THE ADOPTION OF
NEW MAIZE TECHNOLOGY
N PLAN PUEBLA,
THE ADOPTION OF NEW MAIZE TECHNOLOGY
IN PLAN PUEBLA, MEXICO
CENTRO INTERNATIONAL DE MEJORAMIENTO DE MAIZ Y TRIGO 1976
International Maize and Wheat Improvement Center, Apartado Postal 6-641, Mexico 6, D.F. Mexico
Correct citation: Winkelmann, Donald. 1976. The adoption
of new maize technology in Plan Puebla, Mexico. Centro
International de Mejoramiento de Maiz y Trigo. Mexico City.
vi + 24p.
I. Setting the scene 1
Program goals 1
Selection of a site 3
Characteristics of the project area 3
II. Organization of Plan Puebla 5
Experimental work 5
Plan Puebla methodology 5
Financing Plan Puebla 6
III. Recommendations, potential gains, and farmer responses 7
Recommendations and experimental yields 7
Farmer yields 8
Farmer response 9
IV. Factors impeding the diffusion of Plan technologies 11
On the behavior assumptions 12
Access to information and to inputs 12
Has diffusion really been impeded? 20
V. Reviewing conclusions 22
Lessons reinforced by Plan experience 22
Farmer compliance with recommendations 22
Implications for Plan hallmarks 22
FOREWORD:ON LAUNCHING THE STUDIES AND MOTIVATING HYPOTHESES
Launching the Studies
The study described in the following chapters is one of
a series aimed at enlarging understanding of the factors
impinging on the adoption of new maize and wheat
technology. Better understanding of the elements shaping
the diffusion of new cereals technology can help govern-
ments and development assistance agencies to increase
farmer income, hence the interest in the topic. Interest
increased as controversy about effects of introducing new
technologies attracted widespread attention to the theme.
CIMMYT, with its mandate defining its role in the
development and diffusion of maize and wheat technology,
quickly assumed a participant's role in the discussions. The
concern and the interest emanating from the critical im-
portance of the theme stimulated CIMMYT to look for a
modus operandi through which patterns of adoption and
the forces shaping those patterns could be identified.
Better understanding of these relationships would influence
CIMMYT efforts to develop new technology, the orientation
of its training program, and the approach taken in counsel-
ing governments about national programs.
In order to better comprehend what influences farmer
response to new technology, CIMMYT set out to facilitate
the research on which this and the other studies of the
series are based. We decided to examine eight cases in
which maize or wheat technology had been introduced to
farmers. In identifying programs for study, we limited
consideration to those in which the technology had been
available to farmers for no less than five years and in
which no less than 100,000 hectares of land might have
been affected. Eight programs were selected for study. For
maize the focus was on Colombia, El Salvador, Kenya west
of Rift Valley, and Mexico's Plan Puebla. For wheat,
programs in India, Iran, Tunisia and Turkey were consider-
ed. CIMMYT's maize and wheat staff participated in the
selection of these programs. With their knowledge of
programs around the world it was possible to choose a
varied set of experiences-e.g. programs with and without
irrigation, with and without effective price guarantees,
with massive extension effort and with virtually none.
To the extent possible, each of the adoption studies was
under the supervision of an indigenous economist. In only
one case was it necessary to turn to an expatriate and
there we had the good fortune to collaborate with a re-
searcher with several years experience in the area. Each
of the collaborators shared CIMMYT's concern for farmer
response to new technology.
Beyond sharing this concern, each collaborator had an
interest in farm level research done in close cooperation
with agricultural scientists. The importance of this interest
emerges from our conviction that agricultural scientists who
are knowledgeable about a particular maize or wheat area
can contribute substantively to research on the cereals
economy of that area. Their special knowledge about the
interaction between plants and their environments is im-
portant in identifying agro-climatic zones, critical periods
for the crop, and activities which are essential to effective
cultivation. Many agricultural scientists played a prominent
role in these studies; each warrants our gratitude for his
As the studies were completed it became apparent that
much could be said for publishing them in a standard
format. With several serving as Ph.D. dissertations and
others as less formal research pieces, a common format
could only be achieved through reworking the original
monographs. In every case but two, then, CIMMYT's pub-
lication is an abridgement of a longer piece. The Indian
study, itself a review of the findings of several other research
efforts, is being published in its entirety with no effort to
recast it in the form of the others. The Puebla study is also
In making the abridgements we have followed certain
norms. Mathematical proofs have been eliminated, litera-
ture reviews have been included only where they relate to
points which are unique to a given study, and the discussion
of the hypotheses motivating the studies have been dropped.
This last decision arises from recognition of the substantial
commonality of these hypotheses among the studies. This
suggested that, rather than presenting essentially the same
discussion in the text of each abridgement, the hypotheses
could be treated once in an abbreviated form for all studies.
While each of the studies examines a somewhat different
set of circumstances, all depart from the same general
assumption about farmer behavior. The assumption is
that farmers are income-seeking risk averters who are
sensitive to the nuances of the environment in which they
farm and that they are generally effective in their decision
making. For the six studies based on original survey data
and to a more limited extent for the study of Plan Puebla,
this common point of departure leads to a great deal of
similarity in the motivating hypotheses.
Given a farmer oriented by the assumptions described
above, we might expect to see relationship between the
adoption of elements of the new technology and 1) char-
acteristics of the farmer-his age, education, family size,
farming experience, off-farm work, percentage of land own-
ed, 2) characteristics of the farm-its agro-climatic region,
competition of industrial crops, relative importance of
cereals, nearness to markets, farm size, 3) characteristics of
government programs-access to credit, access to informa-
tion (though extension agent visits or visits to demonstra-
Some of the relationships between these variables and
the adoption of elements of the new technology are more
arguable, some less. Least arguable are hypotheses relating
adoption to education, farming experiences, percentage of
land owned, more favored climatic regions, relative import-
ance of cereals, nearness to markets, farm size, access to
credit, and access to information. With other things equal
and accepting our assumptions that farmers are income-
seeking, risk-averting, sensitive, and effective maximizers,
virtually no one would argue that any one of these relation-
ships should be negative.
Somewhat more arguable is the relation of age and family
size to adoption. Even here it is likely that only a few
would argue that these relationships might be positive.
Most arguable are the relationships linking adoption to
off-farm work and competition of industrial crops. With
respect to the former, some hold that the relationship is
positive as more off-farm work implies more income, there-
fore a greater capacity to bear risk, hence a greater willing-
ness to adopt new technologies. Others hold the converse,
arguing that more off-farm work implies less interest in the
farm, hence less willingness to put in the time and energy
associated with taking on new technologies. So too for
industrial commodities, where those who see the relation-
ship as positive allude to greater experience with improved
inputs and larger incomes while the contrary view rests on
capital restrictions and the high opportunity cost of labor.
With knowledge of the relationships among these var-
iables, researchers and policy makers can better develop and
diffuse new technologies. Some of the variables considered,
e.g. age and family size, are beyond the control of these
decision makers. Nonetheless, by incorporating them in the
analysis the effects of variables subject to their control are
more clearly discerned. Knowledge of how these variables,
e.g. agro-climatic zones and extension programs, relate to
adoption can be of critical importance in affecting the
development and diffusion of new technology.
With this rough sketch of the general argument, readers
wanting more detail about the derivation of the hypothe-
sized relationships can turn to the relevant original piece
from which this series of abridgements was drawn. In
all cases the studies feature the effects of agro-climatic
region and farm size on adoption of elements of new
technology. This emphasis is related to the earlier contro-
versy about the effects of new technology where these two
factors played prominent roles.
Before moving to the abridgement, some attention to
the phrase "elements of the new technology" is warranted.
Much has been made of the concept of a package of practi-
ces in the introduction of new technology. We've chosen
to look at this a bit differently, taking the view that the
differences in risk, expected income, and cost of each
element of the technology are large enough to outweigh
the effects of the interaction among these elements. That
is to say, perceptive and prudent decision makers might
well choose to take up only a part of the package rather
than the entire package. For the programs studied, the
two dominant elements in the package are improved seed
and fertilizer. These two were analyzed as dependent var-
iables in each of the studies. Of lesser importance are
such elements as seed treatment, date of planting, method
of planting, use of herbicides, use of pesticides, planting
density, and seed bed preparation. Nevertheless, where
any of these was recommended and where data are adequate,
these are also treated as dependent variables.
While CIMMYT has been associated with these studies
since their inception, the opinions expressed by the authors
are not necessarily endorsed by CIMMYT.
This report describes the adoption of more intensive
maize technology in the Puebla area of Mexico. It is based
on data made available to CIMMYT by Plan Puebla and relies
especially on the annual yield survey. I'd like to thank,
without implicating, several whose comments have helped
the report, among them: James Bemis, Steven Breth, Hugh
Bunting, Ralph Cummings, Jr., Reggie Laird, Gregorio Marti-
nez, and Robert Osler. And a special thanks for Michael
Sainz who worked up the data.
I. SETTING THE SCENE
For some 20 years, from the mid-1940's to the mid-1960's,
Mexico was the scene of active programs aimed at increas-
ing agricultural production, especially of maize and wheat.
These programs were a joint undertaking of the Ministry of
Agriculture and the Rockefeller Foundation. By 1965 the
efforts in wheat had achieved preeminent success, with
average yields moving from roughly three quarters of a ton
to some 2.5 tons per hectare, an increase of over 200 per-
cent. This substantial increase rested on new varieties, wide-
spread use of fertilizer, irrigation, and weed control all
commonplace in the wheat economy. The situation was
sharply different in maize, even though parallel efforts had
been made. Average yields increased from roughly 0.7 tons
to a bit over 1.1 tons per hectare, something under 50 per-
Several critics sought to explain why the one program
worked so well, with virtually 100 percent adoption of new
varieties and dramatic increases in yields, while the other
lagged behind, with less than 15 percent of potential area in
new maize varieties and barely perceptible annual increases
in yields. One explanation [1 j held that the difference
emerged from wheat's cultivation on irrigated land by large
farmers as contrasted with maize's cultivation on rainfed
land by small farmers. These differences in the natural and
economic circumstances of farmers were seen as primary
forces in shaping the very dissimilar evolutions of the two
A second related explanation held that small traditional
farmers are not amenable to change. Something, perhaps an
intransigent traditionalism, made such farmers unreceptive
to new technologies, no matter what these promised in
terms of yields and profits.
