A Discussion of Issues Relevant
to the Development and Implementation
of a Farming Systems Research Program
DAVID D. ROHRBACH
October 3, 1980
dc ~5-c- fte ks /- 3
Introduction................... .............................. 1
I. Issues Relating to the Definition of the Copcept of Farming
Systems Research........................................... 6
1) The Parameters of Farming Systems Research; How Holistic
Should FSR Be? ... ........ ........ ........... .... 6
2) Farming Systems Research 'In The Large' Versus Farming
Systems Research 'In The Small' ........................ 7
3) Distinction Between FSR and Farm Management Studies..... 9
4) The Role of FSR in National Development Planning and
Policy Making; The Assumption of a Fixed Versus Variable
Policy and Infrastructural Environment.................. 9
5) Role of FSR in the Agricultural Research System: Upstream
(Developmental) Versus Downstream (Applied) Approaches., 11
6) Target Population Group Size; How Location Specific
Should FSR Be? .......................'.... ........ .... 12
II. Issues Relating to the Organization of Interdisciplinary
Farming Systems Research.................................. 14
1) Problems Relating to the Integration of the Social
Scientist Into the Agricultural Research Process........ 14
2) Issues of Downstream Staff Maintenance: Appropriate
Levels of Training, Length of Commitment, Rewards....... 15
3) Issues of Organizational Structuring: Program Size,
Funding, Relationship of FSR to Other Research Station
4) Linkages Between FSR Programs and Other National
Agricultural Development Support Institutions........... 17
5) Character of Coordination Between International, Regional,
and National Research Programs......................... 18
III. Issues Relating to Farming Systems Research Methodology.... 19
1) The Identification of Target Area Farmers and Their
Stratification Into Homogeneous Recommendation
Domains .................. ........... ............... 19
2) Identification of Objectives and Character of Data
Collection; Appropriate Types of Field Surveys........ 20
3) Character, Value and Degree of the Ex Ante Screening
of New Technologies................................... 22
4) The Character of On-Farm Testing..................... 23
5) The Evaluation of Research and Testing Results........ 24
6) The Diffusion of Research Findings: Research or
Extension Agent Responsibility......................... 25
7) The Necessity of a Dynamic and Iterative Research
Process.......... .................................. 26
IV. Issues Relating to the Evaluation of the Applicability of
Farming Systems Research to U.S. Small Farms Research...... 27
1) Issues of Domestic Small Farm Development and Small
Farm Technology Needs................................. 27
2) Historical Use of the On-Farm, Systems Research
Approach Reexamined.................................. 29
3) Issue of Expansion Versus Reorientation of Existing
Research and Extension Efforts....................... 30
4) Problems Relating to the Integration of FSR Into the
Domestic Research System ......................... 31
The following discussion of issues relating to farming
systems research (FSR), attempts to describe a series of questions
prominent in the dialogue among FSR practitioners, and between the
practitioners and people interested in understanding FSR. This is
not an outline defining the "state of the art." Nor is it a technical
monograph. Instead, the paper simply attempts to outline a number of
fundamental problems relevant to the development and implementation
of an FSR program.
Much of the information has been gathered from a brief
survey of literature pertaining to the concept of farming systems
research. No presumption is made to its definitiveness. The problems
and issues described, however, provide a useful introduction to an
important series of questions relevant to people interested and
involved in FSR programs. The essential purpose of the paper is to
explain the issues and to describe the variety of arguments being
made about them. Tentative answers are provided to some of the
issues while questions are merely raised concerning others.
To the FSR practitioner, most of these issues are not likely
to be perceived as controversial. Some have even claimed that there
are no major conceptual issues related to farming systems research.
This is due in part to the fact that the term FSR has been used so
generally, but is also due to a degree of agreement over the basic
value and character of this type of research. The leyel of general
agreement present just underneath the surface of debate implies that
those issues which remain most significant to practitioners cannot
be resolved in the process of debate per se. These issues, which
primarily relate to questions of methodology, organization and
implementation, are in the process of being resolved by experience.
It is important to note that there probably can be no
definitive answer to many, if not most, of these issues. Their
resolution greatly depends upon the character and mandate of the
research institute attempting to implement an FSR program. Additionally,
it depends upon the environmental (political, economic, socio-cultural,
and biological) circumstances in which the institute operates.
Accordingly, a number of FSR practitioners have repeatedly stressed
there can be no single FSR methodology for all research institutions.
This paper has been divided into categories of conceptual,
organizational, and methodological issues as well as those specifically
relating to the potential applicability of FSR to domestic small farms
research. Issues relating to any one of the categories, however, cannot
be distinguished strictly from those identified as belonging to others.
The issues and problems are interdependent among themselves and to a
series of wider issues relating to the appropriate character of LDC
and small farmer development.
The ideas and judgments expressed in this paper are largely
abstracted from documents describing FSR which have been identified
in the bibliography. In addition, some clarification of the issues
has been gained directly from a few of the practitioners themselves.
I am particularly indebted to Peter Hildebrand, Elon Gilbert, Donald
Plucknett and Larry Harrington for their comments on an earlier draft
of this paper. I alone, however, should be held responsible for any
misinterpretations or mistakes.
Also, one note of caution should perhaps be addressed to the
readers of this document. As originally written, the paper was
designed to help identify agenda items for a USDA/OICD and AID/DS/AGR
sponsored workshop on farming systems research. The following discussion
of issues related to FSR program development and implementation has
been specifically addressed to those possessing some familiarity
with the subject. Readers seeking an introduction to the subject
may find the issues of some interest. They are advised, however, to
examine some of the diverse perspectives found in different writings
about FSR. The definitions and ideas discussed hereafter mark an
attempt to summarize and draw generalizations about a philosophy and
methodology still evolving through practical application.
Farming Systems Research is a philosophy and methodology
of agricultural research for the development of improved technologies
appropriate to small farmer needs and circumstances. 1/ FSR originally
developed in response to the recognition of the fact that limited
resource farmers in developing countries were not adopting many
improved technologies believed relevant to their needs. The blame
for this situation could in part be attributed to the inefficiency of
national extension services. Evidence also suggested, however, that
many of the disseminated technologies were simply not suitable to
farmer's circumstances. Resource constraints limited the applicability
of some technologies. Others, upon investigation, appeared inappropriate
to environmental conditions or farmer's goals. To overcome these
problems, agricultural researchers sought to explicitly integrate the
consideration of real farm circumstances and the dynamics of farmer
decision making into the research process. This provides the basis
for the FSR methodology.
Three principal elements distinguish FSR from traditional
agricultural research. First, an explicit attempt is made to under-
stand the farm, farmer, and farm environment in a holistic manner--that
is, as a complex system of interdependent parts. This involves the
recognition that the attempt to change any one of a system's components
must be evaluated in terms of its effect on the other parts of the
system. New technologies must be coordinated with the constraints
and capacities contained in the entire range of farming system operations.
Secondly, FSR initiates the research process with an attempt
to determine research priorities on the basis of an analysis of the
characteristics of representative, target group farming systems.
Generally this involves a preliminary descriptive survey designed
to define the farming system and its environment. On the basis of
this information, researchers identify farming system constraints
and capacities which they attempt to overcome or exploit through the
development and adaptation of technological improvements. In other
words, FSR is specifically designed to solve farming system problems.
Thirdly, the entire process of FSR, including the analysts
of the farming system, the technology development and testing, and
the verification of test results, is carried out by interdisciplinary
1/ FSR need not be small farm specific. The ideas and methodologies
are similarly applicable to agricultural research aimed at
developing technologies for large farms. Technologies developed
for small farms may be usefully employed by larger farmers. The
methodology has principally developed, however, in response to the
needs of small farmers in developing countries.
teams of social and biological scientists. 2/ Each contributes a
unique and necessary perspective to the understanding of systemic
constraints and capacities. Each supplies a part of the answer to
the question of why farmers do or do not adopt certain technologies.
FSR does not obviate the need for traditional commodity based
or reductionist agricultural research. The new approach simply attempts
to improve the efficiency of such programs by promoting a closer
coordination between the technology development process and the farmer
or farm system to be developed. The systems approach should thus be
recognized as complementary to traditional disciplinary research.
The practical nature of FSR and its relationship to ongoing
research programs can best be seen through a review of a number of
successfully implemented FSR programs. In fact, the very attempts
to establish a clear definition for FSR have depended, to a large
degree, on the identification of the valuable attributes of the ongoing
farmer oriented, systems based, agricultural research programs. While
such a review lies outside the scope of this paper, certain generalizations
can be made about the variety of different FSR approaches practiced
at a few of the international, regional and national research centers.
