently impossible to consider everything as variable at the commencement of
research: some things must assumed as fixed, outside the scope of research.
The phrase "farming systems research" carries with it a connotation of wholeness,
a feeling that ll management practices in the farming system must be considered
as variable. However, there is no reason why some of these management practices
may be not also be classed as fixed. The issue, then, is how many practices
should be allowed to vary during research -- many or few?
One alternative is to allow many variables, considering for example the
selection of crop or livestock enterprises for a farming system (and the manage-
ment of each) as variable. This has been called "FSR-in-the-large". Another
alternative is to focus research on the management of one (or few) enterprises
in the context of the current farming system. This has been called "FSR-in-the-
small". The point of importance is that this focus on one or few enterprises
is still a valid form of FSR. This is because the design and analysis of research
on these few, selected enterprises takes into account effects of technological
change on other components of the farming system.
A focus on one (or few) enterprises in FSR is likely to be particularly
wise under the following conditions:
(1) The bulk of farmers' resources are used in one enterprise (e.g.,
maize in much of Central America, wheat in N. Africa) so research on this enter-
prise is likely to have the best leverage on such system problems as deficient
income, excessive risk, or seasonal variability in the employment of farmer-owned
resources (Collinson, 1980).
(2) There appears to be some scope for improvement in the management of
the pre-dominant enterprise. In fact, our experience indicates that, in the
case of maize or wheat, improvements in crop management can have substantial
impacts on production and farmer income.
(3) Land-use is not highly intensive (although even here a single enter-
prise focus will often lead to useful innovations as long as system interactions
are carefully considered).
A focus on the most important enterprises) will, in addition, normally
lead to lower research costs. The amount of information on farmer practices and
circumstances needed to plan research (describe representative practices, identify
priority production problems, pre-screen potential solutions to these problems) I
is less when research is well-focused. Similarly less data is needed for evalua-
tion of experimental results. Even farmer assessment of new technology is facil-
itated because recommended improvements on the current farmer practice tend to
be relatively uncomplicated, making it easier for farmers themselves to evaluate
Needless to say, the specific choice of target enterprises) will vary
from target area to target area. The issue to which reference is being made is
one of scope: in general, should researchers try to improve the management of
one (or few) major enterprises in a farming system or should they attempt a
broad-brush re-design of that system? National research programs interested
in FSR should consider a relatively simple and inexpensive form of FSR -- that
3.3 Choice of Target Area
Another major decision in the initiation of FSR concerns the target area,
or region where initial research efforts are to be carried out. While it is
conceivable that an FSR program may wish to initiate research in many areas si-
multaneously, FSR is usually initiated in one or few target areas, both because
of resource limitations and to allow researchers to gather experience in FSR
procedures. The characteristics of the initial target area, however, may influ-
ence the success of the FSR activities therein and, consequently, can affect the
probability that FSR will be extended to other regions, as well as the form it
Several criteria may be considered in the selection of an initial target
area: (1) The possibility of gaining tangible results in a reasonable period of
time. This is particularly important when the credibility of FSR is at stake,
Later, more difficult areas can be selected -- when greater researcher expe-
rience can help compensate for more difficult research problems. (2) The
presence or absence of major bottlenecks (human resources, logistics) to
the implementation of FSR. At least initially, accessible areas are pre-
ferred. (3) The likely effect of a success in FSR on such national ob-
jectives as lower food prices for urban consumers, rural development, income
distribution, etc. For example, when food prices are of primary concern, a
different target area may be chosen than when the welfare of small farmers is
given top priority (Byerlee, Collinson, et al, 1980).
3.4 The Incorporation of Social Scientists in FSR
A key step in the initiation of FSR is the formation of research teams
with effective collaboration between biologicall and social scientists. The role
of the social scientist in FSR is now widely regarded as essential. The social
scientist is responsible for incorporating the "human element", or socio-economic
problems and circumstances that affect farmers' decisions, into the design and
evaluation of new agricultural technology. He must take as much responsibility
as breeders or agronomists for such research activities as choice of representa-
tive farmer-collaborators, selection of high-priority experimental variables,
selection of the level of non-experimental variables, and evaluation of experi-
Effective multi-disciplinary is not, however, easy to achieve. An effec-
tive multi-disciplinary research team is composed of researchers who enjoy a
sound base in their respective disciplines but who nonetheless work to meet a
common objective: the development of new technology useful to farmers. That is,
they must be "task-oriented". Collaboration is facilitated when one discipli-
narian is aware of the questions that fruitfully may be asked of another (Bartlett
and Akorhe, 1980).