How, then, to improve the livelihood of small farmers?
Under what circumstances would they shift to new technol-
ogies? Indeed, would they shift at all?
These were the questions being asked by Mexican spe-
cialists in agriculture and their colleagues in the Rockefeller
Foundation's Mexico office in the mid-1960's. They were
particularly relevant questions given the threat of world
wide food shortages and the related problem of low in-
comes and malnutrition among farming populations. It was
precisely the small traditional farmers of Mexico and the
world who suffered most from low incomes and malnutri-
tion, precisely these farmers who seemed most reluctant to
take up new practices which apparently promised relief
from their straitened condition.
Motivated by the difficulties with maize in Mexico, a
picture reinforced by similar scenes from around the world,
Rockefeller Foundation set out to design a new program. It
was to have a thrust unlike that of previous efforts and it
was to be set in an environment carefully chosen to make
evident its strengths and weaknesses.
The program came to be called Plan Puebla, named after
the valley where it was implemented. This monograph des-
cribes Plan Puebla. It draws heavily on data made available
by Plan Puebla staff and, in its early chapters, on the sub-
stantial number of papers already written about the Plan.
While an attempt is made to describe each of the Plan's
important features and experiences, attention is concen-
trated on the extent to which various classes of farmers
adopted the recommended technologies and the elements
which might have been instrumental in shaping their dif-
The remainder of the introduction describes in more de-
tail the goals of Plan Puebla, outlines the factors leading to
the choice of the Puebla Valley, and presents some charac-
teristics of maize production in the area.
Plan Puebla is a development effort aimed at the traditional
farmer. At its inception it was described as. "an attempt
to tackle simultaneously. ." two development problems
-food shortage and low incomes in agriculture. "by ob-
taining massive increases in yield among small holders..."
From its initiation the program had two overriding ob-
jectives. The first was to develop, field test, and refine a
strategy for increasing yields of a basic crop among small
holders. The second was to train technicians in the use of
this strategy. A third objective, added later and probably
designed to provide a standard against which progress could
be measured, was to double maize yields in the program
Consider the two primary goals in the light of earlier ex-
perience in Mexico and elsewhere. Those involved in devel-
oping and diffusing new maize technologies for Mexico had
apparently followed the classic pattern. They had worked
diligently on experiment stations to develop new varieties
and the results of trials showed that they had been effective
in these efforts, at least for certain areas. Varieties in hand,
and again based on experimental work, they had formu-
lated agronomic practices dates of planting, fertilizer
These new technologies or production strategies seemed
to promise substantial increases in yields over traditional
practices. More important, they apparently promised a
sharp increase in profits. The technologies were then put in
the hands of the extension service for diffusion among
There, in substantial measure, the expectations of re-
searchers were disappointed. Farmers, especially those of
the rainfed altiplano where the bulk of Mexico's maize area
is found, persisted in their old ways, rejecting the strategies
offered by representatives of Government's agricultural
What went wrong? Were the strategies too complex?
Too expensive? Did they not fit the cropping pattern?
Were such essential inputs as viable seed and fertilizers not
available? Had the farmer no credit? Were the promised
profits more apparent than real?
These experiences and questions led some of those asso-
ciated with Mexico's maize program, with the Graduate
School of Agriculture, with The International Maize and
Wheat Improvement Center, (CIMMYT), and with the
Rockefeller Foundation to organize a new experiment. The
experiment postulated that a method for attaining yield
increases could be found. What was required was an experi-
mental approach, an approach which took nothing for
granted but regarded each element in the chain from re-
searcher to farmer as a variable.
There was in all of this one critical assumption, never
made explicit but always there. The farmer was regarded as
purposive in his behavior, goal oriented and with profits as
an important goal. Little heed was given to the possibility
that traditionalism was the force blocking change. Rather,
the difficulty was seen to be on the side of the institutions
charged with developing, diffusing, and supporting new
technology. This view was not widely held in 1966, cer-
tainly not among leaders of those critical agricultural insti-
tutions. Their view, again not often made explicit, was that,
Fig. 1. The Plan Puebla area.
in fact, small farmers were tradition-bound and that the
traditions themselves prevailed against the diffusion of new
technology no matter how profitable the technology nor
how effective the extension service.
With the view that farmer behavior is purposive, with
emphasis on profits, what was needed was a system for
developing and delivering useful technology to farmers. Dis-
covering that system was the primary objective of the pro-
Once the new system was operational and demonstrably
effective in fomenting yield increases, the second primary
goal would become critical. That goal was to train others,
from Mexico and the rest of Latin America in particular, in
the use of the system. These newly trained people were to
then transfer the methodology to their own regions or
countries, promoting yield increases there.
All of this, of course,. would lead straight away to in-
creases in food production and to improvements in the nu-
tritional status of the adopting farmers. It would also lead
to other changes in practices as researchers, extensionists,
and farmers applied the lessons learned in maize to other
crops. The system, then, would be a vehicle for introducing
a dynamic element into traditional agriculture.
And what of those suspicious decision makers, those
who pointed to traditional farmers as recalcitrant conserva-
tives? Their attitudes would necessarily change as changing
practices and yield increases signaled the farmers' willing-
ness to adopt new technologies.
These were the goals that oriented the efforts of those
associated with the project -ambitious but pragmatic,
-emerging from their own frustrated attempts to foster
Selection of a Site
Once committed to the idea of the project a site had to be
found. Clearly it was neither feasible nor wise to think in
terms of the entire country. Given the intentions, what was
needed was an area with appropriate characteristics.
After careful deliberation among professionals represent-
ing several disciplines e.g. soils science, maize breeding,
communications a set of characteristics was identified. In
a general way it was thought necessary to site the project in
an area which offered the possibility of obtaining a marked
increase in maize yields.
More specifically, the experiment would best be under-
taken in an area: 1) with a large number of small farmers;
2) where maize played an important role in the cropping
pattern; 3) where the probability of loss due to drought,
frost, or hail was small; 4) where government infrastructure
in the form of credit agencies, market roads, and crop insur-
ance was available; 5) and where farmers had ready access
to markets with stable prices [3, pp. 130-1 ].
Given these characteristics, it was decided to locate the
project in the Puebla Valley. The area is in a high valley,
ranging in altitude from 2150 to 2700 meters above sea
level, some two hours drive east of Mexico City. (See Fig.
The list of desirable characteristics again signals the con-
viction of the project's initiators that, with the right cir-
cumstances, small traditional farmers would change their
farming practices. What this list emphasizes is potential
average profits and ready access to credit, to inputs, and to
product markets. This is a tacit assertion that farmer behav-
ior is purposive, oriented toward profits. What was needed
was a profitable technology and a system for diffusing it. A
unified strategy for achieving these ends was the aim of the
Characteristics of the Project Area 1
Maize yields in the area selected appeared to average be-
tween 1 and 1.5 metric tons. In 1967 they were 1.3 tons
per hectare according to data from a bench mark survey
carried out by the project. Four weather stations in the area
reported average rainfall from just under 800 mm to over
850 mm in the seven months from April through October.
Soils vary but over 60 percent of the arable land is in essen-
tially problem-free soils while about 35 percent is in soils
where yields might be limited because compacted or imper-
meable horizons restrict root growth. Even so, given the
rainfall, soils, and the growing season, it seemed clear that
substantial increases in yields could be achieved through the
introduction of new technology.
Farms and farmers. Looking now at the other character-
istics thought to be desirable, the area fits all of the criteria
well. The roughly 117,000 hectares of cultivated land were
divided into farms which averaged 2.7 hectares according to
surveys made in 1967 and 1970. This implies 43,300 farms
in the area. Two thirds of the farms had less than 2.5 hec-
tares of cultivated land.
Only 27.5 percent of the farms were completely pri-
vately owned. The remainder were in ejidos or were a com-
bination of ejido and private property. The ejido is a pecul-
iarly Mexican tenure form. Emerging from Mexican expe-
rience in land reform in the 19th century, its distinguishing
characteristics are that it invests usufruct rights in the farm-
er but does not permit him to sell, rent, or mortgage the
land. Roughly half of Mexico's cultivated land, both irri-
gated and rainfed, is in ejidos.
While the amount of land in ejido farms varies from
place to place, such farms are generally small. In the project
area farms with only ejido land averaged a fraction over 2
hectares. Again about half of the total of cultivable land
was in ejidos. It should be noted that ejidos are usually
farmed individually; something under 2 percent of Mexico's
ejidos are collectives and none of these are in the project
About 75 percent of the farmers had 3 years or less of
schooling. Many of them worked off the farm, with off-
farm wage income making up nearly one quarter of the
estimated net income. Crops contributed some 30 percent
to net income. Nearly all of the area's farmers had heard of
chemical fertilizer and over 80 percent of them had used it
prior to the project's initiation.
Importance of maize. Maize occupied nearly 70 percent
of the cultivated land according to the surveys of 1967 and
1970. The next crop in importance was beans pole beans,
bush beans, and scarlet runner beans, all of the genus Phas-
eolus occupying just over 15 percent of the cultivated
area. Maize and pole beans were often found intercropped.
Weather. Analysis of weather data suggested that prob-
abilities of loss due to weather damage were moderate to
slight. Two of the weather stations reported no frost in May
and June, one station reported frost in May in 17 percent
of the years and June frosts in 5 percent of the years.
Frosts occur mainly in October through March causing little
or no damage to maize. In 1974 an early September frost,
one of the earliest on record, did cause extensive damage to
the area's maize.
While hail occurred occasionally it didn't appear to be
particularly worrisome. Three of the area's four weather
stations averaged one hail storm in July and August, half
that number in September.
Rainfall patterns were such that project specialists esti-
mated the probability of severe drought at 0.1, with an esti-
mated 60 percent reduction in yields, and moderate
drought at 0.3, with an estimated 30-60 percent reduction
in yields. The estimated drought frequencies and losses
were thought to be low enough as to not influence average
yields unduly, hence not significantly affect farmer re-
sponse to a new technology.
Access to credit. Mexico has long had agricultural agen-
cies catering to the needs of some of its small farmers.
