Title: Clarifying the learning curve
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00080617/00001
 Material Information
Title: Clarifying the learning curve
Physical Description: 7 leaves : ; 28 cm.
Language: English
Creator: Wake, John L
Publication Date: 198-?
Copyright Date: 1980
Subject: Agriculture -- Technology transfer -- Evaluation   ( lcsh )
Agricultural extension work -- Florida   ( lcsh )
Farmers -- Effect of technological innovations on -- Florida   ( lcsh )
Genre: non-fiction   ( marcgt )
Statement of Responsibility: John L. Wake.
General Note: Typescript.
General Note: Caption title.
 Record Information
Bibliographic ID: UF00080617
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 153293738

Full Text

TO: P. Hildebrand

FROM: John L. Wake

SUBJECT: Clarifying the Learning Curve

In the FSR/E meeting on June 7th Dr. French handed out for
discussion a May 25th memo from Dr. Andrew concerning questions
Dr. Woeste had about FSR/E North Florida. Question number nine on
page two reads "Can we say something about the importance of the
learning curve to extension agents and how can that learning
curve be better understood in all of its dimensions?" Since the
learning curve is my area of expertise I would like to address
Dr. Woeste's question. Much of this view of mine is a synthesis
which I will not bother to substantiate with references.
Hopefully this paper will give you some good ideas on ways to
answer Dr. Woeste's question.

Information, Experience and Learning

There are two ways a farmer "learns" how to use a new
technology efficiently. First, he can gain information before he
ever uses the innovation. This is information from farm seed
dealers, extension agents and the plethora of extension
information services, from talking to neighbors, from just
watching a neighbor try something new in his field, from farm
magazines, radio, television, newspapers and probably from every
other conceivable form of social communication. The farmer learns
a helluva lot just by gathering these sources of information.

Information is important. One of the explanations of
differential adoption of technologies by farmers is the cost
involved in gathering this important production information. The
information cost may be explicit such as, gas expense or magazine
price, or the information cost may be an opportunity cost such as
the lost time driving to a meeting or reading a farm publication.
The extension service greatly lessens the information costs; it
is free to everyone who asks (except opportunity cost), it's
unbiased, highly competent and readily available.

Information though, is not the only way a farmer learns how
to use a new technology, a farmer also learns a technology by
just using it. The hands on experience teaches a farmer how the
innovation must be adapted to his soil, his machinery, his labor
supply, his management and to all the resources that are unique to
him and his farm. The farmer first learns how to use the
technology and then he learns how to use the technology cheaper.
Farmers are constantly trying to improve their production
practices. All people are always trying to find ways to get more
for less. Once a farmer adopts a technology he may never truly
"perfect" it because each year the technology reacts differently
to different weather conditions and the farmer then learns some
more about the technology and adjusts his practices. The farmer's
resources change through time also and it takes time for the

Page 2

farmer to adjust his use of the technology to his changing land,
labor and capital situation. Prices change and the farmer must
learn the new least cost production practices.

As the farmer learns a technology it has ramifications on his
whole farming system forwards and backwards. He adapts his
rotation, his chemical use, his equipment use, his labor use and
he adjusts not only the new technology but he adjusts his
preceding crop, his following crop, his inputs and how he handles
his outputs. The farmer continuously integrates technologies into
his own dynamic production system. Agriculture is the
biologically based production system. The strength of the
interactions between the components is often incredible and the
mechanisms often mysterious. Farmers work with nature so
production is not straight forward like it is in the industrial

Agricultural production is too complex for farmers to
completely understand a new technology before they ever try it.
Farmers can reduce the amount they have to learn through
experience and hard knocks by gathering information about a new
innovation before they try it. For example; two identical farmers
have two identical farms. The only difference is that one farmer
has more information about the new technology both are going to
try this year for the first time. The farmer with less
information necessarily makes more errors which lower yield,
lower quality, cause him to use the wrong inputs; or to put it
more generally, his errors raise his unit cost of production.
This less informed farmer also necessarily learns from his
errors, corrects them and lowers his unit production costs with
experience. The twin farmer with more information has a lower
unit cost of production to begin with but he too makes errors,
learns, corrects errors and thus improves his production cost
with experience. The point is that there are two ways a farmer
learns a technology; through information and through experience.