Team effectiveness can be influenced by experience (or its surrogate,
training) and team organization. The more experience researchers have in multi-
disciplinary, problem-solving research, the easier subsequent collaboration
becomes. In the initial stages of FSR, however, such experience is in short
supply. Training team members in FSR philosophy and methods can help overcome
this initial hurdle. Although some IARC's have begun FSR-related training pro-
grams, there is still a scarcity of such programs that focus on the practical,
multi-disciplinary aspects of FSR. (See Gilbert et al, 1980, Chapter 7 and
Team organization can influence the quality of multi-disciplinary colla-
boration by pre-determining both the role and status of social scientists. In
one country, economists of a junior level were recruited to work as staff members
on experiment stations, in order to encourage and collaborate in research on
farmers' fields. Their relatively low status in comparison with their agronomist
colleagues, however, proved to be a serious obstacle to their attempts to initiate
FSR off-station. In another country, social scientists were involved in the
planning and evaluation of on-farm research -- but as a "support unit", not as
full-fledged members of a target area research team. Although useful information
on new technology was produced, the absence of social scientists during the ex-
perimentation stage led to inadequate contact with farmers during that stage,
with a considerable loss of farmer feedback on the characteristics of new tech-
In yet another case, agronomists and economists were organized as a target-
area-specific research team, with joint responsibility for research decisions.
This is normally regarded as the best form of team organization. Nonetheless,
effective multi-disciplinary collaboration proved to be elusive because the
relatively inexperienced economists demonstrated a preference for independent
projects. While formally sharing responsibility for team decisions, in practice
they preferred to work alone in conducting extended farm surveys that were only
marginally related to on-farm experimentation.
In summary, a trained, experienced and task-oriented team of biological
and social scientists, with joint responsibility for research decisions in a
given target area appears ideal. Frequently, however, national program circum-
stances and policies will not allow the use of this ideal. In extreme cases,
agronomists may be specially trained to provide a social science input. When
a few social scientists are available, they may be either attached to experiment
stations or placed into a "support unit" to back-up several research field teams.
In these cases, however, the social scientist will usually be forced to cope with
status problems and with an increased difficulty in acquiring an intimate acquain-
tance with farmer circumstances for a given target area.
3.5 Moving Research Onto Farmers' Fields Farmer Collaboration
With few exceptions, research whose purpose is to provide useful new tech-
nology to farmers within a reasonable period of time must be conducted on-farm.
This is because experiment station circumstances (e.g., soil fertility, weed
and insect population, irrigation and drainage) are normally unlike those faced
by target farmers.
Within a target area, then, the primary responsibility of the selected
multi-disciplinary research team is to plan, conduct and analyze on-farm experi-
ments. Armed with a knowledge of farmer problems and circumstances, researchers
are in a good position to field-test promising solutions to important production
problems. However, the team must avoid the temptation to conduct "experiment-
station research" on farmers' fields -- i.e., in isolation from farmers.
On-farm research begins with a review of secondary data on the target 1
area and surveys of target farmers, to ascertain their agro-climatic and socio-
economic circumstances. This information can be used in a variety of ways in
planning on-farm experimentation: farmers may be pooled into roughly homogeneous
groups, or recommendation domains. For each domain, important production problems
can be identified and possible solutions to these problems pre-screened for com-
patibility with the current farming system and for effect on farmer welfare.
A wide variety of experiments can be used in on-farm experimentation.
Complete factorials (and in some cases "super-imposed trials") aid in identifying
production limiting factors. Economic levels of inputs and practices that affect
production can be estimated through researcher-managed, replicated experiments.
Large-plot, unreplicated trials can verify the attractiveness of an innovation
when compared to the current farmer practice (Palmer et al, 1980). Finally,
farmer-managed trials allow farmers themselves to assess the attractiveness of
The key to effective on-farm research, however, is the maintenance of
close contact with farmers. During the initiation of FSR, research teams often
lose sight of this point.
During on-farm research, farmers should collaborate in various ways:
respond to surveys, loan their fields, provide a reference point from which the
effects of innovations may be measured (farmers' practice), comment on alternative
experimental treatments, completely manage fields in which promising innovations
are added to their current practice, and provide feedback on the reasons for
liking or disliking proposed new technology.
Clearly, on-farm research without effective farmer collaboration can
easily degenerate into a series of sterile exercises. Continuous contact with
farmers is usually needed to maintain the "farmer-orientation" that characterizes
4.0 Low-Cost FSR
The feasibility of FSR for national research programs is a question that
is currently attracting considerable attention. This question frequently takes
the form of the cost-effectiveness of FSR, given its apparent expense. Some ways
in which resource-poor research programs can initiate FSR activities have already
been presented: (1) FSR can be initiated as a small-scale project instead of a
large-scale program, (2) FSR can focus on important issues in the management of
major crops (in the context of the farming system) instead of attempting the
wholesale re-design of that system, (3) FSR can be initiated in an accessible
target area in which prospects for success within a reasonable period of time
are high, (4) FSR researchers can practice effective multi-disciplinary collabo-
ration and close collaboration with farmers.