Notable among these was Banco Ejidal, a Government
agency established in 1935 and reorganized several times
since.2 The Ejido Bank served Mexico's ejido farmers, i.e.
the ejidatarios. The services it offered varied enormously
from one part of the country to another.
There were also Government banks which served private
farmers, the so-called pequeios propietarios who operate
virtually all farms not classed as ejido. Chief among these
banks was Banco Agricola. As might be imagined, banking
services available to the smallest private farmers were lim-
In 1967 the Ejido Bank in Puebla made loans to only
925 farmers in the entire state. The Agriculture Bank made
60 loans, few of those to the smallest farmers. At least two
factors contributed to the small number of loans. Bank pro-
cedures were cumbersome and at times demeaning, so that
farmers were something less than eager to seek credit. The
banks, on the other hand had suffered low repayment rates
on their loans to small farmers. The Ejido Bank in Puebla
reported repayment of about 40 percent of the loans made
in 1967 [6, p. 196
In short, then, while banking services existed in Puebla,
they were not widely used by the area's farmers, neither
ejido nor pequeiio propietario.
Insurance. Crop insurance was also available, indeed was
mandatory for those borrowing through the government
banks. This service in Puebla is a part of a national program
under Aseguradora Nacional Agricola y Ganadera. That
program has been roundly criticized by Puebla farmers as
not paying out when payments are justly due and/or being
too expensive. Moreover, until recently the insurance cov-
ered the loan rather than the crop itself. Critics of the pro-
gram say that it is less an actuarially oriented insurance pro-
gram, more oriented toward redistributing incomes from
richer to poorer farming regions. Nonetheless, farmers
wanting loans from the Ejido Bank or from the Agricultural
Bank were required to buy insurance 16, p. 128 ].
Markets. Market roads in the area were good. Several
paved roads ran through the project's domain, connecting
all of the larger towns. Virtually every small town lies along
or near to the feeder road.
Finally, markets with stable prices were available.
Mexico has long had a system of price supports for basic
commodities maize, wheat, beans, and rice. Individual
farmers were not always able to sell at these prices as CO-
NASUPO, the agency responsible, was at times short of
funds, or storage space, or particularly keen on enforcing
quality control. Nonetheless, the very existence of the
agency kept prices up and ensured that annual price fluc-
tuations were not notable. The guaranteed price of maize
remained at $940/ton from 1967 through 1972, and then
went to $1200 in 1972, $1350 in 1974, and to $1750 in
1975. Meanwhile, area market prices ranged from a low of
abouit $750/ton in 1967 to a high of over $2000/ton in
early 1976 (12.50 pesos -- 1 dollar).
What follows is a review of the experience of what has
come to be called Plan Puebla. The next chapter treats the
organization of the project's activities. The third chapter
deals with what happened by describing the promise of the
new technology, as manifested in the demonstrations and
trials on farmers' fields, and the response of farmers to the
technology, as manifested in its adoption. The fourth chap-
ter treats factors which might have limited adoption of the
technology. A final chapter offers a summary and con-
The report focuses on Plan Puebla experiences during
the period, 1967 through 1975. Its early chapters rely on
several studies, especially , , , and . It also
relies heavily on two sets of data developed by Plan Puebla.
The first comes from surveys undertaken in 1967 and in
1971 and the second from annual yield surveys,
1. Discussion in this section relies heavily on [4,p. 1-7).
2. In 1975 three Mexican agricultural credit banks were merged in
what is now the Banco Nacionat de Credito Rural.
II. ORGANIZATION OF PLAN PUEBLA
It was evident from the discussions that led to the founding
of Plan Puebla that several levels of activities would be
emphasized as work evolved. The two major questions were
how to organize the experimental work so as to formulate
recommendations and how to diffuse the information
among the Plan area's farmers.
Early consideration suggested that work on formulating
technology should feature plant breeding and agronomy.
This emphasis led to immediate incorporation of the rele-
vant professionals on the Plan's staff. Research was initiated
in 1967 with two plant breeders and two agronomists. Each
was assisted by several field workers, often themselves farm-
ers from the area. Their work was reinforced by profes-
sionals from CIMMYT headquarters, nearby in the Valley
Recommendations were formulated for the 1968 season.
In that year they were tried on the farms of 103 farmers,
the first participants in the Plan. The recommendations
emphasized agronomic practices and relied on local maizes.
These recommendations aimed at maximizing the average
profits from fertilizer use.
Prior to Plan Puebla the Ministry of Agriculture had a
single recommendation for rainfed maize in the area. This
was used as a point of departure by Plan professionals in
designing their experimental work. At the conclusion of the
first year's experiments this recommendation was revised in
a second approximation to suitable recommendations.
Successive approximations were made in 1969 and 1970. In
1971 there were additional recommendations for compac-
ted soils and for an additional region. In 1972, recommen-
dations were made for sixteen producing systems. For each
system two recommendations were given, one for limited
and one for unlimited capital. Some of these recommenda-
tions were altered slightly in ensuing approximations.
Work on varieties proceeded through 1972 when it was
decided to phase out the program on varietal improvement.
The decision was largely a consequence of evidence that
some local varieties had extraordinarily high yield potential,
on the order of five times the area's average yields.
It was decided to expand work on communications slowly,
starting with one specialist in 1968, to four specialists in
1969, and to five specialists in 1970. This number was dic-
tated by the decision to divide the area into five zones, the
fifth coming under program recommendations for the first
time in 1971. Each extension agent worked with field as-
sistants, many of them farmers from the area. Supplemental
professional expertise was available from CIMMYT.
In addition to disseminating information on the recom-
mendations the extension technicians worked with groups
of farmers to help them obtain credit, achieve timely deliv-
ery of inputs, and repay loans. This set of activities became
so important in the early 1970's that the number of demon-
stration plots and field days was actually cut back because
the extension force lacked time to both work with farmer
groups and to mount demonstrations.
All of these activities must be seen in the context of the
Project's goals to develop and field-test a methodology
and to train others in its use. During the first years the
methodology evolved, a method which came to distinguish
Plan Puebla from other programs with similar intent.
Plan Puebla Methodology
By 1969 the methodology was essentially fixed. It featured
experimental work on farmers' fields; the use of farmer
groups for disseminating information, getting credit, acquir-
ing inputs, and repaying loans; socio-economic evaluation;
and administration aimed at coordinating the work of re-
searchers, extensionists, agencies providing credit and in-
puts, and farmers.
Experimental work on farmers' fields was a part of the
Plan from the outset. Working through farmer groups was
introduced in 1969. Socio-economic evaluation and coor-
dination were part of the project from the beginning but
their roles expanded as new potential for these functions
became evident. Coordination especially increased in im-
portance as work with the several government agencies in
the area expanded.
After 1974 emphasis within Plan Puebla changed. Coin-
ciding with a growing concern by the government for in-
creasing production in traditional agriculture, Staff began
to concentrate even more on the training function. Exten-
sion workers from all over Mexico were brought to Puebla
where they were trained in the Plan Puebla approach to
facilitating change among small farmers.
In effect, after 1974 Plan Puebla operated two sets of
activities in tandem. The one, through the operational arm
of Plan Puebla, continued to emphasize increasing the area's
production of maize and responding to other demands of
the area's farmers. The second, through a training center,
emphasized training advisors for other programs in Mexico.
Collaboration between the two is close with Plan Puebla
providing the expertise and the laboratory for the training
Training of Mexican and foreign nationals was initiated in
1970. Over the next four years 45 Mexican and 25 foreign
national participated in the training program. Of these, 44
received six to nine months of training at the technical level
while 22 were trained in technical issues at Puebla in con-
junction with work on Masters' degrees in the Graduate
College at Chapingo.
The orientation of training clearly bore the stamp of the
experiences and philosophy of a program aimed at develop-
ing and diffusing new technology. First, it featured concen-
tration on interdisciplinary work, work combining the func-
tions of breeding, agronomy, evaluation, and coordination.
Second, since coordination and planning were recognized as
critical functions within Plan Puebla, the program aimed at
developing trainee skills in project planning and manage-
ment. Finally, just as training was emphasized in Plan
Puebla, the training program sought to develop the training
skills of the trainees so that they might in turn serve as
With the change in program financing in early 1974 and
the accompanying change in program emphasis, the training
program has also shifted its orientation to one more in
keeping with the Plan's new responsibilities.
Financing Plan Puebla
For the first years of its operation Plan Puebla was financed
by grants from the Rockefeller Foundation through
CIMMYT, by CIMMYT itself, and by the Graduate College
of Chapingo. During that period, from 1967 through 1973,
Rockefeller Foundation contributed $560,000, CIMMYT
contributed $333,000, while the Graduate College contrib-
uted $30,000. This is a total of $925,000 for activities in
Puebla aimed at developing and field testing the method-
ology. Related activities and costs were $87,000 for con-
sulting services to the Project team, $112,000 for consult-
ing with other programs within Mexico and in other coun-
tries, and $168,000 for training.
In 1974 the program was taken over by the Mexican
Ministry of Agriculture and directed from the Graduate
College at Chapingo, a dependency of the Ministry. At that
time, as noted previously, the aims of the program changed
with much more emphasis put on the training of Mexican
The initiators of Plan Puebla regarded the selection of staff
as a crucial element in determining the success of the Plan.
It was evident that the work would be physically demand-
ing, that it would require a great deal of knowledge about
the decisions faced by farmers, and that it would require a
sound knowledge of the biological phenomena impinging
on maize production in the area. This list of qualities meant
that every effort had to be made to attract and hold ener-
getic, well trained, and innovative people.
The criteria employed in staff selection led to a team
than was not representative of the general level of compe-
tence found in agricultural programs within developing
countries. In a sense, then, here the experiment loses gener-
ality. If, that is, the methods developed required qualities
different from those generally found in national programs,
then the resulting methodology would have limited appli-
cability. While recognizing this possibility and emphasizing
the crucial role of competent staff, the program's initiators
believed that developing and field testing the program re-
quired more skills than would be required to operate the
program once a system was established. Given this perspec-
tive, they could, at one and the same time, insist on high
standards for staff and still maintain that the methods de-
veloped would be generally applicable.