FSR programs can generally be said to fall into one of two
categories: 'upstream' or 'location neutral,' developmental research
and 'downstream,' 'location specific,' applied research. The distinction
between these categories relates to the purpose and methodology adopted
by the FSR unit. Upstream FSR programs aim to identify prototype
solutions to major regional farming system constraints. These programs
concentrate on developmental or basic agricultural research most of
which takes place on the research station, The identification of
upstream research. priorities, and the organization of the testing process,
however, depends upon an understanding of actual farm system circumstances.
Upstream researchers must gauge the effects of the tested technology
on the system as a whole. Therefore upstream programs are to a great
degree dependent upon information gained from on-farm and/or downstream
FSR generally classified as downstream aims to develop or
adapt agricultural technologies to improve location specific, target
group production in the short run. The entire research process is
carried out on farmer's fields with the direct participation of the
farmer. In most cases, potentially productive technologies developed
in upstream or commodity research programs are identified for testing
2/ Interdisciplinary implies the combined effort of representatives
from different disciplines. FSR team members are assumed to have
competency in their own disciplines. This is in contrast to
multidisciplinary which implies an individual's competency in
a number of different disciplines.
under farmer's conditions. The ultimate success of the research process
is determined by acceptability of the improved technology by the farmers.
Having noted the difference between these two types of
programs, it is perhaps important to point out that the distinction
may not be so clear-cut in practice. Accordingly, some practitioners
suggest that upstream and downstream FSR can best be understood as
representing two ends of a continuum for systems research. The
character of the FSR approach a research center chooses to establish
must depend, to a large degree, on the character of resources available
to it and the goals sought. In this context, international research
centers with regional if not global mandates for developmental research
and greater research resources should probably concentrate on upstream
efforts. By contrast, most national agricultural research centers
should emphasize downstream programs. In the case of LDC's, location
specific, applied research techniques are probably most commensurate
with the limited resources and more problem specific scope of the
Insofar as this paper concentrates on examining a series
of issues related to national program design and implementation,
attention will primarily be focused on the suitable character of
downstream efforts. Unless otherwise specified, future references
to FSR will refer to downstream programs.
Four components have been commonly recognized as necessary
stages of a downstream methodology. These are generally known as the
descriptive, design, testing, and vertfication/extension activities.
The success of each of these activities depends upon the coordinated
contribution of the research team, the farmer, and the extension service.
1. The descriptive stage involves the examination -
of the nature of a series of representative
farming systems in or-erto gain an under-
standing of systemic characteristics and
constraints. The interdisciplinary team
of FSR scientists can complete this responsi-
bility in a number of different manners
depending upon the time and resources available
to them. Most often the analysis of farming
systems targeted for assistance first requires
the review of any existing secondary information,
including baseline resource and climatological
data, and then involves the execution of
formal or informal farm surveys. Information
must additionally be collected regarding the
quality of available support services and
2. The design stage involves the evaluation of
the specific technology needs of the farming
systems, and the identification of technologies
which might be developed or adapted to increase
farm system productivity. These technologies
must also be judged in terms of their conformity
with societal needs and goals. Field tests
are then designed to evaluate the appropriate-
ness of these technologies to farmer's needs
3. The testing stage encompasses the actual trials
on farmer's fields of the chosen technologies.
These trials commonly come under various
degrees of researcher and farmer management.
By the final phase of the testing process the
farmer generally supplies all the inputs and
holds full responsibility for test management.
4. The verification/extension stage involves a
final evaluation of the acceptability of the
technology to the farmer, and the provision
of information about the technology to the
extension service for general dissemination.
The distinction between these stages need not and probably
should not be clear-cut. A further analysis of the farming system
often takes place during the testing process, and the kinds and
character of trials may often change. FSR practitioners have commonly
drawn attention to the iterativeness of the research process. This
appears to be one of its most important attributes. By definition,
the change of one element of a farming system effects and may change
others. New constraints or capacities may be created. Resultingly,
the development of a system can be a continuing process.
FSR practitioners suggest the changes that can be expected
from this approach to agricultural research are likely to be piecemeal
and accumulative. FSR does not seek to change the entire farming
system at once. Nor does FSR seek optimal solutions for any particular
farm's location specific circumstances. Some degree of optimization
must be sacrificed in the attempt to adapt technologies to the needs
of groups of similar small farmers. Yet evidence shows that even small
or non-optimal technological changes can have major consequences in
terms of farm productivity and farmer welfare.
Some practitioners have additionally noted the potential
implications of FSR findings for national agricultural policy determination.
Certain major farming system constraints may be found to result from
inadequate or nonexistent agricultural support services or problematic
agricultural pricing policies. Information gained through the analysis
of farming systems could possibly be usefully employed by policy
makers as a basis for their decision making. Non-technological
developments fostering agricultural productivity might then represent
an additional FSR benefit.
I. Issues Relating to the Definition of the Concept of Farming Systems
1) The Parameters of Farming Systems Research; How Holistic Should
In the broadest sense, FSR involves the consideration of the farm
system as a production and consumption unit composed of crop, livestock,
and off-farm subsystems. The contribution of each subsystem or component
thereof to the whole farm pattern or process must be gauged in the
attempt to understand the system and design improved technologies in it.
In practice, however, most FSR programs have concentrated on one or
another of these subsystems. New technologies have been developed to
overcome subsystem specific constraints or exploit subsystem specific
capacities. As a result, questions have arisen concerning whether
crop, cropping or livestock systems research can be called FSR, and
more significantly, how holistic should the FSR process be.
Several common responses are made to these questions. First,
as ordinarily used, FSR is a generic concept encompassing crop research,
cropping susbsytem research, livestock subsystem research and whole
farm research. The important element interrelating these various
research orientations is the fact that while the research may be
designed to develop a particular type of technology, it is based upon
a holistic understanding of the farm system itself. Thus the attempt
to improve a cropping pattern depends in part upon the understanding
of the relationship between such a pattern and the complex system of
other farm processes, e.g., its effects on the livestock system, the
farmer's goals and preferences, the character of the farm's entire
configuration of resource limitations, etc.
FSR does not seek to change an entire farming system. The
research approach simply aims to develop and adapt improved technologies
to a farming system in a more efficient manner. Such efficiency depends
upon some degree of recognition of the relationship between proposed
technologies and the functioning of the farming system as a whole.
There are arguments both for and against the consideration of
a variety of different subsystems or systems components in a single
research domain. FSR practitioners note that the complexity and
expense of the research process increases the more holistic the view
and goals for technology development taken. These people suggest that
a truly whole farm research orientation is simply beyond the means of
the agricultural research centers to implement. CIAT's original FSR
program, functioning between 1973 and 1975, is said to have failed in
part because the scope of Its systems research orientation was too wide.
The advantage of a wider research perspective, however, lies in the
greater possibility to exploit systemic interdependencies. Practitioners
argue that without the attempt to develop technologies relevant to various
subsystems, the interrelationships between these subsystems will not
be fully understood.
The limited extent of many research center mandates discourages
the adoption of a whole farm approach. A number of other centers,
however, have recently made a greater effort to gear their technology
development process to more than one subsystem. This seems particularly
true with national research centers with wider mandate responsibilities.
Experience gained through FSR programs has stimulated this sort of
2) Farming Systems Research 'In The Large' Versus Farming Systems
Research 'In The Small.'
In addition to determining the extent of the farming system
or subsystem to be developed, the FSR team must determine which specific
systemic components will be defined as management variables (potentially
changeable) and which defined as environmental parameters (not changeable).
An FSR program dealing with a small number of variables as opposed to
parameters (e.g., one with a crop specific research orientation) can be
called FSR 'in the small.' A program with a large number of variables
in relation to parameters (e.g., one with a multiple cropping system
or cropping-livestock system research orientation} can be called FSR
'tn the large.' There are advantages and disadvantages to each.
A research center's adoption of FSR 'in the large' or FSR
'in the small' largely depends on the extent of its mandate, expertise,
and available resources. 'In practice, both types of programs identify
limited numbers of constraints for resolution at any one time. Both
types of programs attempt to understand the character of these constraints
in relation to their wider systemic interrelationships. The distinction
between them primarily relates to the range of potential variables as
opposed to parameters from which the problem proposed for technological
resolution is chosen. For example, the mandate of CIMMYT, practicing
FSR 'in the small,' demands a research emphasis on maize and wheat and
their directly assoctatiye problem areas. The mandate of CIAT,
practicing FSR 'in the large,' demands research to promote food production
At issue is the scope of focus of the FSR efforts. FSR
'in the small' focuses from the initial stages of the research
process, on predetermined enterprises deemed of major significance to
the farming system. The types of problems identified and technologies
developed are largely determined by the enterprises chosen for development.