FSR practitioners claim that their approach is more "cost-effective"
than traditional research in terms of "technology adopted per unit of money
spent". They believe that traditional, on-station research in developing
countries has led to relatively little adoption of new technology by small
farmers. Nonetheless there is still a current of concern about the expense of
FSR. It may be a worthy exercise, then, to list more ways in which FSR expense
can be reduced.
A principal way to increase the effectiveness and reduce the cost of FSR
is to stratify the farmers within a target region into relatively homogeneous
target groups or "recommendation domains". A recommendation domain is merely a
group of farmers facing similar problems and circumstances, operating similar
farming systems, and for whom similar recommendations will be appropriate (Byerlee,
Collinson et al, 1980). By placing all on-farm research in this context, one
avoids the two extremes that serve as alternatives: (1) assume that research
results will be appropriate for all farmers in a target area, heterogeneity in
circumstances notwithstanding, (2) formulate separate recommendations for indi-
The acquisition of data on farmer problems and circumstances provides
another area of possible cost reduction. Informal, non-probabilistic surveys
and well-focused, single-visit small-sample formal surveys are preferable to
large-sample surveys or frequent visit surveys in this connection. The criterion
for the selection of survey instrument should be that of "the lowest cost com-
mensurate with the degree of understanding that is necessary." (Norman, 1980).
Another area of possible cost reduction is that of expanding the universe
for which FSR results are applicable. This may be performed by determining the I
transferability of one set of results to other similar environments (i.e., extra-
polating recommendation domains into new target areas). In this fashion, some
(but rarely all) of the steps in FSR may be skipped.
Finally, it should be pointed out that it is unnecessary for FSR to pro-
duce the "best" new technology for farmers. Insofar as it discovers anything
"better" than the current farmer practice, it will be useful. Thatis, FSR need
not engage in the fine tuning of the farming system, but rather may concentrate
on seeking the best of readily available solutions to important problems.
FSR procedures have been seen as fairly adaptable to the varying circum-
stances of national research programs. As a consequence, national programs
wishing to commence FSR are faced with serious decisions. Among these are the
questions addressed in this paper:
Should FSR be initiated as a large-scale program or a small-scale
Should FSR focus on the management of major existing enterprises in
the context of the current farming system or should it attempt large changes
in that system?
What are the desirable characteristics of a target area for the initia-
tion of FSR?
To what extent should effective multi-disciplinary collaboration be a
priority objective? How can it be achieved? How should training be organized
to prepare field team researchers?
To what extent should FSR be conducted on-farm? How, and to what
extent, should farmers collaborate in FSR?
How may FSR be made more cost-effective?
No single methodology for FSR can be defined for use by all national
research programs. However, even resource-poor national programs operating in
an institutional environment not wholly sympathetic to FSR can and must afford
to "give FSR a chance".
Arndt, T.M., D.G. Dalrymple and J.W. Ruttan, 1977. Resource
Allocation and Productivity in National and International
Agricultural Research. Minneapolis: University of Minnesota
Bartlett, C.D.S. and J.A. Akorhe, 1980. "Interdisciplinary Co-
operation to Identify Innovations for Small Farmers -- The
Role of the Economist". IITA.
Bernsten, R.H., 1980. "Cropping Systems Research in a National
Program: Indonesia", presented at the AAEA Annual Meeting,
Champaign Urbana, July 28-30, 1980.
Byerlee, Collinson, et al, 1980. Planning Technologies Appro-
priate to Farmers: Concepts and Procedures. CIMMYT Economics.
CIMMYT, 1978. "CIMMYT Training". CIMMYT Today No. 9 .
Collinson, M.P., 1980. "Some NoLes on the Farmer as the Client
for Research" presented at the Workshop on Methodological
Issues Facing Social Scientists in On-Farm Farming Systems
Research. CIMMYT, April 1-3, 1980.
Consortium for International Development, 1980. "Guidelines to
Assist National Government in Implementing Farming Systems
Research and Development Programs Aimed at Farmers with
Limited Resources", second draft.
Gilbert, E.H., D.W. Norman and F.W. Winch, 1980. "Farming Systems
Research: A Critical Appraisal", E. Lansing: Michigan State
University, Rural Development Paper No. 6.
Harrington, L.W., 1980. "Methodological Issues Facing Social
Scientists in On-Farm/Farming Systems Research", CIMMYT
Norman, D.W., 1980. "General Overview of Farming Systems Research",
presented at the Workshop on Methodological Issues Facing
Social Scientists in On-Farm Farming Systems Research.