Table 1 shows the staffing pattern from 1967 through
1975. Two classes of personnel are represented in the table.
The first is people associated with the project on a full time
basis. Some 53 man years of staff time, involving 43 differ-
ent people were absorbed by the project from 1967 to
1973. In 1974 and 1975, over 30 man years of professional
time were contributed by staff. The second is additional
professional counsel from technical advisors. Most of these
were staff members of the Post Graduate College at Cha-
pingo or of CIMMYT. As a final note on staffing, many of
those associated with Plan Puebla in the past are still in-
volved in agriculture in Mexico, most with programs aimed
at small farmers.
Plan Puebla's organizational format and modus operandi
emerged early in the life of the project. The format fea-
tured a coordinated effort in agronomy, communications,
and evaluation. In terms of methods, the hallmarks of Plan
Puebla are research in farmers' fields, diffusion of technol-
ogy and inputs through groups of farmers, continuing eval-
uation and feed back to the professional staff, coordination
of the interests of farmers, plan staff, and local institutions.
In one sense the experimental phase of Plan Puebla, i.e.
Table 1. Professional personnel associated with Plan Puebla from 1967
1967 1968 1969 1970 1971 1972 1973 1974 1975
Staff 5 6 10 11 12 10 9 16 25
Coordinator 1 1 1 1 1 1 1 1 1
Agronomy 2 1 2 2 2 2 2 2 2
Breeding 2 2 2 2 2 1 -
Evaluation 1 1 1 1 1 1 1 1
Extension 1 4 5 5 5 5 5 5
Trainers 7 10
Advisors 2 3 4 5 5 6 6 4 4
the phase in which methods were identified, tested, and
evaluated, was brief. By 1970, after some two years of
effort in the area, the methodology was largely fixed. More-
over, it appears that the research for alternative approaches
was limited. Still, there is evidence that emphasis on activi-
ties changed as experiences accumulated.
III. RECOMMENDATIONS, POTENTIAL GAINS, AND FARMER RESPONSES
The intent of this chapter is to compare the yields asso-
ciated with the recommended technology with that of the
technology typically used in 1967. It will be seen that
apparent potential gains are substantial. By contrast, farmer
adoption of the technology does not meet the expectations
engendered by yield differences.
Recommendations and Experimental Yields 1
A ministry of Agriculture recommendation was available to
Puebla farmers in 1967. It featured the hybrid H-28, 80
kg/ha of nitrogen, and 40 kg/ha of phosphorus with 40,000
plants per hectare. This recommendation was used as the
point of departure for plan sponsored experimental work
According to the 1967 bench mark survey some 62 per-
cent of the farmers applied fertilizer on maize. Some 72
percent used less than 40 kg/ha of nitrogen [3, p. 134 ]on
maize. The average fertilizer application was 34 kg/ha of
nitrogen, 14 kg/ha of phosphorus, some potash. Approxi-
mately 25,000 plants were being grown per hectare. Many
farmers were applying a fertilizer mixture of nitrogen, phos-
phorus, and potash supplied by the national fertilizer
monopoly, Guanomex. Few farmers used hybrids, less than
one percent in 1967, but most farmers knew of them and
many had tried them.
In 1967 agronomic work was initiated on farmers' fields.
The experiments used both local maizes and recommended
hybrids. The results suggested that hybrids had little advan-
tage over the local maizes, that fertilizer levels could be in-
creased, and that planting densities could also be increased.
For 1968 the Plan recommendation was 130 kg/ha of nitro-
gen, 40 kg/ha of phosphorus, and 50,000 seeds per hectare
of the farmer's local variety.
Experimental work in 1968 and 1969 led to three
recommendations in 1970. By that time the area had five
extension zones. Two of the recommendations applied to
four of the zones 130 of nitrogen, 50 of phosphorus, and
50,000 plants for deep soils and 110 of nitrogen, 50 of
phosphorus, and 50,000 plants for areas with a compacted
horizon. The third recommendation applied to the fifth
zone and featured 80 kg/ha of nitrogen, no phosphorus,
and 40,000 plants. A reduction of the relative price of ferti-
lizer in 1971 led to an increase in the recommended nitro-
gen levels, from 80 kg/ha to 100 kg/ha, and planting densi-
ties from 40,000 to 50,000 plants in the fifth zone.
New sets of recommendations have been developed --
e.g. recommendations for those with limited capital and
recommendations incorporating local peculiarities. In
general, however, the recommendations described are the
basic strategies for the area. As compared to practices in use
when the project was initiated they feature more nitrogen,
more phosphorus, and a greater planting density.
The new strategies also call for early application of some
nitrogen and all of the phosphorus as contrasted with the
then prevalent practice of applying all fertilizer with the
first cultivation. Application of phosphorus at the first cul-
tivation makes that nutrient virtually unavailable to the
plant at the time when it is most critically needed.
The additional yields promised by increasing fertilizers
and plant densities were notable. In 1968 Isee 4, p. 24 ]for
deep soils the average estimated yield was 7462 kg/ha of
maize using profit maximizing levels of 187 kg/ha of nitro-
gen and 81 kg/ha of phosphorus. This compared with a
yield of 1028 kg/ha for the control treatment. Even so,
such high levels of inputs were not recommended because it
was recognized that 1968 was an extraordinarily favorable
year for maize.
Based on yields and yield differences for a longer period
the potential promised by the recommendations is lower
but still significant. These differences are shown in Table 2.
The yields are based on data from 14, Table 3.11, p 33]
and include estimates for two soils types and three planting
Even after deducting the farmgate cost of the extra ferti-
lizer the potential advantage of the Plan recommendations
seems attractive. This is evidenced in Table 3 where the cost
of fertilizer has been put in terms of maize and the extra
fertilizer implied by the Plan recommendations deducted
from the extra yield attained.
Table 2. Estimated average experimental yields (ton/ha)
1967-73 for various treatments in three zones of Plan
Zone technology technology c
1 -4 (deep soils) a 1.72 3.91
1 -4 (compacted soils)a 1.88 3.18
5 2.50 4.55
a/ 14, p. 33] reports experiments for several soils types and planting
dates. Two soils types with three planting dates each make up
55 percent of Zones 1-4, An average of the six, weighted by
area, was calculated and then used as described in footnote c.
Remaining soils types in the region were characterized by few
observations so were excluded from the averages. b/ [4, p. 33]
reports zero N: zero P, 50 N:25 P, and 80 N:40 P. The average
fertilizer application in 1967 was 34 N:14 P. This is called the
Conventional Technology. Observations on the three points were
plotted and the yield of the Conventional Technology estimated by
interpolation. c/ Data reported in [4, p. 33] are estimated farmer
yields, 80 percent of experimental yields. These were increased by
25 percent to give experimental yields.
As noted in the previous section, researchers were able to
demonstrate substantial increases in yields with the recom-
mended technologies. In this section the yields obtained by
farmers are examined.
In each year from 1969 to 1975 Plan staff conducted a
yield survey of farmers' fields. Samples were selected from
each of two classes of farmers, from all farmers in the pro-
ject area (general sample) and from farmers who were ob-
taining credit for maize under Plan auspices or through offi-
cial banks (participant sample). Each sample was randomly
selected. In the first case a two stage sampling procedure
was used which identified first a segment and then a strip of
maize within a segment. The farmer associated with the
strip was then interviewed. In the second case farmers were
selected from the list of credit users and then a strip of
maize was identified for each selected farmer.
In 1969 and 1970 only the yield and plant density at
harvest were estimated for each strip. From 1971 through
1975, in addition to yield and density, each farmer was
asked about the application of nitrogen and of phosphorus.
For 1972, 1973, and 1975, farm size was also recorded.
The discussion of this section is based on yield data from
1971 through 1973. Early frosts, the earliest in over 50
years, make 1974 an extraordinary year. Yield data for
1975 were not available when this report was written.
To better reflect the change in yields accompanying the
use of the recommended technology each observation, irre-
spective of sample, was considered for one of four groups.
The first group approximates the 1967 conventional tech-
nology in Zones 1 through 4 and includes all of those
applying less than 50 kg/ha of nitrogen, less than 25 kg/ha
of phosphorus, and with between 15,000 and 35,000 plants
at harvest time. The second includes the same observations
but for Zone 5. The third group is made up of those obser-
vations from Zones 1 through 4 with 90 to 160 kg/ha, of
nitrogen, 30 to 70 kg/ha of phosphorus, and 40,000 to
55,000 plants at harvest time. The fourth includes observa-
tions from Zone 5 with 70 to 140 kg/ha of nitrogen, less
than 40 kg/ha of phosphorus, and 35,000 to 55,000 plants
at harvest time.
Groups one and two are representative of prevailing
practices in 1967 while groups three and four approximate
the practices recommended by the Plan. The average yields
over three years 1971 through 1973 are seen in Table 3.
Earlier years, 1968 through 1970 were excluded because of
the absence of data on fertilizer use. In all cases the experi-
mental yields for trials on farmer's fields exceed the esti-
mated farmer yield. The difference is far more notable in
Zone 5 than in Zones 1-4.
To compare farmer yields with those of Table 2 it is
necessary to average the two yields given there for Zones
1-4. Data in 14, p. 33 ], can be used to establish the pro-
portions of land in deep soils and with connected horizons.
These proportions can then be used as weights. With 60 per-
cent of the area thought to be in deep soils the resulting
averages are 1.78 tons/ha for the conventional technology
and 3.62 tons/ha for Plan recommendations. These averages
represent Zones 1-4.
Comparison of the yields in Table 4 again show that
those using roughly the recommended treatments achieving
substantially higher yields than those using 1967's conven-
tional practices. Even after deducting the cost of fertilizer
(see Table 5) the yield advantages are still evident.
Yield increases in Zones 1-4 under farmer circumstances
(Table 5) compare with those under experimental circum-
stances (Table 3) in Zones 1-4. For Zone 5, the increase
under farmer circumstances is somewhat smaller than under
researcher control. In either case yield increases appear to
be large enough to appeal to the area's farmers.
The first effort to attract farmers to the new technology
was made in 1968 when demonstration plots were set out
on the fields of 103 farmers. These plots were under the
supervision of Plan professionals who worked closely with
farmers to ensure that all recommendations were carried
By 1969 four extension agents were assigned to the Plan,
one to each of the first four areas incorporated in Plan acti-
vities. These men, each supported by one or two field assis-
tants, were involved in a multi-media campaign incorpora-
ting radio, newspapers, a movie, and portable sound sys-
tems. Their aim was to attract farmers to the Plan, to then
help interested farmers form groups, to instruct the groups,
and to then help groups to obtain credit and needed inputs.