By contrast, FSR 'in the large,' without such a limited initial focus,
is likely to recognize a wider variety of systemic problems and develop
or adapt a greater variety of technologies. A research team practicing
FSR 'in the large' might then have a better change to exploit systemic
complementartties and interdependencies.
This adoption of one or the other of these approaches should
not simply depend upon the extent of a research center's mandate. Most
LDC research centers must maintain responsibility to a wide variety
of different kinds of farmers with a wide variety of different types
of production problems. These farmers might best be served by a FSR
team with no deliberate preconceptions concerning problems to be
addressed. The advantage of a smaller research domain, however, is
that the limited resources of a national research station on FSR team
may not be stretched too thinly. This could a)so allow a research
station to develop its expertise in dealing with these particular
enterprises using most farming system resources.
An extension of this debate relates to the issue of the
degree and speed of change FSR seeks to promote. FSR practitioners
have often noted that the farmer is more likely to adopt an incremental
series of limited technological improvements than a larger single
system transformation. Certain valuable systemic improvements, however,
may not be achieved in an incrementaltst fashion. For example, in
cases where the improvement of an existing component enterprise may
not prove feasible, small farm welfare may best be served by the
introduction of a wholly new crop or crop mix. If the research
team aims to promote the rapid development of the small farm system
and major increases in national agricultural production, larger changes
in small farm agricultural production practices may prove necessary.
The practice of FSR 'in the large' probably best supports at least the
consideration of such major transformations.
This is not to say that major productivity advancements are
not possible on the basis of small changes in production technologies.
A great deal of evidence suggests that minor changes can and do often
have substantial impact. Most technological improvements fostered
by FSR will likely fall in this category. Some may not.
An associative issue relates to the need to promote outside
support for the FSR program in the form of both funding and administrative
assistance. FSR practitioners have expressed the fear that if expecta-
tions are too high or if the changes fostered by the FSR approach are
too small, such support may not be gained. They have noted the particular
need in the initial stages of FSR implementation for rapid, clearly
identifiable, cost justifying results.
Yet these problems can also be seen to characterize traditional
research and development approaches. The basic question is whether
FSR should merely seek to improve existing farming practices, or should
major changes be sought in the types of commodities being produced?
This must depend upon the character of identified constraints and
capacities of existing farming systems. The expectations placed on
FSR, however, must be realistic. The research approach must be judged
in terms of its ability to foster increased research efficiency.
3) Distinction Between FSR and Farm Management Studies.
Some farm management researchers claim FSR to be no different
from the farm systems development approach they have been practicing
for years. They question the need for a 'new' research approach.
FSR practitioners have responded with the claim that FSR involves a
distinguishing concentration on technology development as opposed to
the reorganization of existing management practices and existing patterns
of resource use. While farm management researchers tend to assume the
value of the system they examine, FSR scientists aim to seek solutions
to specifically identified problems.
Some farm management scientists have suggested that they also
have been involved to some degree with technology development. FSR
practitioners stress, however, that the goals and structure of the
FSR work process are different. Teams of interdisciplinary scientists
jointly perform holistic farm system analyses seeking problems for
technological resolution, These efforts are explicitly linked with
ongoing technology development research. The fulfillment of the objectives
of the FSR process depends upon farmer acceptance of improved technologies,
4) The Role of FSR in National Development Planning and Policy
Making; The Assumption of a Fixed Versus Variable Policy and
This issue entails two related problems. First, what relationship
should exist between the objectives of the FSR process and the national
agricultural development process as a whole? Some practitioners envision
FSR to represent an interface between national development priorities
and local level development needs and opportunities. FSR has primarily
evolved to orient the technology development process in response to
local level farming system constraints. Research target areas are
often identified by the central government as regions of development
priority. Likewise, technology research priorities must conform with
national needs. Additionally, the successful dissemination of improved
technologies often depends upon the efficient operation of national
agricultural support services. The need to coordinate national and
local level agricultural development efforts thus appears extremely
To some degree this coordination can be attained if the FSR
team and development support bodies simply understand each other's
objectives and responsibilities. In certain instances, however, the
establishment of a distinct coordinating entity may prove useful. This,
for example, could include representatives from agricultural policy
making research, extension, credit supply and marketing groups, Infor-
mation gained from farming system analyses could then be systematically
shared, and the efficiency of efforts to develop and disseminate technologies
could be improved. The problem, however, is that the effort necessary
to play this role may be extremely difficult, time consuming, and costly,
drawing valuable resources away from the research process itself.
Secondly, the issue involves the related question of whether
the FSR team should adopt an interventionist or submissive approach
to support service development. The interventionist approach calls
for the explicit promotion of institutional support services reform
in order to help promote FSR improved technology adoption. The sub-
missive approach claims this to be beyond FSR team capabilities, and
that, therefore, policy and institutional factors exogenous to the
farming system should be regarded in the constraint identification
process as fixed. Innovations should not depend on reforms in support
service operations which may or may not take place. If exogenous
systemic factors are counted as variables, a greater degree of systemic
change may be possible. An example of this would be a case where an
FSR team identified the infertility of soils to be a major cause for low
crop productivity. This could be remedied by the application of a
certain type of fertilizer. Yet the region's fertilizer distribution
system is notoriously poor. The scientists perceive themselves faced
with the options of attempting to test and introduce the fertilizer
and pressing for the development of the distribution system or seeking
higher yields without the fertilizer. The interventionist team would
seek support service reform. The submissive team would seek alternative
routes to higher yields. The difficulty and complexity of attempts
to change exogenous factors represents a strong argument for regarding
them as fixed.
Most FSR practitioners have suggested the value of a compromise
between the interventionist and submissive approach. In this case,
FSR teams maintain some responsibility,for identifying policy or infra-
structural constraints and communicating them to the appropriate
administrative representatives. Such claims can be substantiated by
written program reports describing FSR field activities and summarizing
survey data. Most FSR activities, however, concentrate on developing
and adopting farm technologies of probable acceptability without change.
This issue is extended in the question of how potential changes
in the agricultural support system or those actually being promoted
should be regarded. An example would be a case where a development
project is about to be implemented aiming at the reform of a support
service. Should the success of such a reform attempt be assumed? If
the hoped-for changes in the support system do not take place,
innovations dependent on the change may no longer be viable. TechnQlogical
changes which are not reliant on reforms in the exogenous development
support system will likely still be beneficial. The degree and
character of development sought by policy makers, however, may thereby
5) Role of FSR in the Agricultural Research System: Upstream
(Developmental) Versus Downstream (Applied) Approaches.
As noted in the introduction to this paper, the distinction
between upstream and downstream FSR is not absolutely clear. Upstream
programs primarily aim to develop prototype solutions for major regional
constraints. Downstream programs seek to develop, and more importantly
adapt, agricultural technologies to improve specific target group
production and overcome specifically identified farming system
constraints. Upstream research primarily takes place on research stations.
Downstream research primarily takes place on farmers' fields.
While both types of FSR depend upon farmer input, farmers are more
extensively involved in downstream programs. Accordingly, the
information and experience gained from downstream programs should play
a major role in orienting upstream FSR. Similarly, downstream programs
should seek to adopt technologies developed in upstream efforts (as
well as technologies developed in reductionist or commodity research
programs) to target group circumstances.
tn practice, most FSR programs can best be viewed as lying
somewhere on a continuum between the upstream and downstream ideals.
The distinction, however, remains important in the determination of
the type of program a developing country should implement.
Some practitioners have suggested the value of a division of
labor between international research centers concentrating on an upstream
approach and national research centers which should concentrate on
downstream programs. :Proponents of this viewpoint claim that the
greater expertise and funding levels of the international centers
are more supportive of basic developmental systems research than the low
resource levels of national centers. These, by contrast, are in a
better position to develop location specific downstream programs,
adapting the findings of international centers to local circumstances.
In addition, the broader mandates of the international programs imply
greater opportunity costs for location specific work. Under a coordinated
division of labor, national centers can provide international centers with
information concerning major local level constraints and systemic
characteristics tn exchange for information about basic technological
There are also arguments, however, for not maintaining this
separation of responsibilities. Researchers involved in upstream
programs may only gain an understanding of and respect for the complex
nature of actual farming system constraints if they gain some degree of
direct experience with downstream efforts, This problem particularly
relates to the difficulty of understanding the complex trade-offs
involved in the small farmer's decision making process. Similarly, a
national downstream program might have need for basic component technology
research unmet by either international upstream programs or commodity
The decision on what type of FSR program to implement at a
national research center must depend on the actual character of national
goals and resources. Factors to consider include the quantity and
quality of available research personnel, the degree of commitment
placed in the FSR effort, the size and character of the technology
base available for applied on-farm research, and the types of constraints
to be resolved. Most practitioners believe that in the initial
stages of FSR program establishment, the downstream approach is probably
most valuable and viable.