CIMMYT, April 1-3, 1980
Palmer, A., A. Violic and F. Kocher, 1980. "In-Service Training
in Maize Production Agronomy at C.MMYT" presented at the
Annual Meeting of the American Society of Agronomy, Detroit,
November 30 to December 5, 1980.
Problems of Interdisciplinarity in
Farming Systems Research*
Scientists argue about the definition and meaning of "farming systems."
However, there is general agreement that.farming systems research requires a
wholistic approach and this in turn implies a team or interdisciplinary re-
search effort. In this paper I propose first to define the problems asso-
ciated with interdisciplinary research in farming systems. Then I will dis-
cuss the implications for training. However, before discussing the problems,
I would like to characterize the various forms of interdisciplinary research.
In the early part of this century there was little if any formal in-
terdisciplinary research. However, the research staffs of the college of
agriculture were small. As a result there was a good deal of informal in-
teraction among disciplines, just as there is today in the international
agricultural research centers. Under this environment each individuals'
disciplinary research is likely to reflect to an important degree the in-
fluence of other disciplines.
As the disciplines grew, however, communication among disciplines in the
agricultural colleges declined until today it is not uncommon for scientists
to work and socialize exclusively with people in their own discipline. Formal
interdisciplinary research studies in agriculture began in the 1940s. The
earliest that I am aware of is the. research by Jensen et al. (1942) on imput-
output relationships in milk.. In the 1950s and 60a there was a great deal of
* Paper prepared for the AID-USDA sponsored Workshop on Farming Systems Re-
search, December 8-10, 1980, Washington, D.C.
**Professor of Agricultural Economics, Cornell University.
interdisciplinary research of this type conducted between agricultural eco-
nomists and biological scientists particularly at Iowa State and Michigan
State Universities (see for example, Hoffnar and Johnson, 1966). The de-
gree of formal collaboration among researchers varied considerably from pro-
ject to project. In some cases individuals worked together under a broad man-
date that allowed each to pursue.his/her own interests. In other cases a
tightly managed group worked under a project director meeting frequently to
discuss the design and execution of the research and integrating the results
into a coordinated product.
Today it is still more common to find multidisciplinary projects loosely
organized in such a way to enable each individual scientist to do his/her own
thing. An obvious reason for this is the difficulty of conducting truly
integrated interdisciplinary research. The obstacles encountered in under-
taking such research are described below.
Defining the Obstacles
There are three elements in farming systems research which when analyzed
help to explain the difficulties associated with conducting interdisciplinary
work in this area. These three, which I refer to as: (1) group dynamics,
(2) the systems approach, and (3) the farmer'.s system, are discussed in the
subsections which follow.
A considerable amount of research has been conducted on group dynamics
(see for example Hare et al. 1955). Whether one is. engaged. in research or
other types of activities, factors such as size, composition, and leadership
make a considerable difference in the efficient functioning of the group.
Interdisciplinary research adds still another obstacle, the problem of
I communication. In this respect distinction is often drawn to the gap between
social and biological sciences. y .own experience is that there is often even
more difficulty in communication .within the social sciences. Problems of
interdisciplinary communication.unfortunately are as prevelant among scien-
tists from the developing as the developed world. The former have inherited
the disciplinary biases through advanced training in developed countries.
It is my judgement that there are definite limits to the size of a
farming system team. The value of adding an additional member or discipline
must be weighted carefully against the diseconomies of coordinating a group of
larger size. The upper limit would seem to be about six. In the sondeo
approach to establish the research agenda a relatively large group may func-
tion efficiently in the manner described by Hildebrand (1978). But in the
actual conduct, evaluation, and reporting of research even six may be too
large for the core group.
In the selection of team members important factors to be considered are
the willingness and ability of individuals to participate in a team, applied-
research, systems approach and the individuals knowledge and understanding of
farmers and farming. An understanding of systems and farmers is a prere-
quisite for farming systems research.
The systems approach
Adding to the difficulty of working together as a group is the need to
understand and master what for most scientists is a totally different approach
to scientific research, systems analysis. My remarks in this session draw
heavily on a recent article by Dillon (1976) which I recommend as must reading
to anyone concerned with the general topic of farming systems research.
Most of us have grown up in a world.in which the furtherance of the
discipline is taken as more important than the solution of people's problems.
This is but the inevitable consequence of specialization and .reductionism.