From 1970 on each of the Plan's five Zones had an ex-
tension agent with assistants. Their work, which started
with emphasis on demonstration plots and field days,
moved even more toward assisting groups in obtaining in-
puts and repaying loans. By 1972 and 1973 the number of
demonstration plots actually declined as extension agents,
by now heavily committed to helping groups, found them-
selves with insufficient time to maintain the number of
demonstration plots [6 1
The combination of activities experimenting, demon-
strating, promoting the Plan, facilitating input flows
-undertaken by Plan staff, led to rapid expansion in the
number of farmers associated with the Plan. Since that
early flourish the number has increased at a lower rate.
Participation in the Plan has been measured in several
ways but two predominate. One is the number of farmers
who received credit for producing maize. A second relates
to the number of farmers following some portion, perhaps
all, of the Plan's recommendations. The second measure
seems the most appropriate, because some farmers receiving
credit are not following recommendations while some farm-
ers not receiving credit are doing so. Even so, data on the
first measure is instructive. Table 6 describes a series of
Table 3. Estimated adjusted yield increase
(ton/ha) of Plan-like strategy over conven-
tional strategy under experimental circums-
Zone Net yield increase
1 4 (deep soils) 1.35
1 4 (compacted horizon) 0.58
a/ Adjustment is made for the cost of extra
fertilizer by converting each year's farm price of
fertilizer to that year's farm price of maize, multi-
plying and subtracting. Calculations based on [5,
pp. 67-72], prices are average for 1967, 1971, and
Table 4. Estimated farmer yields (kg/ha)
with 1967-like and Plan Puebla-like strategies,
averages for 1971 through 1973.
Zone technology strategies b
1-4 1685 3434
5 1946 3113
a/ less than 50 kg/ha of N, less than 25 kg/ha of
phosphorus, and 15,000-30,000 plants/ha in both
areas. b/ 90-150 kg/ha of N, 40 to 70 kg/ha of
phosphorus, 40,000-55,000 plants/ha for Zones 1-4
and 70-140 kg/ha of N, under40 kg/ha of phospho-
rus, 35,000-55,000 plants/ha for Zone 5.
Table 5. Adjusted yield (ton/ha) increase of
Plan-like strategy over conventional strategy,
under farmer circumstances, average 1971-73
Zone Net yield increase
a/ Average yields for the conventional strategy
were subtracted from average yields for the re-
commended strategy (see Table 4) and the on farm
maize cost of additional fertilizer for each year was
Table 6. Number of farmers and area receiving credit for
maize through Plan-associated agencies.
Farmers a Area b
Year No. Percent Ha. Percent
1968 103 0.2 76 0.1
1969 2,561 5.9 5,838 7.3
1970 4,833 11.1 12,601 15.8
1971 5,240 12.1 14,438 18.0
1972 6,202 14.3 17,533 21.9
1973 7,194 16.6 20,604 25.8
1974 8,159 18.8 26,351 32.9
1975 8,701 20.1 28,140 35.1
a/Total number of farmers estimated at 43,300. b/Total maize area
estimated at 80,000 ha.
Table 7. Percentage of plots in annual general survey
receiving different levels of nitrogen, phosphorus and
reporting different pre-harvest plant densities, Zones 1-4.
1971 1972 1973 1974 1975
low 46.0 42.4 42.4 32.8 35.6
medium 18.5 12.4 19.4 17.6 16.1
high 35.5 45.2 382 49.6 48.3
low 51.8 47.5 62.5 64.1 56.4
medium 5.3 11.3 7.6 6.1 5.4
high 42.9 41.2 29.9 29.8 38.2
low 43.9 22.6 33.3 32.8 28.2
medium 29.6 32.8 32.7 29.8 37.6
high 26.5 44.6 34.3 37.4 34.2
a/ low, 0-50 kg/ha; medium, 50-80 kg/ha; high, over 80 kg/ha;
b/ low, 0-20 kg/ha; medium, 20-30 kg/ha; high, over 30 kg/ha;
c/ low, 0-30,000 plants/ha, medium, 30,000-40,000 plants/ha; high,
over 40,000 plants.
variables related to credit lists.
While it can be expected that the number of farmers and
the area in maize varies from year to year, no estimates of
annual values are available. They were assumed to be con-
stant, at an estimated 43,300 farmers and 80,000 ha. of
maize, in calculating the percentages of Table 6. The num-
ber of farmers must vary less year by year than does the
area in maize. Given this, it is likely that changes in partici-
pation are better reflected by the percentages related to
farmers than to those related to area in maize.
Looking now at the second measure of participation,
that based on use of the elements of the recommended
technology, Table 7 represents the developments from
1971 to 1975 for Zones 1-4 and Table 8 for Zone 5. The
data are from the annual general survey of yields. Fertilizer
rates and plant densities used to fix limits on categories are
arbitrary and parallel those set by Plan staff in their reports
[2, 4 ]. They differ slightly, except for the high category of
Tables 7 and 8, from the Plan technology of Table 4.
The high category includes a number of plots receiving
very high levels of nitrogen. This usually occurred on very
small plots and sometimes exceeded 300 kg/ha. These ob-
Table 8. Percentage of plots in annual general survey
receiving different levels of nitrogen and reporting different
pre-harvest plant densities, Zone 5.
1971 1972 1973 1974 1975
low 33.3 12.0 15.8 31.6 16.4
medium 18.2 28.0 21.0 0.0 8.9
high 48.5 60.0 63.2 86.4 74.7
low 72.7 64.0 15.8 21.0 34.2
medium 18.2 36.0 47.4 63.2 45.6
high 9.1 0.0 36.8 15.8 20.2
a/See table 7.
servations, included in High in Tables 7 and 8, were ex-
cluded from Table 4 because of the notable effect they
would have had on the average net yields presented there.
The next chapter will show that applications of nitrogen
above, say, 120 kg/ha had little effect on yields. Why, then,
did some farmers apply so much? These were usually cases
in which small quantities of fertilizers were applied to very
small plots. The quantities involved were so small that cau-
tion in their use promised only minimal absolute gains.
If participation in Plan Puebla is measured in terms of
the application of plant nutrients and of plant densities, the
data of Tables 7 and 8 evidence changes in participation
since 1971. More tenuous comparisons can be made with
data emerging from the 1967 survey, but they are not
Looking first at nitrogen, and thinking of participation
as being related to the High category, participation in-
creased by 36 percent in Zones 1-4 and by 54 percent in
Zone 5. Similarly, using plant density as the measure of
participation, the increase was 29 percent in Zones 1-4 and
over 100 percent in Zone 5.
With somewhat more stringent qualifications the number
of plots on which recommendations can be said to be fol-
lowed shrinks appreciable. Starting with nitrogen as the
major element of the new technology, and recalling that
recommendations range from 110-130 kg/ha in Zones 1-4
and are 100 kg/ha in Zone 5, define an adopter as any
farmer who reports on at least oneplot more than 100
kg/ha of nitrogen per hectare in Zones 1-4 and more than
90 kg/ha in Zone 5. Second, define adoption in terms of
nitrogen, as above, and plant density at harvest of more
than 40,000 plants per hectare for Zones 1-4 and 40,000
plants in Zone 5, both on at least one plot. Recommenda-
tions for both are a seeding rate of 50,000 plants. Finally,
define adoption in terms of nitrogen and plant density as
above and add phosphorus of over 40 kg/ha, in Zones 1-4,
all on at least one plot. Recommended levels vary from 50
to 60 kg/ha in Zones 1-4 and no phosphorus is recom-
mended in Zone 5.
With these definitions of adoption the proportion of
adopters are given in Table 9 for 1971 and for 1975. Data
limitations make it impossible to include 1967, 1968 and
1969 and the data source for 1970 differs from that of
1971 and 1975, hence the presentation of 1971 and 1975.
While comparison with the first years of the program
must be speculative, it is probable that few farmers fit the
second description of adoption (nitrogen and density) in
1967 and 1968 and it seems unlikely that more than 10-15
percent fit the first definition in 1968. Less than 2 percent
used over 100 kg/ha of nitrogen in 1967, but this was an
acknowledged bad year.
If impressions concerning 1968 are correct, then adop-
tion of nitrogen in Zones 1-4 between 1968 and 1971 was
substantial. The number of farmers applying 100 kg/ha of
nitrogen nearly doubled, from 5600 thought to be doing so
in 1968 to an estimated 11,000 in 1971. From 1971 to
1975 the increase in the absolute number of adopting farm-
ers was roughly the same but the relative change was lower
than in the first four years. With adoption measured in
terms of nitrogen application and plant density the absolute
number of adopters is smaller and the year to year changes
are also smaller. Adding the third element to the definition
reduces both percentage changes and absolute changes dras-
tically. With adoption defined in terms of all three ele-
ments, only 6.7 percent of the farmers, roughly 2500 indi-
viduals, had adopted the recommendations by 1975.
In Zone 5 it appears that fertilizer use was more wide-
spread in 1967 than in Zones 1-4. Although Plan activity
started in Zone 5 only in 1970, by 1971 42 percent of the
farmers surveyed reported using over 90 kg/ha of nitro-
gen. The proportion increased to 67 percent, which is to
say an additional 25 percent of the Zone's farmers, by
1975. Almost equally notable was the increase in the
number of farmers reporting over 40,000 plants atharvest
time. From zero in 1971 this increased to 13 percent of the
Zone's maize plots by 1975. Defining adoption in terms of
nitrogen use and plant density, 13.3 percent of the sur-
veyed farmers were adopters. Extrapolating to the popula-
tion this is over 700 farmers.
What conclusions can be drawn from the data of the
annual surveys? They indicate that only a small proportion
of the area's farmers are following Plan recommendations
completely, even given the apparent potential profits and
eight years of Plan efforts. For Zones 1-4 only 6.7 percent
of the plots featured all three major dimensions. Mean-
while, in Zone 5 only 13.3 percent of the surveyed plots
had the two elements of the recommended technology.