6) Target Population Group Size; How Location Specific Should FSR Be?
This issue underlies the larger problem of cost effectiveness.
Downstream research results are specifically geared to the needs of
particular target groups of farmers. This ensures that developed tech-
nologies are appropriate to the actual nature of farming system circum-
stances, and thereby helps ensure they will be adopted. The limited
number of farmers reached, however, entails two problems. First,
as the research process becomes more location specific, performance
costs rise. The on-farm testing component of the FSR process, practi-
tioners note, is the most expensive part of the process. The greater
the costs, the greater must be the improvements to small farm production
and welfare. Secondly, the value of the location specific improvements
must outweigh the opportunity costs of not reaching a greater number
of other small farmers.
FSR practitioners commonly suggest two responses to these
problems. First, they note that FSR does not seek optimal solutions
to each farmer's location specific problems, Instead, research efforts
aim to develop improved technologies for farming systems or parts of
farming systems containing simila- characteristics across reasonably
similar regions. The basic task of FSR is to develop technologies which
promote large enough improvements in the productivity to be adopted by
large enough groups of farmers. Practitioners point out, however, that
developed technologies which may be superior over a broad area are
often inferior to farmers'location specific traditional technologies
developed from generations of practice. This experience provides a
basic justification for the location specificity of FSR.
Secondly, practitioners assert that the true value of FSR
must be judged by comparing its ability to generate appropriate technologies
for farmers with that of traditional research approaches alone. From
this viewpoint, the opportunity costs of effectively assisting location
specific farmers are less than those attributable to programs developing
more widely 'adoptable' yet inappropriate technologies.
Little cost-benefit analysis relating to FSR has been done
to date. While the benefits of FSR .aye.been commonly noted in.terms
of rapid technology development and diffusion, little quantitative data
exists. The associatiye costs of FSR have not been carefully examined.
Some question the feasibility of such analysis. They question whether
FSR can be adequately judged In strict quantitative terms. Such quanti-
tative documentation, however, may at least prove useful in the determina-
tion of program design.
II. Issues Relating to the Organization.of Interdisciplinary Farming
Systems Res.earch .... ... ..
1) Problems Relating to the Integration of the Social Scientist
Into the Agricultural Research Process.
The establishment of a cooperative working relationship between
social and biological scientists, particularly when social scientists
are newly being integrated into an agricultural research institution,
is difficult and complex. Several major problems can offset the attempt
to achieve a synthesis of knowledge, understanding, and practice.
First, a proper mix between the number and quality of social and biological
scientists must be determined for the FSR team. If biological scientists
dominate, research may be oriented toward particular biological constraints
and socioeconomic and socio-cultural considerations may not be given
full enough attention. The opposite may occur if social scientists
dominate the research team. In addition, tt has been suggested that the
emphasis placed on various components of the research process can shift
depending on the orientation of the team. For example, a social science
orientation may tend to place greater emphasis on description and
technology design based upon existing technological knowledge. This
team might seek a rapid passage through the testing stage to the diffusion
process. A biological orientation could place less emphasis on the
description and design stages and greater emphasis on testing and
evaluation processes. Distinct orientations can also appear in the
evaluation stage itself--the agronomist being more concerned with yields
and the social scientist more concerned with income or welfare levels.
Secondly, the development of a cooperative process of working
interaction between various types of scientists can be difficult unless
each works to understand each other's terminology and disciplinary
perspective. This may or may not require some degree of specialized
training. The problem, however, underlines the value of a specialized
team leader, perhaps with some degree of multidisciplinary experience
and an understanding of different disciplinary perspectives. The team
leader apparently must have the responsibility for and ability to
bring the diverse disciplinary perspectives into a common focus. The
potential usefulness of a farm management scientist in this role has
Thirdly, in order for team members to seriously evaluate each
other's opinions and understanding, some practitioners stress that the
academic training of various disciplinary representatives must be
congruent. FSR team members with greater understanding and experience
may not seriously consider the ideas of those with lesser training. Social
scientists, however, have commonly had a difficult time proving their
value to technical scientists. Therefore, the initial introduction of
a social scientist into the agricultural research process may, in fact,
require a particularly well trained and articulate disciplinary
representative with strong knowledge of the concerns of technical
scientists. It cannot be understressed, however,.that a serious
commitment to the implementation of an FSR program depends upon an
initial recognition of the-importance of socio-cultural and economic
2) Issues of Downstream Staff Maintenance: Appropriate Levels of
Training, Length of Commitment, Rewards.
Staff development and maintenance problems are commonly cited
by FSR practitioners as, among the most difficult program implementors
face. Developing country research institutions commonly possess few
highly trained scientists and limited monetary resources. Therefore,
field terms will likely be composed of researchers with Bachelor's
degrees or post high school technical training, A brief period of
instruction tn the concepts and methodology of FSR can be of significant
value. Of greater importance, however, is a bastc competence in the
practical application of disciplinary knowledge.
The quality of field personnel should be judged in terms of
the needs of farmers. The farmer must have some respect for the researcher's
technical abilities. Likewise, the disciplinary representatives must
maintain a sincere willingness to work with and learn from the farmer.
In addition, field team scientists must. be evaluated for
their ability to work with research station scientists. The adequacy
and effectiveness of on-farm research in part depends upon strong
linkages with on-statton, development research programs. The establishment
of FSR program credibility wtll be dtfftcult unless station scientists
have an understanding and respect for field team efforts. If the
relationship between the two groups simply appears as an interaction
between junior and senior scientists an effective linkage may be difficult
This problem directly relates to the issue of rewards and
opportunities for professional advancement of field team scientists.
On-farm research must, not be simply identified as a training program
for junior scientists, Yet if there is a dichotomy between experienced
scientists practicing developmental research and less experienced
researchers performing adaptive field trials, this may be difficult
to avoid. If field scientists are to be rewarded by promotions to
research station work the cooperation between team members could be
threatened. In this case each researcher might work to prove his
-----~-.. ~-~lr .rrrr.-~~r-u-- ^.-rarr+.-C.
The problems of professional advancement threaten both junior
and senior FSR scientists. Practitioners have commonly noted that while
the prestige and possibilities for advancement are known and straight-
forward for scientists involved in basic research, such is not true for
scientists working in an interdisciplinary manner at the farm level.
Willingness to work on downstream teams could significantly be compromised
by the lack of adequate opportunities for professional development.
The rewards for downstream tnyolvement need to be explicitly identified,
and a system of reward distribution must be established. This is easier
said than accomplished, however. Present opportunities for peer
contact, professional review and publication are minimal. Current
professional disciplinary distinctions therefore cannot but threaten
the character of interdtsctplinary cooperation.
3) Issues of Organizational Structuring: Program Size,Funding,
Relationship of FSR to Other Research Station Activities.
A number of developing countries have expressed an interest
in the idea of FSR. The question remains whether they will back that
interest with the necessary degree of financial and administrative
support to implementt something more than a short term development
project. Such backing must come from both the governments and the re-
search institutions to be Involved,
In relation to this question, some practitioners have noted
their fear that high expectations or the lack of adequate understanding
concerning the prospects and process of FSR can significantly compromise
the value of the program. Therefore, initial attempts to implement
FSR should probably be small in scale. The minimal viable size of an
FSR program is open to question, as is the most effective staff size
and level of staff expertise. One practitioner has suggested that field
teams may be as small as two individuals. The ideal staff size must
largely depend on the goals of the program and area to be covered.
The linkages between an FSR program and other research
station programs are extremely important to the success of an FSR effort.
Research stations hold baseline data and information about improved
technologies crucial to the systems analysis and testing process. In
addition, data gathered by the FSR team can be usefully employed to
help determine station research priorities. Some FSR practitioners
argue that the establishment of a distinct organizational unit is necessary
to maintain the interdisciplinary integrity of the FSR program,
Accordingly, a separate budget should probably be established to
reinforce this commitment. Yet such. a mode of organization .could
threaten the possibility of a cooperative pattern of linkages with the
other research programs. This threat might be avoided if the FSR unit
initially evolved as an ad hoc group of representatives from the ongoing
There is also some question about whether the funds for an
FSR budget can be diverted from.existing.research center programs and/or
whether new sources of funding must be found. While practitToners have
noted that the lack of additional monetary support could hinder the
development of an FSR program, no evidence exists describing the size
of commitments needed. In.general, the.costs of off-station research,
particularly when this involves a great deal of travelling between
different research sites, have'been said to be higher than those.for
traditional on-station research. FSR programs are likely to require
new investments in vehicles for transport to widely dispersed farms, and
a larger variety of research personnel, 'et presumably, as a greater
degree of a research center's effort is directed off the experiment
station, money used for on-station programs and station maintenance
can be diverted to service off-station activities. Specific quantitative
data needs to be assembled.
