On-farm research : organized community adaptation and learning for efficient agricultural technology innovation

Material Information

On-farm research : organized community adaptation and learning for efficient agricultural technology innovation
Hildebrand, Peter E.
Peter E. Hildebrand

Record Information

Source Institution:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.
Resource Identifier:
212626591 ( oclc )

Full Text

/6. -70-
c, -151


Po+or F Wi dphranr

Professor, Food and Resource Economics Department
University of Florida
Gainesville, Florida 32611

Prepared for presentation at the
International Multiple Cropping Systems Conference
Jiangsu Academy of Agricultural Sciences
Cosponsored by the Chinese Academy of Agricultural Sciences
October 8-11, 1985, Nanjing, Peoples' Republic of China


Peter E. Hildebrand

Professor, Food and Resource Economics Department
University of Florida
Gainesville, Florida 32611

Thirty years ago rural sociologists in the United States talked about

innovators, early adopters, late adopters, and non-adopters of improved

agricultural technology (Bohlen and Beal, 1957; Rogers, 1962). Innovators

were the ones who adapted technology to community conditions and they,

along with early adopters, tended to be the primary beneficiaries of this

technology. Late adopters benefited very little while non-adopters usually

were affected negatively and were considered as laggards who failed to

recognize the benefits of modernizing by adopting improved technology.

Researchers and extension agents studied the characteristics of the inno-

vators and early adopters, and tended to translate these characteristics

into model users of the products and services of agricultural research and

extension institutions. This philosophy has directed technology

development more and more toward innovators and early adopters. At the

same time, many who were late adopters and non-adopters have left farming

and been absorbed into industrial and urban life.

The environment for agricultural development in the United States in

the 1950s and 1960s created the "pro-innovation bias" (Rogers, 1983). This

bias was based on the thought that new technology Is good and should be

adopted. The fact that new technology might not be adoptable by all

farmers was seldom considered. That even the "better" farmers might need

to adapt the new technologies to their specific conditions was given some

consideration. But emphasis was placed on developing "Improved" technology

and then selling or diffusing the technology to "enlightened",

"successful" or "good" farmers.

The purpose of this paper is to argue that the methodologies of

agricultural technology development and dissemination through on-farm

research have now advanced to the point that they can be blended into a

highly efficient process that serves all farming systems in a community

simultaneously. These methologies can improve the social distribution of

the benefits from public investment in agricultural research and extension

and at the same time improve the efficiency of these activities.


Community adaptation, the adaptation or adjustment of a new technology

to the agro-socloeconomic conditions of a community, is what Rogers (1983)

calls reinventionn". In the past, adaptation of technology depended upon

those farmers with better resources and management skills. This process

resulted in technology better adapted to the resource and management

conditions found on the farms of those innovators and early adopters. Late

adopters, those farmers who had management and other resources somewhat

different from the innovators and early adopters, were forced into

additional adaptation of the technology to make It suitable for their less

favorable situations. Non-adopters were comprised of those whose

management and other resources were sufficiently different that the

technology could not be adapted to meet their conditions. Many high

yielding varieties (HYVs) of the Green Revolution of the 1960s and 1970s,

which produce superior yields on good soils with abundant moisture, have

been disappointing, or even produce less than traditional varieties on

infertile or arid soils. Modification of farm conditions to meet the

requirements of the HYVs could not be achieved on many farms and hence

those farmers became the non-adopters and did not benefit from this new



At the same time that community adaptation of technology is In

progress, the community is also becoming familiar with or learning about

the new technology. This learning takes two forms. One is the hands-on

or experiential learning of the Innovators or others using the technology.

S'" cc#--nd Tc eheorrv+trin !orntnn +hr->iinh lrfvrmnp+'nr. nainrpd h\. >(>

servatlon and other forms of study by those who are not involved actively

in the use of the technology (Wake, 1984). A learning curve, as

Illustrated In Figure 1, relates achievements toward reaching potential

results (such as potential yield from a new technology) with numbers of

attempts at using, or experience with, the technology.