And these are not exactly the recommendations of the Plan
but count those employing somewhat less intensive prac-
Table 9. Percentage of plots by region receiving part or all
of the recommended technologya in 1971 and 1975.
Nitrogen Nitrogen, density,
Year Nitrogen and density and phosphorus
Zones 1 4
1971 29.1 7.4 4.2
1975 34.5 13.4 6.7
1971 42.4 0.0 0.0
1975 67.1 13.3 13.3
a/ Recommended technology is defined in terms of the following
three elements: Nitrogen-over 100 kg/ha in Zones 1-4, over 90
kg/ha in Zone 5; Density-over 40,000 plants/ha at harvest time in
all zones;-Phosphorus-over 40 kg/ha in Zones 1-4, zero or more
in Zone 5.
tices. With the recommendations themselves as lower
bounds, the adoption rate shrinks further.
The discussion of the chapter describes an apparent
anomaly. On the one hand Plan recommendations appear to
promise significant increases in average profits; this is evi-
dent from the researchers' trials and, more importantly,
from the experiences of farmers. On the other hand, farm-
ers are not adopting the recommendations in the measure
that might be expected, given their apparent profitability.
1. This section uses information from .
IV. FACTORS IMPEDING THE DIFFUSION OF PLAN TECHNOLOGIES
Before looking at factors which might have restricted the
spread of Plan Puebla recommendations it is worthwhile to
look again at the assumptions on which the analysis rests.
These assumptions, set out in the preface, are that: 1)
farmers are purposive in their behavior, and, more specific-
ally, that they are income seeking risk averters; 2) that
farmers are sensitive to the nuances of the environment in
which they farm; and 3) that they are reasonably efficient
in managing the resources at their disposal. This is not to
suggest that peasant farmers are agricultural savants. It is to
say that traditionalism itself rarely restricts the spread of
This view of the farmer signals the elements which might
be intervening in diffusion of new technology, i.e. in farm-
ers following Plan recommendations. The discussion which
follows reconsiders the assumptions, then looks at access to
information and to inputs, profits and risks. In considering
these points, emphasis will be given to official institutions,
drawing on the work of Heliodoro Dfaz , to profits and
risks, drawing on the work of Edgardo Moscardi [7 ], and
to how the latter are influenced by the opportunity cost of
family labor, drawing on the work of Manuel Villa Issa 18 1
A final section will deal with the sense in which diffusion
has, in fact, been impeded.
On the Behavior Assumptions
This discussion takes as given that purposive behavior is a
characteristic of small farmers and that they are sensitive to
the nuances of their farming environment. This is not to say
that cause and effect are understood to the degree that they
might be in the modern world, but farmers certainly know
well the agricultural cycles and the implications of aberra-
tions for diseases, insects, choice of variety, and so on.
What needs to be discussed further is the assumption
about the ends to which purposive behavior is directed. The
view presented here is that first income and then risk aver-
sion are most influential among farmers' ends.
The literature citations which could be adduced to
support the assumption that farmers seek greater incomes
are legion. Certainly among economists income is regarded
as the primary element influencing behavior. There are, of
course, those who argue that income's role has been vastly
overstated and that social elements play the dominant role
in shaping farmer response to their environment.
The position taken in this paper is nicely summed up in
the following assertion by George Foster, cultural anthro-
pologist: "In earlier chapters examples were given showing
how social and cultural factors have caused people to fore-
go economic gain:... Yet in the final analysis these atti-
tudes appear to be delaying or holding actions, rather than
a definite barrier. Sooner or later the economic pull seems
certain to out-weigh other factors" [9, pp. 150-1 1.
The influence of risk aversion on farmer behavior is less
certain. The argument has a strong intuitive appeal i.e. as
average incomes approach subsistence levels then, in the
absence of facilities for borrowing against future income,
survival requires that one give heavy emphasis to achieving
Table 10. Percentage of surveyed plots receiving one, two,
or three elements a of the recommended technology by
extension zone in 1975.
element Zone 1 Zone 2 Zone 3 Zone 4 Zones 1-4 Zone 5
N 37.5 45.2 23.3 30.0 34.9 67.1
N + D 16.0 12.0 3.8 17.5 13.4 13.3
N+D+P 10.0 6.1 0.0 4.0 6.0
a/ Nitrogen-over 100 kg/ha in Zones 1-4, over 90 kg/ha in Zone 5;
Density-over 40,000 plants/ha; Phosphorus-over 40 kg/ha in
yearly or even seasonal subsistence requirements with high
probability, placing less emphasis on longer run averages.
This emphasis on the near future occurs because, without
access to vehicles for transferring income from period to
period (e.g. assets or loans) income below subsistence re-
quirements in one period can make contemplation of long
run averages academic.
On an empirical note, interviews with a sample of farm-
ers from the Puebla area reported in 110 ] are consistent
with this intuition. Nineteen farmers selected at random
from a larger group of non-adopting farmers were asked
why they didn't follow Plan recommendations. Seventeen
responded with answers indicating the unwillingness to go
into debt when the weather, hence the resulting produc-
tion, is uncertain. Empirical studies undertaken in other
countries, e.g. [11 ] and [12 ] support the view that farm-
ers are risk averters.
It will, then, be taken as given that the behavior of farm-
ers is shaped by income and risk, not completely deter-
mined by these variables, but in large measure conditioned
by them. It will also be taken as given that farmers are
sensitive to their environment. The degree to which farmers
are efficient in managing scarce resources must vary from
farmer to farmer, some doing so quite efficiently, others
not so well. On balance, we assume that farmrnrs are reason-
ably effective in allocating resources.
Access to Information and to Inputs
Provision of information and inputs to the farmers of
Puebla is a function of a group of institutions operating in
the area. Information is the responsibility of Plan staff, of
extension agents, and of the technical staff of the official
credit agencies. Access to inputs comes through a private
fertilizer dealer and through government banks. These insti-
tutions and their effect on the diffusion of the recom-
mended technology is the subject of this section.
Some have argued that a major factor impeding the dif-
fusion of new technology is inadequate service from the
institutions responsible for supporting agriculture. In the
case of Plan Puebla, this view is most cogently argued by
Diaz. The motivating hypothesis of his study is "that the
very roots of backwardness are found in the dysfunctional
institutional structure which serves agriculture in Mexico".
'6, p. 4].
In treating institutional impediments to Plan progress
Diaz details relationships with the Plan staff, relationships
between the staff and other service institutions, and be-
tween both staff and agencies and farmers.
In examining farmer experience in acquiring services
from Banco Agricola and Banco Ejidal, Diaz reports that:
1) procedures for obtaining and repaying loans were
cumbersome and time consuming, 2) promised inputs
often arrived late and 3) little technical assistance was
offered to clients by the banks. In functional terms this
meant that the flow of information through official chan-
nels was inadequate, that the cost of obtaining credit
through official sources was too high, and that returns to
fertilizers were sometimes reduced by late deliveries.
With the advent of the Plan, efforts were made to im-
prove the services of the banks and of the private fertilizer
dealer. These efforts focused on speeding the processing of
loans and repayments, ensuring timely delivery of promised
inputs from the banks, and improving the capacity of bank
technical staff to counsel farmers on the use of modern in-
A significant measure of success accompanied these
efforts as the banks added staff for processing loans, im-
proved the working conditions of technical staff, and waged
successful campaigns to improve deliveries. One manifesta-
tion of the improved services offered to farmers was their
vastly improved repayment of loans they profited from
the loans and wanted to maintain them. By 1973 both
Banco Ejidal and Banco Agricola were enjoying over 90
percent repayment rates on loans made to Plan participants
as compared with repayment rates of 40 percent for Banco
Ejidal prior to the Plan's initiation.
A second function provided by Government for the
farmers of Puebla is agricultural insurance. Farmers criticize
this service because the charges, even though heavily sub-
sidized by Government, seem high, the coverage seems
limited and, indeed at times, capricious. While attempts
were made to introduce new formats for self insurance and
other innovations, none of these were deemed acceptable
by those administering the insurance program. Their re-
sponse, coupled with regulations which tie insurance to the
type of bank loans most commonly made to the farmers,
meant that whatever problems characterized insurance at
the outset were still evident several years into the project
Diaz also comments on the adequacy of efforts to dis-
seminate information about the program to area farmers.
His assessment [6, p. 379 ] is that the change of emphasis
of extensionists from demonstration and field days to loan
and payment processing had a negative influence on dif-
fusion of the technology. A second critic of Plan Puebla
questions the efficacy of the communications program
[13 1, arguing that too few farmers are aware of even the
existence of Plan Puebla. In this same vein, Avila Dorantes
[14, p. 106 ] argues that the primary reason that partici-
pants in the Plan, i.e. those appearing in Table 6, are them-
selves not following recommendations exactly is that they
do not know what is being recommended, especially as re-
lates to seeding density. He concludes by presenting the
possibility that farmers might will have heard of the recom-
mended densities but were unconvinced of their utility and,
being unconvinced, have forgotten that recommendation.
Diaz summarizes his attitude towards the role of institu-
tions by saying that, even after the changes induced
through the efforts of Plan staff, ". the existence of insti-
tutional problems have prevented the project from reaching
the vast majority of campesinos inhabiting the region". [6,
p. 496 ]
It is evident that the services offered by institutions
charged with supplying Pueblafarmers could be further im-
proved in particular that the time required to obtain and
repay loans could be shortened, that input delivery could
be made more timely, and, that Plan staff could relate more
closely to farmers. It is also likely that such improvements
would lead to higher rates of adoption of Plan recom-
mendations as they would increase awareness of the Plan's
advantages and reduce the cost (essentially the value of the
time spent in dealing with administrative issues) of follow-
ing the recommendations. Even so, however, it does not
appear that these institutional shortcomings are the major
impediments to more widespread farmer acceptance of Plan
The argument to support the proceeding assertion rests
on three elements. The first Is the marked difference in the
adoption of Plan recommendations among the five zones of
the Plan region. (See Table 10). The second is an assump-
tion that the services offered by supporting institutions,
whether adequate or not, did not vary from one zone to
another in a way consistent with the variation seen in Table
10. The third element is that area farmers have access to
information and to inputs through sources which are not a
part of the formal apparatus of the project.