4) Linkages Between FSR Programs and Other National Agricultural
Development Support Institutions,
The establishment of adequate institutional relationships
between various agricultural development support agencies can be
crucial to the success of an FSR program, The effective operation of
these agencies can determine the options available for technological
improvement and the value of the developed technologies. Some FSR
practitioners have suggested, therefore, that these relationships
be explicitly structured with channels of information exchange and
administrative overview. It remains questionable, however, whether the
responsibility for promoting the coordination of the agricultural development
effort should be taken up by the research institute. The problems with
such an attempt have been identified above,
One linkage, however, which merits closer examination due to
its distinct importance is that with the extension service. A close
working relationship between an FSR team and extension service representatives
is generally said to be of major value to on-farm research efforts,
Extension agents should participate in all stages of the FSR work process.
Agents possess useful information for farming system analysis and test
design. Their participation in the testing and verification stages can
significantly improve the ultimate dissemlnatton process.
Some people interested in FSR question the appropriate
extent of responsibility of FSR scientists over the extension and
diffusion of their research findings. Insofar as an element of the
FSR evaluation process involves the analysis of the acceptability of
the team's technological Innovations, there seems a tendency for
downstream scientists to involve themselves in the diffusion process.
This may in part be attributable to the fact that the successful
adoption of new technologies determines the success of the entire FSR
effort. The costs of a researcher's extensive involvement in the
dissemination process are likely to be high,
~.~.- .-'.--"Yi.m. .-1 ..ii-CI C111-- .--- I- ~~-~. -~-I
Practitioners point out, however, that verification trials
under total farmer management do not represent a usurpation of extension
responsibilities. While these tests may be perceived by nonparticipating
farmers as demonstrative trials, they are in fact important means for
the final assessment of innovations. Once such evaluations are completed,
the extension service holds full responsibility for the dissemination
of research findings.
5) Character of Coordination Between International, Regional, and
National Research Programs.
Three basic elements underlie the issue of coordination
between various levels of research institutions. The first is the
question of an appropriate division of labor. The values and problems
associated with this idea have been discussed earlier. Suffice it to
say that some degree of division of responsibility already exists
and is valuable. The actual character of this division must depend
upon the degree to which international centers serve national needs.
It seems clear that national centers must maintain some responsibility
for the development of basic technologies commensurate with the demands
of local circumstances,
Secondly, FSR practitioners have suggested that an important
vein of interdependence lies in the area of information flow between the
various sorts of institutions. The success of both international and
national FSR programs can be enhanced by the establishment of explicit
channels of FSR related information flow--of information concerning
developed systems technologies downward from the international research
centers and information about farm system constraints upward from the
national research centers. It has also been suggested that these linkages
be made on a multilateral rather than a bilateral basis. This implies
that formal channels of communication be established between international
and national research centers. Such flows of information should not be
dependent on informal contacts between scientists at various research
Thirdly, international centers are already taking up the
responsibility for training national FSR scientists. The length and
character of these training programs should perhaps be evaluated to
maximize effectiveness. There has been some suggestion that methodologies
of FSR used and taught at the international level are not relevant
to local level resource constraints and needs. The methodologies of
international centers may be more complex, or rely upon a greater degree
of field staff expertise than the national centers are able to provide.
III. Issues Relating to Farmihg SysteimsResearch Methodology
1) The Identification of Target Area Farmers and Their Stratification
Into Homogeneous Recommendation Domains.
The identification of a target area for agricultural development
is principally a political decision. Theltarget area delimits the
region in which an SR team conducts its search for technological Improyer
ments. This may commonly be an area in which-ltttle research has been
previously done. The identification of target area characteristics through
the examination of previously collected baseline data, secondary source
reytew and on-farm surveys then supplies the basis for the determination
of likely recommendation domains.. These are made up of groups of farmers
with roughly similar circumstances and problems. There may be one or
more recommendation domains in a target area. Generally, however, a
single FSR team will work to improve production in a single recommendation
domain at any one time. Presumably, an innovation approved by a repre-
sentatiye group of farmers within the domain will be appropriate to
most other farmers in that domain.
The determination of the actual character and extent of
recommendation domains can be a difficult problem. In order for
FSR to be cost effective, the size of a group of farmers capable of
adopting an improved technology must be reasonably large. As noted
earlier, however, the domains cannot be so large that the technological
improvements offered are less than significant for each individual farmer,
or will not be rapidly accepted.
Recommendation domains can be determined on the basis of a
wide variety of criteria. They are in part dependent on the amount of
base data information gathered and the types of farming system variables
deemed most important. The extent of domains can also depend on the
general types of innovations sought. Resource and climatological
baseline data previously collected by a research center can provide
an initial basis for the identification of boundaries. Information
gathered from on-farm surveys and secondary source interviews can then
be evaluated in terms of government, scientist and farmer identified
priorities and development objectives. Research priorities established
on the basis of this information finally determine the range of farmers
for which i'nnoyations might be appropriate, The point is that decisions
made determining the amount of information gathered and the needs to
be addressed can significantly affect the type and number of farmers
to benefit from the research process.
One additional clarification perhaps needs to be stressed.
While recommendations ofdomains generally include similar sorts of farming
systems, different systems may have similar characteristics in relation
to any particular innovation. Ultimately, the group of farmers to
whom an improved technology is recommended may differ for each technology,
Therefore, the.identification of a 'similar'k group of farming systems.
on which an FSR.team might concentrate its efforts should be always
open to reexamination and reform. The FSR team must not simply confine
its, efforts to those farmers:Jnitially identified as similar..
2) Edenttftcation of Objectives and Character of Data Collection;
Appropriate Types of Field Surveys ,
Before going into the field.FSR;teams should acquaint
themselves with whatever relevant recorded information already exists
about the target area, Research stationstoften have some amount of
climatological and resource survey data relating to the team's
region of interest. The uses of :uch information can significantly
inform the field research process,'and;provide important links between
the FS.R scientists and the researchers who.collect this data. The
coordinated review of such data:canoalso serve to initiate the field
team to the demands of a cooperative effort.
The primary purpose of base-data.Surveys in the field is to
acquaint the FSR team with the real nature of small farm circumstances.
FSR practitioners often point out that the small farmer has developed
his farming system through generations,of experience in order to make
optimal use of his limited.resources. The base data farm survey can
acquaint the researcher with the.character of the environment in which
the farmer operates as well as the'farmer's understanding of that
environment, The survey can be used to.determine how and why a farmer
operates as he does. The understanding of the complex interrelationships
of small farm enterprises and the character of farmer decision making
i. necessary to determine if and.how a farmer can be assisted to improve
his productivity. The effects of numoerous exogenous influences inr dluding
market character and quality, price levels, input availability, extension
quality etc. must also be taken into consideration. The evaluation of
collected survey information then provides the research team with a
basis for choosing and evaluating potentially appropriate innovations.
Presumably, the greater degree of understanding the researcher
has of farming system circumstances, the better he is able to develop or
adapt agricultural technologies to farmers!.needs. FSR practitioners
commonly note, however, that data collectors working to identify
systemic constraints and design on-farm research must guard against
gathering too much information. These scientists have found a strong
tendency to collect more information than is necessary for research
design and more information than can be effectively digested in a reasonably
short period of time. This- needless information has been found to include
both the measurement of useless variables and the unnecessarily precise
measurement of useful variables. .Clearly, a proper trade-off must be
established between the speed and amount of data collected. A method
must be identified for gathering a necessary and sufficient amount of
information at the least possible cost,
The determination of the nature-and extent of a necessary
and sufficient amount of information i :extremely difficult;. 'SR
practitioners have found that each member of a team desires to collect
excessive amounts of information pertinent.to his own discipline.
Therefore, researchers note the value of es-tablishing strict guidelines
defining necessary information:before the:team approaches the field.
Such guidelines, however, must not be so strict as to prevent the
recognition of important unexpected -ariables.
In addition to determine ng a proper amount of information
needed, practitioners must identify an appropriate method for its collection.
Researchers commonly argue about the yalue of various sorts of survey
methods. Rapid, informal surveys, are:said.to allow a team of researchers
to gather a great amount of relevant: information cost effectively.
Scientists are given the opportunity to.directly acquaint themselves
with a large number of small farmers~and to continually redirect their
questions in response to what they observe and to articulated farmers'
concerns. Scientists using informal surveys may be less inclined to
Orient their questioning in terms of preconceived notions about farming
system conditions and constraints.