Since the 1950s there has been a growing.reaction to this trend and expan-
sionism, teleogy, and synthesis are now being recognized by science as ne-
cessary modes of thought for understanding the world. It follows that a
systems approach, based on teleological concepts and means-ends analysis, is
slowly being introduced as a necessary corollary to the inadequacies of the
old hypothetic-deductive method of research as a means of assessing goal
seeking and goal setting systems. In this new approach, man consumates the
system, and defining the "right" objective becomes more important than making
the "right" choice between alternatives.
Dillon (1976) refers to the systems approach as a new technology. He is
quick to point out that rapid adoption is unlikely since academics represent
perhaps the most conservative purposive element of the agricultural system.
There is, of course, not a single system within agriculture but rather a
whole hiearchy of systems of which the farmer's system is only one sub-system.
However, given the dominant role of the farmer in this sub-system and of man
as either producer or consumer in higher order systems, the implication is
that the socio-economic sphere is more important than the physical and bio-
logical sphere in the choice.of research directions.
Traditionally our research has been production oriented. .This production
orientation has been transferred. to the developing countries through the
training of scientists in our tradition and is still dominant today in the
international agricultural research network. The success of this system in
terms of the traditional orientation of.research has tended to emphasize its
short-comings in the modern context of the systems approach. Improving merely
a part of the system or a sub-system cannot be presumed to lead to enhanced
performance of the overall system. Although it is politically more expedient
in the short run for agricultural scientists to ignore the socio-economic
consequences of their work, there is obviously no guarantee that research
progress in the traditional mode will lead to social justice. The growing
interest in farming systems research stems in large measure from a recognition
that research efforts of the recent past have not resulted in social justice
for the resource poor farmers of either the developed or the developing world.
The farmers system
Gotsch (1977) suggests that "the farmer's system" is a more appropriate
term than "farming systems" since it lays the emphasis where it should be.
Enterprising graduate students in the Agricultural Economics Department at
Cornell recently have been selling for one dollar a pin which reads "No Farm
Background." Most of our graduate students and a good many of our faculty are
qualified to purchase such a pin. And if we were to change the title to "No
Contact with Farmers this Year" even more would qualify.
During the early part of this century students and faculty alike came
primarily from farming backgrounds.. and they brought with them an appreciation
of rural values and a capacity to communicate with farmers. I would argue that
a lack of experience in American farming.is a handicap to many scientists
engaged in applied research to assist the farmers of developing as well as
Often the weakest link in farming systems research is the communication
(or lack there of) between scientist and farmer. This problem and some
suggestions to improve communications are the subject of an excellent paper by
Chambers (1980), "Understanding Professionals: Small Farmers and Scientists."
Unfortunately, much of what passes for communication today takes the form
of an interview questionnaire now referred to professionally as the "survey
instrument". The trend toward the use of the survey questionnaire is as
pronounced in the developing as the developed countries, since surveys tend to
be rather cheap and easy to administer. The problems come later when one
attempts to transfer the data to computer tape and analyze the results.
Without minimizing the important and utility of good surveys, we need to
consider more carefully alternative ways of communicating with farmers parti-
cularly in identifying their goals and objectives, in evaluating their methods
of classifying resources, and in valuing indigenous technological knowledge.
In this task it would appear that rural sociology and anthropology should have
an increasingly important role to play in the farming system team.
Implications for Training
Dillon (1976) indicates that the adoption of the systems approach to
research will require a complete shift in the emphasis on professional
training. He suggests instead of the typical pattern, an initial (one year)
introduction to the systems approach followed.by A (two year) period of dis-
ciplinary specialization, capped off with. the bringing together of different
disciplines in the context of.-some.relevant agricultural system.
In its most fully developed form, this approach would call.for a new set
of majors to replace traditional disciplines such as agronomy, plant breeding,
and agricultural economics. Systems majors might include such titles as crop-
soil systems, plant-animal systems, or farming systems. Research funds would
be reallocated accordingly to multidisciplinary groups or systems teams.
Dillon's view may represent nothing more than a.vision of the distant future
which many scientists would regard with fear and trepidation. But there are
already signs that some of these changes. are coming to pass, not only in the
newly formed international research centers, but also in the more hallowed
universities and colleges of agriculture.
Ealey (1979) describes a program for training environmental managers to
work in multidisciplinary teams. Monash University in Australia offers a
Masters of Environmental Science. A core staff of three faculty organize and
administer the program and coordinate the activities of some 80 other staff
members from all faculties who teach or supervise over 100 candidates.
The program consists of a course work component and a research component.
The course work is intended to broaden insight and provide an opportunity to
improve the depth of understanding in areas of previous training. The research
component is designed to provide practical training in a multi-disciplinary
group research. Teams of three to five candidates, each one from a different
discipline, work together on a part time basis over a two year period. Each
candidate is involved in the production of two documents, a group report
integrating the work of the team and. a minor thesis which.details the work
which each individual performed as his/her. contribution to the.group report.