The above example makes it easy to see why lack of
information can discourage adoption. The farmer with less
information had a higher "cost" of learning the technology i.e.
higher unit production cost. The higher cost and less informed
twin may become discouraged, plan not to use the new technology
anymore and badmouth the new technology, or at the very least he
will not expand his use of the new technology as fast as his more
informed and lower cost twin. Characteristically a farmer will
try a new technology as a small experiment first and then as he
learns the technology and its benefits become apparent, the
farmer will expand his use of the new technology. In the end, the
higher cost of learning of the less informed farmer causes him to
adopt less rapidly. In addition, the example shows how important
the information provided by the extension service is; it lowers
the cost of learning to the farmers who use the service, it
increases adoption, diffusion and ultimately the wellbeing of the
community. It is as if the extension service absorbed the high
learning costs of the farmers by providing production
information. The lowered cost of learning then increases

Page 3

technological diffusion and hopefully social wellbeing.


Now we've seen how information and experience affect how a
farmer learns a technology and his adoption but we still do not
know how the technology itself affects learning, adoption and
ultimately the diffusion of the technology. The technology is the
other actor in technology adoption. The farmer learns the
technology but the technology itself determines how it is learned
by the farmer. Some technologies may be extremely simple, say a
new disc plow that is exactly the same size and shape as the old
disc but the new disc lasts twice as long and costs the same as
the old disc. The amount of learning the farmer has to do in this
case is minimal, he has to get the information that the disc
exists but not much more. The new disc performs exactly the same
in the field so there is not much to learn with experience. The
farmer gains all the benefits from the technology almost

At the other end of the scale would be an exemely complex
technology like soybean production. The farmer int only has to
get enough information to decide to adopt soybeans but he also
must get information about the multitude of production practices
used in the technology. The farmer has to learn the optimum; land
preparation, fertilizer and practices, varieties, liming
practices, nematicide, seed treatment, planting date, planting
equipment, row spacing, planting rate, planting depth, planting
method, inoculant and practices, herbicide type, herbicide
equipment, herbicide effects on following crop, when to
cultivate, cultivating equipment, if to irrigate, when to
irrigate, insecticide type, insecticide application method,
insecticide application time, fungicide type, fungicide
application method, fungicide application time, combine time,
combine adjustments and marketing, to name some examples.

The information the farmer does not get before he starts
production has to be learned through experience. The strong
component interactions with soybeans make it just that much more
difficult to learn the best production practices. For one
example, if the farmer decides to plant late he should also
change the soybean variety, plant spacing and harvest date, as
well. As mentioned earlier yearly weather differences
complicate matters; herbicide practices and varieties which are
fine in some years are disastrous under different weather
conditions. The cruxt of all this is that the more complicated
the technology the more learning is involved and the more
learning involved the greater the number of errors involved as
farmers learn the technology.

The cost of learning, caused by errors made in production
practices, is higher the more complex the technology, no matter
how much information a farmer has to begin with. As technologies
(although profitable once they are learned) become more and more

Page 4

complex they tend to have slower and slower diffusion and their
adoption becomes more and more restricted to the larger farmers.
Adoption becomes restricted to the larger farmers because they
are the ones who can better afford the high costs involved in
learning these complex technologies. In the same way that the
less informed twin farmer had to pay a higher price in errors as
he learned the technology, a more complex technology causes all
farmers to pay a higher price in errors as they learn the
technology. With higher costs of learning the farmers need more
experience (time) with the technology before they can learn it
well enough to see its benefits and expand their use. This of
course, slows diffusion and the flow of social benefits to be
gained from the technology adoption. Therefore the more complex
an innovation the slower the diffusion for all farmers and the
more likely that smaller farmers are discouraged from adoption.