Questionnaires, by contrast, have the advantage of producing
at least minimal amounts of quantitative data useful for the justification
of research priorities and test designs, as well as the substanti'atton
of claims for policy or itnfrastructural reform. Also, greater amounts
of more random, representative data can be collected by this means,
Questionnaires, however, can reinforce preconceived notions about farming
Some researchers, stress the value of simple participant
observation techniques over th.elength of a cropping season. A great
amount of information can hereby be obtained about farmers' needs and
practices which would most likely remain unobtainable through short and/or
formal surveys. Placed in this' situation, the researcher might
be more inclined to recognize the value of the farmer's traditional
technologies. A valuable 'give and:take' rapport can be established
between the researcher and a group of farmers in a region. This process,
however, might be said to be needlessly costly, particularly under the
assumption that such information and rapport can be gained in th.e technology
In-depth case studies provide valuable information about the
wide variety of interdependent .relationships underlying critical systemic
variables over time. Random sample surveys; display the representativeness
of yarious systemic characteristics. The. only conclusion it.seems
possible to draw. is that every type of survey has its advantages and
The resolution of the problems of data collection must finally
be worked out by each FSR team. Several key factors can be identified,
however as iMportant determinants effecting the choice between survey
methods. These include; a) the amount of information initially
available to team members about th.e target area; b) the team's degree
of experience with. the farmers in the region; c) the amount of a team's
experience with FS'R; d) the amount of resources available to a team
to collect and analyze its information; el the circumstances under
which th.e team operates--cultural factors, language barriers, etc.; and
f) the time the team allots itself to accomplish the task of data
The evaluation of the trade-offs involved in initial attempts
to define and understand a farming system is slightly easier with the
recognition that farming system analysis continues throughout the
research process, The research team may find it useful to employ
several survey ethods of different stages of the process. Additionally,
the results of initial surveys need not all be evaluated before.the
testing stage begins, FSR involves an iterative sequence of events
in terms of both attempts to learn about farming systems and attempts to
promote their development.
One final point perhaps should also be made with regard to
the issue of data collection, Some practitioners strongly caution
against an oyerconcern for quantification of base data information.
They claim FSR should not serve as a means to gain benchmark studies.
Instead, FSR scientists should concentrate on understanding the farmer
and his circumstances. The felt need to gather quantifiable data can
distract the researcher from this purpose, Other practitioners note,
however, that quantified data could supply valuable credibility to
the research process and could be necessary for the professional develop-
ment of the scientists involved.
3., Character, Value and Degree of the Ex Ante Screening of New
Ex ante screening of technologies involves the attempt to
determine what technologies are appropriate.to test on the basis of
initial data survey information. Two distinct issues are entailed in
this process. First, the value of a great amount of reliance on
ex ante screening has been questioned. The initial screening process
itself is necessary as a starting point for technology test design,
Farmer's needs and circumstances are herein being evaluated in
relation to societal objectives. Adaptable technologies can then
be chosen from the variety of known production tools. At issue, however,
is the degree of time and effort which should be entailed in the screening
process. The belief underlying the question is that much more can be
learned in the technology testing process about farm system circumstances
than in extended initial attempts to analyze that system. The
iteratiyeness of the FSR development process insures the continued
reeyaluatton of test designs on the basts of increasing knowledge
about farming circumstances. Therefore, the initial data collection
and analysis need not be too complex.
Secondly, a question exists concerning the relative value
of adaptive as opposed to developmental on-farm technology testing.
The adaption of known agricultural technologies has been said to
result in more rapid and significant farm system improvements.
The suggestion has even been proposed to assess improvements already
developed or adopted by the most innovative farmers for their wider
acceptability. There will always be certain system constraints,
however, which can only be resolved by means of basic developmental
technology testing. tt may be profitable to perform this developmental
work in the field. Yet researchers must be careful not to threaten
small farmer production or welfare. The final answer to this question
largely depends on the extent of the availability of adaptable technologies.
41 The Character of On-Farm Testing,
Several types of testing methods are used in most FSR programs.
These can include researcher managed on-farm tests, farmer managed on-
farm tests, and verification trials etc. A variety of different
factors must be taken into account in designing these trials ranging
from the determination of variables to be accounted for (including
non-experimental variables measuring effects on the farming system as
a whole), to the identification of appropriate test sites and plot
size, The evaluation of the advantages of various testing techniques
lies outside the scope of this paper. Such an evaluation must be
coordinated with an assessment of the influence of country specific
goals and circumstances. Two major issues, however, can be highlighted.
The first concerns the degree of farmer participation in the
research process. Such participation i;s crucial. The farmer must be the
ultimate judge of the results of research trials and thereby the value
of the improved technology. The greater his involvement in the testing
process, the more likely test results will be appropriate to his
circumstances. Researchers must recognize, however, that farmer
involvement entails a degree of risk to.the integrity of the trials.
There is always the chance that the farmer may mistakenly compromise the
test process, for example,.by harvesting a crop before it can be
measured. In addition, researchers ,must recognize that farmer involvement
entails a degree of risk to the farmer himself. A failed on-farm trial
can reduce a farmer's basic food supplies or level of subsistence.
Technologies for on-farm testing must be carefully screened, and
compensation must be assured for losses.
Another issue whichtimerits careful consideration is whether
participating farmers should be chosen for their representatiyeness,
or for the fact that they are innovative. In most cases, farmers
who volunteer to participate in on-farm testing programs will most
likely be those who have been innovative in the past. Such farmers
are often innovative, however, because they are subject to different
sets of constraints than representative non-innovative farmers. Even
in the case when these farmers operate similar objective circumstances
to those of non-innovators, subjective differences may be significant.
As a result, improved technologies accepted by these farmers may be
perceived aye inappropriate by others, The value of using innovative
farmers, however, lies in the greater likelihood that they would understand
the nature of on-farm trials, and contribute useful information during
the test process and in their evaluation of the results,
The use of representative farmers would probably more effectively
simulate common farmer circumstances. Technologies so developed might
then be perceived as appropriate and valuable by greater numbers of
farmers. In certain cases, explicit attempts to work with nonstnnovators
could make them more receptive to innovations in the future.
5) The Evaluation of Research and Testing Results.
The principal issues relating to the process of evaluation
concern how and by what criteria the evaluation should be carried out.
FSR practitioners. note that the best ultimate measure of the yalue and
appropriateness of improved technologies is the farmer's adoption
rate. The measure can be compromised, however, by the lack of an effective
dissemination process'. The influence of an inefficient extension
service on the rate of technology diffusion can be difficult to
distinguish, from the 'influence of the quality of the technology itself..
Practitioners note that a number of other evaluative
criteria can also prove useful. Different bases of judgment are often
employed at different stages, in the testing process, The appropriate
character of such judgment criteria must be carefully determined to
conform with the particular character and:objectives of the testing
process. Practitioners must guard against the traditional tendencies
of agronomists to solely evaluate the success of their experiments
in terms of yields and input-output ratios. In many cases, these may
be less important measures than soctoeconomir factors such as income,
employment generation capacity or risk reduction.
Some FSR practitioners have also specifically noted the
importance of accounting for the long-term consequences of adopted
technologies. Such considerations may conflict with the need to
generate rapid, significant farm system developments. Yet the lack
of this type of evaluation can ultimately cause great ham) to the
farmer. The assessment of long-term consequences should include
consideration of the influence of technology adoption on environmental
quality, market circumstances and price levels, sbcio-cultural
traditions, and national development goals.
6) The Diffusion of Research Fi'ndings: Researcher or Extension Agent
As noted earlier, extension agents should participate in all
phases of the research process. They can contribute useful information
for farming system analysis and help foster the development of ties
between research teams and farmers. Involvement in the testing process
can also productively inform extension agents about the technologies
they will later be responsible to implement,
There.remains some question, however, about the degree of
ultimate reliance which should be placed upon a notoriously inefficient
extension service for the dissemination of research findings. To
what degree should the FSR team concern itself with. the character
and quality of the technology diffusion process?
On-farm trials, particularly in the later stages of the
testing process could be viewed as demonstration plots. This, however,
is not their major intention. If extension inefficiency does appear
to be a common problem, FSR researchers could take a more explicit
and active role in extension agent training. While this may appear to
be an expensive use of a scientist : time, the lack of effective
technology dissemination must be recognized as. a greater loss.
Minimal amounts of extension training is probably more productive and
less costly than the researcher's attempt to take responsibility for
the extension process himself.
7) The Necessity of a Dynamic and Iteratiye Research Process.
The.dynamic and iterative nature of the FSR process is an
extremely important characteristic. In essence, FSR is developmental.