Efforts at interdisciplinary research and training at Cornell have been
more modest than those described above, being based on individual projects
rather than on a program. Perhaps,. the most ambitious of these involved 14
faculty and graduate students from 7 disciplines in a study of nitrogen and
phospherous in the environment (Porter. 19751.
Another project involved six disciplines at Cornell linked with the
CIMMYT corn program. PhD students from each of .these disciplines -- agri-
cultural economics, agronomy, biometry, entomology, plant breeding, and plant
pathology did their research.on various aspects of corn production.with
field work being conducted in Mexico (Contreras et al. 1977).
A course in water management was initiated in the mid-1970s with in-
struction from faculty in three disciplines agricultural engineering,
agricultural economics, and rural sociology. The faculty teaching this course
developed a US/AID supported research project, "The Determination of Develop-
ing Country Irrigation Project Problems," with field work being conducted in
South and Southeast Asia. The course provides a training ground for graduate
students who have later participated in the US/AID project or have returned to
conduct water management research in their own countries.
Our only ongoing farming systems research project is entitled, "Tech-
nology Introduction into Traditional Farming at High Elevations." Two graduate
students, one from rural sociology and one from plant breeding, are just
completing a study of farming systems in the Andean Mountains of Equador where
they have lived in a small village with their families for the past two years.
Following the model used in water management, we have been experimenting
this fall in the development of a course in farming systems. Ten faculty and
twenty-four students have been involved in.this experiment. Approximately 50
percent of the time was devoted to. weekly.discussions usually based on pre-
sentations of one or two faculty members.. The other 50 percent.of the time
was spent in small group exercises.. Students were divided into four mixed
discipline teams of six members and asked: (.) to present a case study of
farming systems research, and (2) to identify and study a local (New York
State) group of farmers in order to be.able to describe their farming system
and suggest a research agenda.
I will not go into any further details regarding this course, which is
just now coming to an end, other than to. say that it has tended to high light
the barriers to interdisciplinary research that I have described in this
paper. Faculty presentations tended to emphasize the familiar discipline
approach rather than the systems approach. The field exercises proved to be
valuable if for no other reason than providing most graduate students with
their first opportunity to interview American farmers.
In this paper I have not dealt directly with what we understand by the
term "farming systems." But as a minimum it would seem that farming systems
research should consist in its initial stages of an interdisciplinary team
undertaking: (1) a wholestic systems look at the farm and farm family in-
cluding non-farm activities, and (2) an interaction or dialogue with the farmer
or farm family. Both of the above elements should have the purpose of aiding
in the identification of the appropriate researchable issues. Based on these
criteria, there are very few studies in the literature today which could
properly be classified as farming systems.. New concepts of interdisciplinary
training are clearly needed.to prepare .people for research in farming systems.
Chambers, Robert. "Understanding Professionals: Small Farmers and Scientists."
IADS Occasional Paper, International Agricultural Development Service,
New York, 1980.
Contreras, Mario Ruban, Daniel Lee Gait, Samuel Cephas Muchena, Khalid Mohammad
Nor, Frank Byers Peairs, and Mario Santos Rodriquez P. "An Interdiscipli-
nary Approach to International Agricultural Training: The Cornell-CIMMYT
Graduate Student Team Report," Cornell International Agricultural Mimeo-
graph 59, 1977.
Dillon, John L. "The Economics of Systems Research." Agricultural Systems 1
Ealey, E.H.M. "Training Environmental Managers to Work in Multidisciplinary
Teams." unpublished mimeo, Graduate School of Environmental Science,
Monash University, Australia, 1979.
Gotsch, C.H. "The Concept of Farming Systems in the Analysis of Agricultural
Research and Development Programs." mimeo paper presented at the Middle
East and Africa Agricultural Seminar, Tunis, Tunisia, February 1-3, 1977.
Hare, P.A., E.E. Borgatta, and R.F. Bales, eds. Small Group Studies in Social
Interaction. Knopf, New York, 1955.
Hildebrand, Peter E. "Generating Technology for Traditional Farmers A Multi-
Disciplinary Methodology." CIMMYT Asian Report No. 8, October 1978.
Hoffnar, Bernard R. and Glenn L. Johnson. Summary and.Evaluation of the
Cooperative Agronomic-Economic.Experimentation at Michigan.State University
1955-63, Michigan State. University Research Bulletin 11, 1966.
Jensen, Einar, John W. Klein,. Emil Rauchenstein, T.E. Woodward and Ray H.
Smith. Input-Output Relationships in Milk Production. United States
Department of Agriculture Technical Bulletin No. 815, May 1942.