(In the above example of technologies with varying
complexities we have assumed that the technologies were all equal
in every respect except complexity. In reality the amount of
benefits derived from using a new technology have a profound
effect upon adoption and diffusion. For example, when soybeans
were at $12 per bushel a farmer could make a lot of mistakes his
first year and still benefit from adopting soybeans the very
first year. At today's price of $6 per bushel the same mistakes
in the first year would cause a tremendous cost and would put
into doubt the long run benefits from adopting soybeans. In other
words, the learning effect only explains the supply end of
adoption, demand is equally important in explaining adoption and
diffusion of technologies but demand is not considered here.)

A general theory of agricultural diffusion when learning is
present might work something like this. First of all, the public
undertakes agricultural research. The researchers spend years
investigating a technology while slowly learning how to improve
it. The government agricultural researchers slowly "perfect" the
technology which lowers the cost of learning to use the
technology to the farmers. Eventually (hopefully) the technology
is "perfected" to the point that larger farmers can profitably
adopt the technology. The larger farmers then in turn continue
"perfecting" the technology until the cost of learning becomes
low enough for smaller and smaller farmers to adopt. This pattern
of diffusion, larger farmers first, then medium and then smaller
farmers has been observed in some studies although the reasons I
give here for its occurance differ from the reasons usually

Nevertheless there may be a kink in this diffusion pattern
which may further disadvantage the smaller farmers. To the degree
that larger and smaller farmers have different resource bases,
the "perfecting" of the technology by the larger farmers will not
necessarily "perfect" the technology for the resource base of the
smaller farmers. The smaller farmers may not benefit from the
larger farmer learning when the two groups of farmers are very
much different. This can explain the differential adoption seen
in the diffusion process. It is true that researchers may

Page 5

sometimes develop a technology that is so profitable or so easy
to learn that the cost of learning is minimal and discourages no
one from adopting but usually research is a methodical step by
step improvement process. (We have assumed that all of these
technologies are scale-neutral. If a technology is not
scale-neutral then of course, smaller farmers are overtly
discouraged from adopting.)

This brings us to the situation as it is today. An
agricultural researcher investigates a new agricultural
technology, he finally gets the technology to a point to where he
thinks it would be profitable to farmers and then the researcher
starts to move his new technology off of the research station and
into the farmers' fields. The researcher is not a superman, he
does not know exactly how the new technology will react under the
various and sundry conditions farmers actually have. Therefore
the farmers end up making mistakes with the new technology and
the returns don't reach farmer or researcher expectations.
Smaller farmers using the new technology may not be able to hold
out but some farmers probably will stick with the technology
while slowly learning how to lower unit production costs and
integrating the technology into their entire agricultural
systems. These first adopters, presumably larger farmers, will
share their knowledge with the community, lowering the
community's cost of learning, increasing adoption and the
community supply curve. But each new adopter will also have to
learn how to use the technology in his own particular situation.

When a technology is moving off the research station the role
of information is less important. The extension service provides
information free of charge to all farmers regardless of size
which lessens differential adoption. But when the technology is
just moving off the research station there is just not enough
good information about the technology that the extension service
can use. The extension service may have all available information
but the agents know that benefits from the new technology may not
be assured and the agents therefore are cautious in their
recommendations. The extension service wants better information
about the technology but it is just not available. Eventually
production information accumulates but only slowly and with years
of experience.

Take the example of perennial peanuts in Florida. A few
years ago several plantings were made into farmers fields but
today few of those plantings survive. The researcher explains
that perennial peanuts need intense management to become
established, which is to say that perennial peanuts are a complex
technology at its current stage of development. The researcher
himself has no problems growing it but when it comes to general
farmer production there is a lot to be learned. There is great
demand for a forage legume in North Florida yet there isn't
enough good information available about how farmers can grow
perennial peanuts profitably for extension agents to really push