Researchers are called upon to.constantly reevaluate thetr basic
knowledge about the farmer and farm system. Failed research tests
provide useful information for redesigntng:new tests, The adoption of
improved technologies creates farming systems with new characteristics
and often new constraints. An ongoing FSRprogram tn a single
recommendation domain involves a continual progression from testing
and the evaluation of test results to the redesign of on-farm trials
to the testing process again.
The effective operation of feedback loops and a continuous
process of learning are thus of great importance for FSR productivity.
This suggests that FSR teams should maintain a significant degree of
integrity over an extended period of time. Each change in team
membership can involve a loss of valuable knowledge and experience,
In addition, feedback channels between the field and the
research station supply an added dimension of dynamism, Such
linkages should probably be explicitly organized,
IV. Issues- Relating-to the -Evaluation of. the Applicability of Farming
Systems Research to U.S.:-Small'Farms'Research. ; :: ::"
1) Issues of Domestic Small Farm Development and Small Farr
Over the past few years there has been a renewed concern
for the needs of American small farmers.. This has arisen in part
as a result of a reaffirmed commitment tg.th4 value of the small
farm way of life, and a new concern for the:effective utilization
of small farm resources.. In addition, the concern arose out
of the perception that existing national and local level research
and extension programs were primarily oriented to serve the interests
of larger commercial farmers, and therefore were not adequately
meeting small farmer needs.
This interest in domestic small farm development has
largely been directed in two areas. First, there has been a
concern to determine how the national extension service can act
more effectively to assist small farmers. New channels of communica-
tion are being opened, and more deliberate assistance efforts are being
adopted. Secondly, there has been a concern to ensure the accountability
of agricultural research efforts to small farmer needs. This has
involved an attempt to evaluate the appropriateness of existing agri-
cultural technologies to small farmer circumstances, and to identify
new types of small farm specific research needed.
tWithtn this context, two major issues underlie the
evaluation of the applicability of FSR to domestic small farmer
needs. First, small farm researchers have recently noted that
the exact nature of the 'small farm problem' is unknown. The renewed
concern for small farmer welfare has uncovered a lack of knowledge
about the actual nature of small farm conditions and needs. More
information needs to be gathered to identify who the small farmer is,
before major development program funding priorities can be established.
Secondly, an examination of the applicability of FSR demands
a preliminary assessJent of the appropriateness of existing agricultural
technologies to small farmer needs, This must involve an evaluation of
why small farmers do or do not adopt known technologies, and an evaluation
of unmet small farm technology needs. Such an analysis lies outside
the scope of th.s paper. What follows simply represents a series of
judgments which provide a useful starting point for an evaluative
Most agricultural technologies presently developed for U.S.
small farmers: are classified as scale neutral. Clear evidence suggests,
however, that many limited resource small farmers have not adopted many
__*C___I _ I_---I---YI~CI-^-CI.C1-. r--~Cn-Ln~~-4-1(-rl-111--
of these apparently useful technological.developments. Low rates of
adoption are commonly perceived to.result from both a lack of knowledge
about these.technologies, and a simple lack of initiative or desire to
The first of these causes implies an inefficiency in the
dissemination process. This could in part result from the fact that
extension services are generally inadequately, funded and staffed to
meet the needs of every farmer and smaller farmers are often the
first to be neglected, The problem could.4lso result from the inadequate
training of extension personnel.
The second cause for low rates of adoption could be attributable
to a number of factors including the multiple goal orientation of many
small farmers, Yet evidence suggests that in many cases this results
from the fact that scale neutral technologies are simply inappropriate
to limited resource small farmer circumstances and needs. Land, labor,
and capital constraints do in fact limit the adaptability of certain
technologies commonly thoughtof as scale neutral. To many small
farms researchers or extension agents, examples of this are somewhat
common. Fertilizer applications, particularly in the strengths
recommended for large and medium scale commercial farmers, are simply
too expensive for small farmers. A seed variety which requires
Intensive application of water, fertilizer, and insecticides or maintains
a short optimal harvest season can be inappropriate. A highly toxic
insecticide often cannot be used by a small farmer without a license or
an appropriate applicator..
In addition, there is' evidence to suggest that small farmers
have many unmet technology needs which specifically relate to the size
of their enterprises. The most comoroly cited casess are small scale
machinery needs. Yet increasing numbers of additional size specific
needs are beginning to be recognized by people interested in small
farm development, These include needs for new plant varieties,
cropping techniques, weed control measures, etc.
A question, therefore, of obvious relevance is how can such
needs be met.' If agricultural research. institutions have not adequately
served small farmers, how can research efforts be reoriented? FSR
represents one approach. While the specific methodology which has been
developed for dealing with developing country needs and circumstances
may not be strictly applicable (and the point in itself is debatable),
some similar sort of on-farm, systems research method could be.
Apparently small farmer needs and circumstances need to be better
identified and understood. The yalue of the involvement oftechnolqgy
research scientists in this task needs to be examined. Some people
interested in domestic small farm development have.suggested the'
usefulness of interdisciplinary investigations.
A related question demands an assessment of the importance
of location specificity. Some small farm development specialists
note the fact that small farmers in different regions and small
farmers with different goals and circumstances have different needs.
Whether an FSR type of approach represents a cost-effective way to meet
these different needs should, perhaps, be evaluated. The least
that can be said is that FSR represents a potentially valuable option
for the improvement of domestic small farms research and extension.
2} Historical Use of the On-Farm, Systems Research Approach
Many U.S. agricultural researchers and extension represen-
tatives initially exposed to the idea of FSR have remarked that this
type of approach is similar to on-farm, systems research efforts aimed
to assist domestic small farmers in the early part of this century.
Some note that these sorts of endeavors only disappeared from usage
a few years ago, Two opposing reactions, however, correspond to these
First, there is the belief that good reason exists for the
termination of these sorts of efforts. These people claim that the
American farmer has become more knowledgeable and commercialized, and
thereby more independently response to the opportunity to maximize
the productivity of his farm's resources, Most small farmers, it
is argued, have at least potential access to information deemed necessary
to operate their farms efficiently. In relation to developing
countries' circumstances, input and output markets are generally
well developed, as are the research and extension services. In relation
to identified needs, an FSR type of approach currently appears too
costly to support. Such needs, according to this perspective, can
be adequately served by existing agricultural support services.
By contrast, some agricultural researchers and extension
agents have reaffirmed the need for this type of approach. These
people commonly note that the termination of on-farm, systems research
efforts has corresponded with the development of a preoccupation with
medium and large scale commercial farmers, and a corresponding loss
of concern for limited resource small farmer needs. This loss of
concern has been reflected in a number of ways. Proponents of this
perspective claim U.S. agricultural support systems, although well
developed, no longer adequately serve many small farmers, National
and state.economic policies have begun to push these agriculturalists
off the farm. In addition, changed g'als and objectives of many small
farmers Ce.g., the lack of a simple concern for the goal of profit
maximization7need to be specifically responded to in different ways
than those used to assist larger farmers.. The problems faced by many
limited resource farmers and circumstances under which they operate
are so different, it has been argued, from those of larger commercial
farmers, that they must be uniquely dealt wit., The application of
an FSR type of approach might therefore be justified,
3) Issue of Expansion Versus Reorientatton of Existing Research
and Extension Efforts,
Some small farms development specialists have suggested that
small farmer needs can adequately be met simply by means of the expansion
of existing research and extension efforts, This would involve three
elements of change. First, increased funding for extension personnel
and training could foster better assistance directed toward small
farmers. Paraprofessional programs currently or recently operative
in Missouri' and Texas exemplify this initiative, Extra money could
also be used to simply hire greater numbers of county extension agents,
and provide better ongoing training and education programs.
Secondly, more funds could be allocated to fund small farm
specific agricultural research. Initiatives have already been made
in this area also, although questions have been raised about the
appropriateness of funding priorities.
These questions underlie a third significant initiative
characterizing present and possible future expansion efforts, This
involves the attempt to promote better information flows between
extension agents knowledgeable about small farmer needs and researchers
working to develop technologies to meet those needs. Specifically,
information gathered from extension agents is being used to help
determine research priorities.
The question remains, however, whether greater value can be
achieved by getting the scientist out on the farm and involved in an
interdisciplinary systems; research effort. This would mean a clear
reorientation of the small farm development process. The scientist
might thereby learn much more about small farmer needs and circumstances.
The priorities; of the research process and the design of trials would
likely be more relevant to small farmer needs. The resulting technologies
might be very different from those developed on the experiment station.
They would undoubtedly be more rapidly adopted. The costs of this type
of effort, however, need to be carefully evaluated.