If institutional shortcomings are taken to be the primary
factor preventing the region's farmers from following Plan
recommendations, then they must also be the principal
cause of the differences in input.use among the zones. But
there is no evidence that institutional services varied from
zone to zone in a way consistent with Table 10's differ-
ences among zones. Indeed conversation suggests that Zone
5 might well have received less attention, had less access to
institutional services, than did Zones 1-4. Certainly research
and distribution of information was initiated later in Zone
5 than in the other zones.
It might be argued that the differences among zones in
1975 emerged from initial differences in, e.g. nitrogen use
and that these were maintained by institutional restrictions.
Data by zone from the 1967 survey are not at hand but, as
less than one percent of the surveyed farmers reported
using more than 100 kg/ha of nitrogen, the differences
among zones were obviously quite small in 1967.
Turning now to the third element it must be recognized
that farmers do not look only to officialdom for informa-
tion and inputs. Much of the information flow on new tech-
nology is from farmer to farmer. And it is certainly evident
that inputs can be acquired without resort to official chan-
That last point is graphically demonstrated by compar-
ing use of inputs through official sources with overall use of
inputs. For 1975, as a measure of farmer access to inputs
through official sources, Table 6 reports that 20.1 percent
of the farmers were fertilizing 35.2 percent of the maize
area through Plan associated credit sources, principally
through the fertilizer distributor and the two government
banks. Turning to the 1975 yield survey, Table 9 implies
that over 45 percent of the maize area was fertilized with
over 90 kg/ha of nitrogen. More dramatically, the 1975 sur-
vey implies that 89.3 percent of the maize was receiving
some fertilizer. With 35.2 percent of the maize area ferti-
lized through official channels and 89.3 percent of the
maize fertilized, there is clearly a great deal of fertilizer
bought and applied independently of the official institu-
If differential access to services does not explain Table
10's differences in adherence to the recommendations, then
it is unlikely that institutional services are the primary fac-
tor accounting for the overall pattern of adoption charac-
terizing the region. There must be other factors explaining
that pattern. This is not to say that institutional short-
comings have not prevented some farmers from taking up
Plan recommendations, they have certainly played some
role in restricting their adoption. Making them the principal
restriction, however, is questionable, leaving too much un-
explained viz. the interzonal differences and input use
exceeding Plan sponsored use. What is required is an expla-
nation that is consistent with low rates of acceptance of
Plan recommendations and with access to information and
inputs through unofficial channels.
It was posited earlier that farmers are profit seekers. The
data of Tables 3 and 5 suggest that substantial gains in
adjusted yield accrue to those who follow Plan recommen-
dations rather than the conventional strategies. But it must
be remembered that these are not the only strategies open
to the farmers. They can, for example, use an intermediate
strategy with less nitrogen, less phosphorus and a lower
plant density than is recommended but more than with the
conventional strategy of 1967.
Table 11. Average farmer yields in 2 years for two plant
densities with nitrogen levels roughly constant. a
Avg. year Better year
Yield Nitrogen Yield Nitrogen
Density b kg/ha kg/ha kg/ha kg/ha
Medium 3136 110 3017 d 120
High 3116 120 3406 115
Medium 2933c 114 3179 108
High 2500C 102 4166 107
a/ Only those plots receiving 90-150 kg/ha of nitrogen were
included in calculating the averages in Zones 1-4 and only those
receiving 80-150 kg/ha in Zone 5. b/ Medium-over 35,000 to
40,000 plants/ha at harvest. High-over 40,000 to 60,000 plants/
ha at harvest, c/ Less than 4 observations. d/Two observations of
20 were eliminated because of disastrously low yields.
Moscardi [7 ] estimates response surfaces for two soils
types in the Puebla area. One is for the deep soils of Popo-
catepetl and the other for the soils of La Malinche. The
deep soils of Popocatepetl dominate Zones 1 and 2 and
account for about 25 percent of the Plan area's maize. The
soils of La Malinche represent a bit over 25 percent of the
area's maize and make up virtually all of Zone 5.
The data on which the response surfaces are based are
from 1967 through 1971 for the first and for 1967 and
1969 through 1971 for the second. Plan investigators used
the same data in generating recommendations but did so on
an annual basis and for each site. Moscardi pooled data over
years and over sites. Pooling the data, he also included soil
variables measuring organic matter, soil phosphorus, and
soil acidity. Moscardi examined three models commonly
used in response surface analysis quadratic, square root,
and exponential. For exponential models he used ridge re-
Plant density. Looking first at plant density in the deep
soils of Popocatepetl, Moscardi related yields to seeding
densities. For the range between 40,000 and 60,000 plants
he concluded that, other things held constant, yield is
independent of density. For each model, the analysis shows
a slightly negative relationship but with low statistical sig-
nificance. [7, p. 87 ].
In Zone 5 on the soils of La Malinche the results were
essentially the same. Moscardi again found a slight negative
relationship but with low statistical significance [7, p. 87].
In another study of the soils of La Malinche, Hernandez
[ 15, p. 257 ], presents graphs showing a modest increase in
yield as seeding density varies from 25,000 to 50,000
plants. The data are from 1971. For nitrogen constant at
100 kg/ha the estimates for his model suggest an increase of
some 200 kilos of maize, from 3800 to 4000 kg/ha, as seed-
ing density changes from 25,000 to 50,000 plants per ha.
These results suggest that the gains from seeding more
than 40,000 plants/ha are slight, if indeed there are gains at
all, in any given year. It was also found that densities of less
than 35,000 plants/ha contribute to lower yields.
Even so, it is likely that in good years yields will be
higher, other things equal, where density is higher. Poorer
Table 12. Estimates of yields (ton/ha) and of adjusted
gross revenue per hectare a for two soils types at various
levels of fertilizer use.
Soils of Popocatepetl Soils of La Malinche
Level Fert/ha Yield Valueb Fert/ha Yield Value
Profit maximizing 190-21-0 4.30 1894 74-5-0 2.54 1350
0.75 maximizing 140-16-0 3.85 1864 55-4-0 2.40 1349
0.50 maximizing 100-11-0 3.42 1792 37-2.5-0 2.20 1314
0.25 maximizing 50-6-0 2.60 1521 18-1.2-0 1.89 1200
a/ Gross revenue is the farm value of production less the cost of
applying fertilizer and of harvesting, shelling, and transporting
maize to market as estimated in [5, p. 379J. b/ Mexican pesos.
years, on the other hand, might well lead to the reverse.
This idea is supported in Table 11, where a good year is
compared with an average year. Yields with plant densities
at harvest time of 35,000 40,000 plants are compared
with those of 40,000 60,000 plants. Nitrogen and phos-
phorus are roughly constant, with the same ranges used for
each year and each plant density. For the better year, yields
are appreciably higher with higher density. For the average
year, yields are roughly the same. In a poor year, like 1967
for example, it could be expected that yields for lower den-
sities would exceed those of higher densities. In these
examples, with fertilizer use roughly constant, yields are a
good measure of profits.
The studies of Hernandez, for one year, and Moscardi,
over several years, suggest little if any advantage to seeding
densities greater than 40,000 plants/ha. These studies are
based on research done on farmers' fields. Evidence from
farmers' experiences suggest that, if weather is good, higher
densities pay, if weather is poor, lower densities pay. This is
treated again below. Other evidence points to marked yield
declines when densities fall below 35,000 plants.
In 1975, some 56 percent of the fields in Zones 1-4 and
52 percent in Zone 5 had plant densities at harvest between
30,000 and 45,000. Only about 15 percent had densities
over 45,000/ha. In the Plan area, seeding 50,000 plants
should, on the average, leave more than 45,000 plants at
With the evidence at hand, it is not surprising that few
fields are planted at the recommended seeding densities of
50,000 plants. There might well be little benefit on the
average from doing so and considerable loss could result in
the poor year, when yields are low anyhow. The income
seeking risk averter might well decide that seeding rates
slightly below the recommended level but well above
30,000 plants/ha are preferred.
Fertilizer use. Marked differences among zones in ferti-
lizer use were shown in Table 10 of the proceeding section.
These differences are especially evident between Zone 3
and Zone 5, where 23 percent and 67 percent of the plots
received approximately the recommended levels of
nitrogen. Differences are also evident in comparing Zone 5
with Zones 1-4, where 35 percent of the plots receive
roughly the recommended levels of nitrogen.
For insights into potential profit from the recommended
fertilizer strategies we can again turn to the work of Mos-
cardi [7 1 Plan recommendations for the deep soils of
Popocatepetl depend on planting dates but are essentially
130-40-0. Moscardi [7, p. 83, 124 ] estimates that the
average profit maximizing recommendation is 190 kg of
nitrogen and 21 kg of phosphorus for seeding densities be-
tween 40-60,000 plants and with soils variables at average
values. Leaving aside the level for the moment, an interest-
ing aspect of the result is the vastly higher ratio of nitrogen
to phosphorus. While this is on the order of 3 to 1 in the
recommendation for deep soils of Popocatepetl, Moscardi's
results show a ratio of 9 to 1.
The two response surfaces estimated by Moscardi were
used to consider the question of what might happen to
yields and to adjusted gross revenue with lower levels of
fertilizer use. It is not uncommon in response surface analy-
sis that, over the high part of the range for inputs, signifi-
cant reduction in inputs leads to small reductions in adjust-
ed gross revenues. Table 12 shows this to be true for data
from the deep soils of Popocatepetl and from the soils of
La Malinche in Zone 5 as well.
For the deep soils of Popocatepetl, fertilizer use can be
reduced to 75 percent of the average profit maximizing
level while adjusted gross revenue declines by only 2 per-
cent. With half the profit maximizing fertilizer use, the
reduction in adjusted gross revenue is less than 6 percent.
The same comparison applied to the soils of La Malinche is
even more dramatic less than 1 percent and less than 3
percent reduction in adjusted gross revenue for reductions
of 25 and 50 percent in the use of fertilizer.
Two questions can be asked here. What happens to the
conclusion that substantial reductions in input use imply
small reductions in adjusted gross value of yield, if other
values for soils variables are substituted? A different set of
values, representing one of the Zone 5 experimental sites,
was substituted in the estimating equation for the soils of
La Malinche. The estimated profit maximizing values of
nitrogen and phosphorus increased to 131 and 7 kg/ha re-
spectively. But a reduction of 50 percent in fertilizer use
reduced the adjusted gross value of yield by only 2 percent.