Porter, Keith S. (editor). Nitrogen and Phosphorus Food Production, Waste
and Environment. Ann Arbor Science Publishers Inc., Ann Arbor, Michigan,
FARMING SYSTEMS RESEARCH PROGRAM1
STRUCTURE, STAFFING and FUNDING
Jerry L. McIntosh2
Farming systems studies of various kinds by teams of agricultural
scientists have been a consequence of the concern for development of more
appropriate technology. In countries where most of the food consumed
comes directly from crop production, the research has been called cropping
systems. Under these conditions there may not be much difference between
farming systems and cropping systems research. In cropping systems
research, although the emphasis is on crops, all related farm activities
are also usually considered; from land preparation, which includes animals
and family labor, to marketing, which involves transportation and infra-
structural development. Consequently, cropping systems research is a
coordinated and integrated interdisciplinary effort to develop technology
that will enable farmers to increase food production in a way that is
acceptable to them. This simple condition "acceptable to them" must not
be overlooked. This socio-economic aspect has been overlooked many times
in the past and is one of the prime reasons why farmers did not or could
not accept the new technology.
In order to have effective research where limited funds and personnel
exist, much of the research in agriculture should contribute directly to
development needs and government production programs. The research should
Prepared for Symposium on Farming Systems Research. Washington, D.C.
December 8-9, 1980.
Cropping Systems Agronomist, Cooperative CRIA/IRRI Program in Indonesia
br (1) systematic and directed to specific target areas to provide the most
efficient use of the research funds and personnel, and (2) it probably should
be conducted on farmers' fields in order to approximate farmers' conditions
and understand the constraints farmers face. It should be an interdisciplinary
effort that is well integrated with related government agencies. If we can
describe the agroeconomic environments of the target areas well, we should be
able to transfer research technology from one area to similar areas and
save considerable time and research effort. Staffing and funding arrangements
associated with a comprehensive agricultural research effort such as in a
farming systems program are vital to the success of the program. These are
administrative matters that will vary from country to country. But some
guidelines may be given that are based upon experience from existing programs.
The objective of this presentation is to describe a cropping systems
research program and illustrate the role of socioeconomic as well as ecological
determinants in the description, design, and implementation phases of the
research and, ultimately, transfer technology. The approach described is
designed primarily to meet the needs of small farmers in developing countries
where there are limited funds and personnel for agricultural research and
development. Much of the information presented is a condensation from previous
papers describing cropping systems research in Indonesia. Consequently, this
paper will only briefly outline the essential points relevant to a farming
systems research program.
Cropping Systems Research
Cropping systems research is a coordinated and integrated effort to
develop technology that will enable farmers to increase food production.
The technology must be acceptable to the farmers who will use it. In some
cases it may be desirable to identify and remove constraints to the farmers
through government programs. The increased production may result from better
management of present cropping patterns, introduction of an extra crop(s) per
year (intensification), or expansion of crop production into newly opened or
under-used land areas (extensification).
The research is carried out by a coordinated group of scientists from
various disciplines. It is focused upon specific target areas to make more
efficient use of research staff and funds. The selection of target areas for
cropping systems research is very crucial. The Indonesian cropping systems
program emphasizes the following criteria for selection of target areas:
1. Critical areas in terms of food shortages and governmental
2. Large areas having similar soils and climate.
3. Feasibility of intensifying cropping patterns based on prior
4. Availability of markets and infrastructure.
A diagram for the cropping systems program in Indonesia (Fig. 1) shows
five distinct phases, associated research activities and approximate time
frame that follw after selection of the target area. The objectives of the
research within a target area should be specific and attainable within a
defined period of time. The research effort should be allowed to develop a
broader research base and perhaps become a farming systems research program
but only in a logical and stepwise manner. It is in this context that the
farming systems terminology is used in the remainder of this paper.
Staffing and Funding of Farming Systems Research
A cropping systems research program that evolves from the activities of
research scientists is likely to be more effective and long lasting than a
program that is conceived and implemented from the top. But unfortunately,
the evolutionary process may take a long time. A combined effort from the
administrators and from the research scientists would be most desirable.
This kind of approach has been effective in Southeast Asia for development of
programs for Genetic Evaluation and Utilization (GEU) for rice breeding and
Cropping Systems Research. These are interdisciplinary research efforts that
are complementary. The efforts of IRRI and international funding agencies
provided the access to the research administrators. The collaboration
of scientists through regional research networks for GEU and Cropping
Systems provided the stimulus at the research level. Fig. 2 illustrates
how these programs may function within a national food crops research
institute. Staff would be available from the usual divisions (or disciplines)
of the institute, hopefully on a voluntary and part time basis. Funding
through the programs would provide the incentive to work together.