During the process of community adaptation, adjustments in the

technology (such as choosing a subset of components, modifying levels of

Inputs or making the technology more nearly conform to community

traditions) have the effect of facilitating learning. Facilitated

learning, making a new technology more familiar, simpler, or easier to

learn to use, shifts the learning curve to the left, Fig. 1. The ultimate

potential of the technology so adapted may be lower than for the "maximum

yield" or full technological package as shown in the figure. The opposite

can also be true, but probably In fewer cases.

OTHER Potentil as adopted

Leorning curve
for adopted /

Learning curve for
scientific package
4ShWft from

0 2 3 4 5 6

Fig. 1. Shift in learning curve from
community adaptation.

Movement along the learning curve occurs with experience, but

comparable movement can occur through learning by observation. If a farmer

begins to use the technology after having learned about it first through

observation, it is equivalent to a shift of the learning curve to the left,

Fig. 2. If the potential of the technology Is unchanged, similar results

are achieved in a shorter period of time. The learning curve on the right

in Figure 2 could represent an earlier adopter while the curve on the left

could be that of a later adopter who learned through observation, the

equivalent of what the early adopter learned with one attempt at using the



The Farming Systems Research and Extension (FSR/E) approach to tech-

nology development (see, for example: Gilbert et al., 1980; Shaner et al.,

RESULT -Pat .L.s -

/ / Learnin curve for
/h later adopters
/Shift from/
I /
'/ ,Lnq curw for
I I eaorlie adopters
(adopted tecnnoloqy)


0 2 3 4 5 6

Fig. 2. Shift in learning curve from
observation before using the
+w;-hnr)CI <\nv

1982; Hildebrand and Waugh, 1983) Is a means of formalizing community

learning and adaptation. To a community, FSR/E brings the additional re-

sources of outside scientific knowledge and expertise. By combining the

efforts of scientists from several disciplines with those of the farmers in

the community, adaptation to local conditions is accelerated and dissemi-

nation Is more rapid.

In the FSR/E approach, much of the research required for technology

adaptation is conducted on farms, under real farm conditions, and with

farmer collaboration. In the process, some farmers are helping adapt the

technology while they obtain experience with It, Fig. 1. Others, those not

directly Involved in on-farm testing, have the opportunity to learn by

observation, Fig. 2.


In farming systems literature, recommendation domains (Byerlee et

at., 1982) are comprised of farms and/or farmers with roughly homogeneous

farming systems (Hildebrand, 1982). This is similar to what Rogers (1983),

for dissemination purposes, calls homophiless". Reconnaissance surveys or

sondeos are conducted to help define and delimit such recommendation do-

mains. Refining or partitioning recommendation domains, as data from on-

farm research is gathered and analyzed, has also been discussed (Hildebrand

and Poey, 1985).

It is more meaningful and more useful to consider the concept of a

research domain which is comprised of one or more recommendation domains.

A research domain, similar to what Byerlee et al. (1980) call a

"homoqeneous target region", is an ex-ante designation of a rouahlv

homogeneous agro-climatic zone throughout which it could be expected that a

set of potential Interventions (technologies) could have applicability.

For most technologies "the same experimental program may be implemented

over the whole region" (Byerlee et al., 1980, p 61). In many cases all, or

nearly all farms in the target region area will pertain to the research

domain. An example is a new disease resistant variety of a crop which

would be expected to demonstrate resistance throughout the area. An

exception would be with a new implement for large tractors in an area where

some farmers have large tractors and other farmers have no tractors. In

this case the research domain for the implement would include only those

farmers in the area with large tractors.