4) Problems Relating to the Integration of FSR Into the Domestic
The value of both a:multidis-ciplinary systems perspectiye
and on-farm research have been.recognized by many domestic agricultural
researchers and extension agents. Most-agricultural research, however,
has remained on the experiment station,:and within a disciplinary
context. Interdisciplinary interaction has largely come in the
form of informal links and communications,.. On-farm work has generally
been left as the distinct responsibility of the extension agent.
The existence of these circumstances highlights two major problems
which could inhibit the implementation of a domestic FSR program,
First, some degree of disjunction commonly exists between
many agricultural research and extension programs. Channels of communi-
cation between researchers- and extension agents could be improved.
While in some cases the two types of programs are closely tied, in
others relations are fraught with mutual distrust and jurisdictional
jealousy. Mutual respect'and cooperation, however, are important for
the stability of an F$R effort,
Secondly, the task of developing an efficient, formal process
of interdisciplinary interaction could be problematic for two basic
reasons. Domestic academic and professional success depends on disciplinary
excellence. Insofar as FSR. requires that responsibility for farming
system analysis and technology development be shared by a group of
scientists opportunities for disciplinary distinction are limited.
In addition, those same professional and disciplinary pressures cause
research to become-increasingly specialized. Communication across
disciplinary lines can therefore often be difficult. These problems
may be particularly severe when the research base is the academic institution
Bernsten, R.H., 1980. "Cropping Systems Research in a National Program:
Indonesia." Paper presented at the AAEA Annual Meeting, July 28-30,
Binswanger, Hans P., B.A. Krantz, and S.M. Virmani, 1976. The Role of
ICRISAT in Farming Systems Research. Hyderabad, India.
Byerlee, Derek, et al., 1979. "On Farm Research to Develop Technologies
Appropriate to Farmers; The Potential Role of Economists." Paper
given at the Conference of the International Association of
Agricultural Economists, Banff, Canada.
CATIE, 1978. A Farming Systems Research Approach for Small Farmers
of Central America, Turrialba, Costa Rica.
Chambers, Robert. 1980. "The Small Farmer is a Professional."
CIMMYT. 1979. Demonstrations of an Interdisciplinary Approach to
Planning Adaptive Research Programmes; Deriving Recommendation
Domains for Central Province, Zambia. Report No. 4. October 1979.
1979. Planning Technologies Appropriate to Farmers: Concepts
and Procedures. Mexico.
Collinson, Michael. 1979. "Agrarian Change; The Challenge for
Agricultural Economists; Micro-level Accomplishment and Challenges
for the Less Developed World." Paper given at the International
Association of Agricultural Economists 17th Conference. Sept. 3-12,
1980. "Some Notes on the Farmer as the Client
for Research." Paper given at the CIMMYT Workshop on Methodological
Issues Facing Social Scientists in On-Farm/Farming Systems
Research, April 1-3, Mexico.
Consortium for International Development. 1980. Guidelines to Assist
National Governments in the Implementation of Farming Systems
Research and Development Programs Aimed at Farmers with Limited
Resources. Draft, June 1980, Ch. 6-9.
Dillon, J.L.,D.L. Plucknett, and G. Vallaeys. 1978. Farming Systems
Research at the International Agricultural Research Centers.
Technical Advisory Committee, Consultative Group for International
Agricultural Research. Washington, D.C.
Goldman, Richard H. 1980. The Role of On-Farm Testing and Evaluation
in Agricultural Research and Policy Management. Development
Discussion Paper No. 87, HIID.
Harwood, Richard R. 1979. Small Farm Development, Understanding and
Improving Farming Systems in the Humid Tropics. Westview
Press, Boulder, Colorado.
Hildebrand, Peter E. 1980. Comments on Farming Systems Research
and Extension. July Draft.
S1977. "Generating Small Farm Technology: An
Integrated Multidisciplinary System." Paper given at the 12th
West Indian Agricultural Economics Conference of the Caribbean
Agro-economic Society, April 24-30, Antigua.
_ 1979. "Generating Technology for Traditional
Farmers." Paper given at the Symposium on Socio-Economic
Constraints to Crop Production, 1X International Congress of
Plant Protection. Aug. 5-11, Washington, D.C.
1979. The ICTA Farm Record Project with Small
Farmers--Four Years of Experience. Guatemala, mimeo.
1979. Incorporating the Social Sciences into
Agricultural Research: The Formation of a National Farm Systems
Research Institutes. Guatemala, mimeo.
1979. "Summary of the SONDEO Methodology Used
by ICTA." Paper given.at the Rapid Rural Appraisal Conference
held at IDS, December 1979, University of Sussex, Brighton.
IRRI. 1980. Guide to On Farm Cropping Systems Research. March Draft,
IRRI Cropping Systems Working Group, Los Banos, Philippines.
Kerr, Howard W. 1980. A Survey of Current and Expected Research Needs
of Small Farms in the Northeast Region. USDA/SEA, Washington, D.C.
Madden, J. Patrick, Heather Tischbein, and Jerry G. West. 1980. An
Agenda for Small Farms Research; A Report on Phase II of the NRC
Small Farms Project. Staff Paper No. 30, Agricultural Economics
and Rural Sociology Department, Pennsylvania State University,
University Park, Pennsylvania.
Menz, K.M. and H.C. Knipscheer. 1980. "Role of Agricultural Economics
in Farming Systems Research at IITA." Paper given at the
Symposium on Risk and Uncertainty in Decision Processes of
Small Farmers in Less Developed Countries, AAEA-WAEA Joint
Annual Meeting, July 31-Aug. 3, San Diego, California.
Navarro, Luis A. 1977. "Dealing with Risk and Uncertainty in Crop
Production, A Lesson from Small Farmers." Paper given at the
Symposium on Risk and Uncertainty in Decision Processes of Small
Farmers in Less Developed Countries, AAEA-WAEA Joint Annual
Meeting, July 30-Aug. 3, San Diego, California.
S_1980. "Some Issues in Farming Systems Research:
CATIE's Experience in the Central American Isthmus." Paper given
at the CIMMYT Workshop on Methodological Issues Facing Social
Scientists in On-Farm/Farming Systems Research, April 1-3, Mexico.
Norman, David W. 1980. "The
for the Small Farmer."
East Lansing, Michigan:
Farming Systems Approach: Relevancy
MSU Rural Development Paper No. 5.
Michigan State University.
1976. "Farming Systems Research in the Context
of Mali." Summary paper for Workshop on Farming Systems Research
in Mali, November 14-19, Bamako, Mali.
Norman, David W. 1980. "General Overview of Farming Systems Research."
Paper given at the CIMMYT Workshop on Methodological Issues
Facing Social Scientsits in on-Farm/Farming Systems Research,
April 1-3, Mexico.
1973. "Methodology and Problems of Farm Management
Investigations: Experiences From Northern Nigeria." MSU
African Rural Employment Paper No. 8, East Lansing, Michigan:
Michigan State Univeristy.
Norman, David W., and Henry M. Hays.
Technology for Small Farmers."
1979. "Developing a Suitable
National Development. April 1979,
Norman, David W., E.H. Gilbert, and Fred Winch. 1980. "Farming
Systems Research in the Third World: A Critical Appraisal."
Forthcoming as MSU Rural Development Paper. East Lansing,
Michigan: Michigan State University.
Scobie, G.M. 1979. "Investment
Some Economic Dimensions."
Oct. 1979, Washington, D.C.
in International Agricultural Research:
World Bank Staff Working Paper No. 361,
Thomas, John R. 1980. "The Need for Program Specific Methodologies
in Farming Systems Research." Master's Thesis, Univ. of Hawaii.
USAID, CATIE Evaluation Team. 1980. Guatemala Country Report.
Draft. Washington, D.C.: USAID.
____. 1980. Honduras Country Report. Washington, D.C.:
S1980. Nicaragua Country Report. Washington, D.C.:
USDA, Ad Hoc Committee on Small Farms of the Joint Council on Food and
Agricultural Sciences. 1979. Research, Extension and Higher
Education for Small Farms. Washington, D.C.: USDA.
,Community Services Administration and Action. 1978. Regional
Small Farms Conferences; National Summary. Washington, D.C.: USDA.
Science and Education Administration. 1979. The Science
and Education Administration's Research and Extension Programs
for Small Farms. Prepared for the Subcommittee on Agriculture,
Rural Development and Related Agencies, Committee on Appropriations,
U.S. House of Representatives, March 1, 1979. Washington, D.C.:
Virginia Polytechnic Institute and State University. 1980.
Cooperative Extension Small Farm Programs in the South: An
Inventory and Evaluation. Research Division Bulletin 153,
Zandstra, Hubert G. 1979. "Cropping Systems Research for the Asian
Rice Farmer." Agricultural Systems (4), 1979. pp. 135-153.