Page 6

When the extension service does recommend a complex unproven
technology they can hurt their credibility with farmers. For
example, spraying wheat for Septoria. In 1981-82 the acreage of
wheat in North Florida skyrocketed because of the new availability
of a new rust resistent wheat variety, Florida 301. It became
apparent in February/ March 1982 that there were large areas with
Septoria. It was almost a crisis atmosphere. Extension
recommended controlling Septoria the way researchers do. The
treatment consisted of; at least two fungicide sprayings at a
specific stage of growth and with a specific interval between
sprayings. The fungicide spraying technology was complex
especially when compared to its potential benefits. Not
surprisingly most farmers did not get the fungicide applied
correctly and nearly all the farmers that I have talked to who
tried the treatment said they would not do it again. The
extension service did the best they could considering the
circumstances. But more knowledge about controlling Septoria
would have been very beneficial to the farmers in lowering there
unit production costs.


Well, the solution to all this, as you might have guessed, is
a farming systems approach to research. The farming systems
researchers take the technologies that are developed on the
research station and further "perfect" them to the farmers actual
production situation. The farming systems researchers learn how
to use the technology better which, in effect lowers the cost of
learning to all farmers. As the technology becomes perfected to
farm conditions the farmers who adopt the technology have less to
learn. The farmers unit production costs are lowered sooner and
the farmers expand the use of the technology sooner. Since the
cost of learning has been lowered fewer small farmers are
discouraged from adoption. The overall effect is more rapid
diffusion of technologies and more rapid improvement in the
community wellbeing. The farmers production knowledge of a
particular techno gy increases faster so that he can then be
learning the next technology that comes along. In a few years the
farmers' knowledge has snowballed just by increasing the
diffusion of new technologies; like the effect a higher interest
rate has on your savings after a few years.

Take the example of perennial peanuts. The FSR/E group
doubled the number of plantings this year and is continually
learning about the establishment of perennial peanuts on Florida
farms by Florida farmers. How long would the same progress have
taken without the farming systems work? How many years? Would it
have happened at all? Wheat is a much more developed technology
in Florida than perennial peanuts but the farming systems group
is working on lowering the unit production costs in this highly
competitive commodity too. By working on the complete nutrient
needs of wheat in North Central Florida the FSR/S group is
learning production information that farmers by themselves may
never have been able to learn.

Page 7

Farming system research increases the information available
to the farmer by fine tuning the technology to the farmers'
actual production conditions, in effect learning for the farmer.
The agricultural station researcher is happy because he gets to
see his technology be further refined and its adoption increased.
The extension agent is happy because he has a better technology
to extend and the farmer is happy because he benefits from using
the new refined, easier to learn, better technology. Society is
better off because farmers' income is increased and consumers get
a better and/or less expensive product, which is the purpose of
this whole research and extension process.

What about directing the farming systems work towards smaller
farmers? The farming system approach could be directed at any
group of farmers and be effective in increasing the generation
and diffusion of technologies for that group. In this case
directing the work towards smaller farmers lowers the unit
production costs, the cost of learning, for the particular farm
conditions on which the work is done, smaller farms. To the
degree that larger farmers have different resources than.smaller
farmers, the "fine tuning" of the technology for and by the
smaller farmers will not necessarily be appropriate for the
resource base of the larger farmers. Some technologies developed
for smaller farmers may not be appropriate for the large guys and
this overtly favors the smaller operators. Even when a technology
benefits large, medium and small farmers equally, the smaller
farmers still benefit because technology has been developed so as
to not discourage smaller farmers from adopting. So much the
better if all farmers can benefit from the technology.

This paper presents some ideas on how integrating learning
into agriculture can explain commonly observed diffusion and
adoption patterns and it also explains ways to exploit this
knowledge towards increasing the diffusion of improved
agricultural technologies. Hopefuly this will give you some good
ideas on how to explain learning to Dr. Woeste in ways that he
can better appreciate.

"Without fine tuning new production methods
to fit the physical and socio-economic
environment, the probability of farmers'
adoption will be severely reduced and the
benefits derived from investment in
agricultural research and extension will only
be a fraction of their potential." Zandstra,
H. G. "A Cropping Systems Research
Methodology for Agricultural Development
Projects." Paper presented at World Bank
Headquarter Seminar, Washington, D. C., 1
August, 1979.

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

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