In a research domain, alternative new technologies evolve into

treatments to be Included in experiments and trials for station research

and/or on-farm experimentation. If research conducted in the research

domain involves several locations and Is designed to take advange of

modified stability analysis (Hildebrand, 1984), then the biological

research can result In the definition within the research domain of two or

more recommendation domains based on biological response to the treatments,

Fig 3. Other agro-socloeconomic research conducted simultaneously with the

biological research (for example, directed agro-socioeconomic surveys, soil

surveys, or farm enterprise budgets) provides Information needed to

characterize the farms in each recommendation domain, so they can be

identified for further research and/or dissemination. These data are also

valuable for second generation research Into cause-effect relationships

which will help to explain results of first generation research (Swanberg).




e= Environment

Fig. 3. Use of Modified Stability. Analysis
to partition Research Domain into


To the extent that a community falls within a research domain, or that

a research domain can be considered as a community, FSR/E becomes an

organized and structured community learning and adaptation system for

agricultural technology innovation. As an organized system and using the

methodology described above, FSR/E is highly efficient in enhancing

technology innovation in agriculture. First, because modified stability

analysis benefits from the utilization of a wide range of environments,

farmers who were formerly thought of as "innovators", "early adopters",

"late adopters" and "non-adopters" can, and should all be included in on-

farm research. Improved regression estimates of the response of

technologies to environments In modified stability analysis results from

including a wide range of farmers. This can improve the efficiency of

technology innovation for the superior environments (environments of the

innovators) while at the same time providing adapted technology for late

adopters and non-adopters in poorer environments. Hence, community

-, -.....- ,- +=kting place :Imultaneourtly for all strata of the community.

In contrast to results of adaptation only by innovators, learning curve

shifts resulting from adaptation in an FSR/E approach will benefit farmers

in both favorable and unfavorable environments.

Both experiential learning and learning from observation are also

distributed more widely In the community from an FSR/E approach than from

spontaneous community adaptation of centrally developed technology. When

adaptive research or experimentation is being conducted by a community

innovator, poorer farmers In the community may well be reluctant to obtain

Information from the innovator. By conducting on-farm trials over a wide

range of environments in a community, FSR/E facilitates the process of

obtaining information and of receiving hands-on experience with a

technology. In turn, the social distribution of benefits is more

equitable. Particularly relevant to the efficiency Inherent in FSR/E is

that shifts to the left in the experiential learning curve from adaptation

(Fig. 1) and from learning by observation (Fig. 2) can be cumulative.

While a farmer Is gaining Information about a technology, the technology

can also be in the process of adaptation to community conditions. In

summary, FSR/E complements, and makes more efficient, a naturally occurring

technology Innovation process in agriculture community learning and



Bohlen, J.M. and Beal, G.M. 1957. The diffusion process. Special Report
No. 18. Agricultural Extension Service, Iowa State College. Ames.

Byerlee, D.; M. Collinson et al. 1980. Planning technologies appropriate
to farmers concepts and procedures. CIMMYT. Mexico.

Byerlee, D., L. Harrlngton and D.L. Winkleman. 1982. Farming systems
research: Issues In research strategy and technology design. American
Journal of Agricultural Economics, Vol. 64, No. 5, pp 897-904.

Gilbert, E.H.; D.W. Norman and F.E. Winch. 1980. Farming systems
research: a critical appraisal. Michigan State University Rural
Development Papers, No. 6. East Lansing.

Hildebrand, P.E. 1982. Combining disciplines in rapid appraisal: the
sondeo approach. Agrlc. Adm. 8, pp 423-432.

Hildebrand, P.E. and R.K. Waugh. 1983. Farming systems research and
development. Farming Systems Support Project Newsletter, Vol. 1., No.
1. Gainesville.

Hildebrand, P.E. 1984. Modified stabilly analysis of farmer managed, on-
farm trials. Aqron. J. 76:272-274.

Hildebrand, P.E. and F. Poey. 1985. On-farm agronomic trails In farming
systems research. Lynne Riener Publishers, Inc., Boulder, Colorado.

Rogers, Everett M. 1962. Diffusion of innovations. The Free Press. New

Rogers, Everett M. 1983. Diffusion of Innovations. The Free Press. New

Shaner, W.W., P.F. Phlllpp and W.R. Schmehl. 1982. Farming systems
research and development: guidelines for developing countries.
Westvlew Press, Boulder, Colorado.

Swanberg, K.G. 1984. An economic justification for the farming systems
approach. AID Economist Conference, Annapolis, Maryland (Draft for

Wake, J.L. 1984. The cost of learning by doing effect on technology
adoption. Unpublished M.S. Thesis. Univ. of Florida. Gainesville.