Farming Systems Research, on the other hand, would involve scientists
from several research institutes. The research program would likely be
organized at a higher echelon within the ministry of agriculture than for
a cropping systems research program (Figs. 3 and 4). Again funding would
flow through the program to the research scientists as indicated in Fig. 5.
The existing research operations supported by routine funding would continue
to function within the various research institutes and directorate generals.
Special funds as seed money would probably be needed to encourage active
involvement in the programs. This is the place where some outside help may
be needed in terms of technical and financial assistance.
Organization of the Research Program
An effective farming systems program depends to a great extent upon the
Coordinator. An effective coordinator will likely move up quickly to higher
positions within the government. This dilemma can be overcome to a certain
extent by developing collective leadership through a Farming Systems Working
Group. Fig. 6 illustrates how a national cropping systems research program
was organized to direct its research activities and coordinate the activities
of regional research centers which have varying degrees of autonomy (Fig. 7).
A farming systems research program should be an interdisciplinary effort.
But it should also be integrated with other governmental activities. Rural
development programs and extension, local government and irrigation agencies
are examples of government programs and agencies that must ultimately be
involved in the implementation of the results of farming systems research.
Their involvement at early stages of the research programs is vital. Fig. 8
illustrates the level and time of involvement of different sectors of the
bureaucracy. This kind of integration facilitates the development of
relevant technology, implementation of the results and smooth transfer to
other areas with similar conditions (Fig. 9). Fig. 10 illustrates how
government interventation in terms of production programs is helpful to
farmers. Many times farmers can do little without this support. We cannot
expect farmers to immediately adopt technology that was developed over a
period of 5-10 years by scientists under conditions where inputs were available
and where no individual risks were involved. Even though farming systems
research may develop more relevant technology than traditional commodity
and discipline oriented research, the technology will likely be more complex
and dependent upon a wider range of social and governmental group involvement.
It is the feeling of may people working in farming systems research
projects that the real payoff from this research is in manpower development.
A strong sense of accomplishment and enthusiasm develops among many of the
staff. This provides an ideal opportunity to identify those scientists who
are stimulated by real problems that farmers face and who can work in
interdisciplinary teams. Short term training with colleagues from other
countries at international centers such as IRRI provides a good background for
young researchers in new cropping systems programs. This kind of training
should be complemented by opportunities to meet periodically at regional
workshops and conferences. These kinds of training opportunities will
probably need support from out of country funding agencies. It would be
useful to have available a broader range of training programs than presently
exist. Advanced degree studies should also be provided. Usually there is
sufficient funding for this. It is more difficult to identify centers of
excellence for such advanced studies.
McIntosh, J. L. 1980. Cropping systems and soil classification for
agrotechnology development and transfer. Paper prepared for the
Agrotechnology Transfer Workhsop. Soils Research Institute, AARD,
Bogor, Indonesia and the Benchmark Soils Project, University of
Hawaii. Held at Bogor, Indonesia. July 7-12, 1980.
CROPPING SYSTEMS RESEARCH P
SEconomic data collection
STechnology Transfer IN
technology and institutions
Site selection and
Economic and biological
Design and test cropping
Technology Transfer OUT
Technology Transfer IN
'Retrieval and Studies
*Partition target area
*Redesign and test patterns_ .
S Production Problems
Fig. 1. Parallel biological and socio-economic activities. required for the five distinct research and
implementation phases of a cropping systems program with little inflow of technology
SELECTED TARGET AREAS
FRAMEWORK FOR CROPPING SYSTEMS RESEARCH PROGRAM
Fig. 2. Interaction among programs and traditional divisions in a Cropping Systems Program.
AGENCY FOR AGRICULTURAL RESEARCH AND DEVELOPMENT
(PROPOSED ORGANIZATIONAL STRUCTURE)
LOCATIONS OF REUARCH
INST6TUTIS WMAIEfl &Vb
F -1 r8EEARCN INS5TITUTES
Fig. 4. Organizational structure of the Agricultural Research Agency.,
MINISTRY OF AGRICULTURE
A*fNCY FOR ANOIVALTRAL OCKAfl
GOINCY -ofl AkNMltinL 11UCAT
1IKAIko ANO EhICNZION
Fig. 3. Organizational structure in Ministry of Agriculture.
Farming Systems Research Program
GF.L--Genetic Evaluation and Utilization refers to systematic varietal improvement studies for food crops
s'icii as: rice, corn, legumes, etc.
Fig. 5. Interaction among programs and traditional commodity and discipline-oriented research
activities in a Farming Systems Research Program.
R E S E A
NATIONAL CROPPING SYSTEMS PROGRAM
National Cropping Systems Working Group
Fig. 6. Organizational structure of a Cr