For this case, the weak relationship between fertilizer and
adjusted gross revenues at high levels of fertilizer use is not
a consequence of the values assigned to the soils variables.
The second question concerns the extent to which the
result, i.e. that small changes in adjusted value of yields
accompany substantial changes in fertilizer use, is a direct
consequence of the use of an exponential function in fitting
the fertilizer/yield data. Said a different way, would the
same result emerge were a quadratic function used? In this
particular case, a quadratic function was estimated with
similar results. This need not always be true as different
functional forms might well show different results.
This sensitivity of results to the choice of functional
form and the difficulty of properly accounting for the
myriad of elements influencing the relationships between
yields and inputs has led some to argue that the precision
manifested by this kind of analysis is pretentious. That view
is developed and simpler procedures for making recommen-
dations are presented in [16 ] It also underlies the slope/
plateau approach to recommendations seen in [17 ].
And what of the farmer data, what do farmer yields net
of fertilizer costs show? Data for 1971, 1972, and 1973
were sorted into groups for Zones 1-4 and for Zone 5.
(Data for 1974 was omitted because an extraordinarily
early frost affected many plots severely. Yield data for
1975 were not available at the time of writing). For Zones
1-4, the groupings were: 1) plots receiving 120 to 160
kg/ha of nitrogen and 40 to 70 kg/ha of phosphorus along
with 35,000 to 60,000 plants, 2) plots receiving 90 to 120 ing levels of nitrogen occurs within a region of 24,000 ha.
kg/ha of nitrogen and 20 to 50 kg/ha of phosphorus along made up of one soil type [7, p. 118 1
with 35,000 to 60,000 plants at harvest, and 3) plots receiv- Moscardi's evidence of variability is supported by the
ing 0 to 50 kg/ha of nitrogen, 0 to 25 kg/ha of phosphorus data of Table 13. Each of the technologies in each of the
and 15,000 to 35,000 plants at harvest. For Zone 5 the regions manifests a substantial amount of variability. This is
groups were plots receiving: 1) 101 to 160 kg/ha of nitro- the result of weather differences from place to place and
gen along with 33,000 to 60,000 plants, 2) plots receiving year to year and of the heterogeneity of the agronomic cir-
80 to 100 kg/ha of nitrogen along with 33,000 to 60,000 cumstances within regions. It is this heterogeneity which
plants at harvest time, and 3) plots with less than 50 kg/ha prohibits viewing a given frequency distribution as if it rep-
of nitrogen and with 15,000 to 33,000 plants at harvest resented a probability distribution of yields open to any
time. Phosphorus is not. recommended in Zone 5. Setting given farmer within a region.
the lower bound at 33,000 plants, rather than 35,000 as in Nonetheless it is interesting to notice that, for each re-
Zones 1-4, made several more observations eligible for gion, the conventional technology had the largest propor-
inclusion and the shortage of observations in Zone 5 made tion of observations at very low adjusted yields, 39 percent
these necessary. Frequency distributions of yields adjusted of the plots reported estimated adjusted yields below 1200
for the farm cost of fertilizers are presented in Table 13. kg/ha in Zones 1-4 and 46.2 percent in Zone 5. This con-
Average yields are presented in Table 14. Yields adjusted trasts dramatically with intermediate technology plots, 5.5
for the farm cost of fertilizer are in Table 15. percent below 1200 kg/ha in Zones 1-4 and none in Zone 5.
Before looking at farmers' yields and their relation to A weaker assumption, which would still permit the same
the diffusion of various technologies, certain caveats must statements about the consequences of changing plant den-
be considered. If it could be assumed that each grouping, sity or fertilizer use for average yields, is that the observa-
i.e. Zones 1-4 and Zone 5, makes up a completely homo- tions included in the calculations proportionally represent
generous region then it could be said, e.g. from Table 11 each region's micro-environment. This is clearly unlikely
that in the average year those Zone 1-4 plots with over even if representativeness characterized the larger sample
40,000 plants at harvest and yields of 3116 kg/ha would from which the included observations were selected.
have had yields of 3133 kg/ha had density been 35-40,000 Looking now at the average adjusted yields of Table 15,
plants or, from Table 14, those Zone 5 plots using inter- how well do each of the technologies fare within the re-
mediate levels of fertilizer with yields of 3133 kg/ha would gion? In Zones 1-4, both the intensive and the intermediate
have had yields of 3523 kg/ha with larger applications of technologies had far greater adjusted yields than did the
fertilizer, conventional technology on the average. The intermediate
It is known that each region is not homogeneous. As a technology averages slightly higher than did the intensive
measure of the variability of Zone 5, Moscardi employed technology. A profit seeking farmer with average plots
sets of values for soil variables, one for each of the 20 sites would choose the intermediate technology. Within Zones
on which experiments were conducted in the Zone 5 soils 1-4, however, some farmers fared better with the intensive
of La Malinche. He then solved for the estimated profit technology than did others with the intermediate strategy,
maximizing level of nitrogen use. The estimated level rang- e.g. see the frequency distribution in Table 13 for adjusted
ed from 36 kg/ha of nitrogen at one site to 131 kg/ha at yields over 3600 kg/ha. Assuming that 1971-73 were repre-
another. All of this variability in estimating profit maximiz- sentative years, what all of this suggests is that, while some
Table 13. Frequency distribution of average adjusted maize yields a for
three maize technologies in the area of Plan Puebla.
Yield Zones 1-4 Zone 5
kg/ha Conventional Intermediate Intense Conventional Intermediate Intense
0- 600 14.8 8.0 15.4
601-1200 24.2 5.5 1.6 30.8 2.4
1201-1800 32.6 8.3 14.2 23.1 11.8 4.8
1801-2400 14.8 19.3 20.6 15.4 29.4 14.4
2401-3000 5.3 38.8 27.0 23.6 28.6
3001-3600 5.3 19.3 15.8 15.4 23.5 26.2
3601-4200 3,1 5.5 11.1 11.8 16.8
4201-4800 2.8 1.6 7.2
a/ Yields reduced by maize cost of fertilizer on the farm. b/ Conventional: Zones 1-4
0-50 N, 0-25 P, 15,000-35,000 plants; Zone 5, 0-50 N, 15,000-33,000 plants (per ha.)
Intermediate: Zones 1-4, 90-119 N, 20-50 P, 35,000-60,000 plants; Zone 5, 80-100 N,
33,000-60,000 plants (per ha.) Intensive: Zone 1-4, 120-160 N, 40-70 P, 35,000-60,000
plants; Zone 5, 101-160 N, 33,000-60,000 plants (per ha.).
profit seeking farmers of Zones 1-4 can be expected to use
over 120 kg of nitrogen, over 40 kg of phosphorus, and
over 35,000 plants per hectare, it does not seem likely that
the majority will do so.
For Zone 5 the intensive technology has higher average
adjusted yields than has either the conventional or the
intermediate technology. The profit seeking farmer with
average plots would tend to choose the intensive technol-
ogy. Given the averages of Table 15 and the variability seen
in Table 13, it might be expected that the bulk of the farm-
ers of Zone 5 would turn to the intensive technology.
These conclusions widespread adoption of intensive
technology in Zone 5 and only limited adoption in Zones
1-4 are entirely consistent with the survey data of 1975.
Those data show that 64 percent of the plots in Zone 5
received over 100 kg of nitrogen/ha while only 38 percent
of the plots in Zones 1-4 received over 110 kg of nitro-
gen/ha. (Recall that recommended nitrogen is 100 kg in
Zone 5 and averages roughly 115 kg/ha for Zones 1-4). An
effort was made to relate use of the intensive technology to
adjusted yields within zones of Zone 1-4. While adjusted
yields explain inter-regional differences in the use of in-
tensive technology, they are only roughly related to differ-
ences in use among zones within Zones 1-4. The limited
number of observations for some technologies in some
zones averages based on one or two observations un-
doubtedly contributes to the weak relationship between the
two measures among the zones of Zones 1-4.
There is one final point which needs to be made about
profits and the increase in fertilizer use evident in Table 7
and 8. Since 1971 there has been a steady decline in the
maize price of fertilizer as maize prices have increased more
rapidly than fertilizer prices. Starting at 5.6 kg of maize/kg
of nitrogen in 1971, the maize price of nitrogen had de-
clined to 3.2 kg by 1975. With the assumption that farmers
are profit seekers, this decline in the relative price of ferti-
lizer could, itself, increase the use of fertilizer. Nonetheless,
the rate of increase in the Puebla area has exceeded the rate
of increase in the country's other maize growing areas.
Opportunity cost of labor To this point discussion of
the relative merits of technologies has abstracted from the
possibility that labor the farmers' or his family's or hired
labor has a cost. If labor is costly, in an opportunity
sense or in a direct sense, then the returns from the technol-
ogies must be adjusted to include a cost for whatever labor
is utilized to reflect differences in labor utilization.
The Plan area is characterized by a good bit of industrial
and commercial activity. These two, coupled with work in
farming itself, offer area farmers many opportunities for
off-farm work. In the 1967 benchmark survey the average
income from off-farm work was 75 percent of the average
income from crops [4, p. 7 J In a more recent survey Villa
Issa [8, p. 55 ]reports that 87 percent of those farmers not
following the recommended technology worked off the
farm as compared with 56 percent of those roughly follow-
ing the Plan recommendations. Villa Issa also estimated that
Table 14. Average yearly yields a for two regions, Zones
1-4 and Zone 5, of Plan Puebla for 1971-73 for three
production strategies. b
a/ Yields (kg/ha) by plot were averaged for each year and then the
yearly averages were summed and divided by 3 for each technology.
b/See footnote b, Table 13.
the average non-adopter worked 251 days in 1974 while the
adopter worked an estimated 222 days [8, p. 65 ]. Both
adopters and non-adopters report hiring labor for farm
work, an average of 41 days for non-adopters and 40 days
for adopters [8, p. 82 ].
All of this gives evidence that in comparing the technol-
ogies some charge must be made to represent the additional
labor needed by the intermediate and intensive technolo-
gies. Two parameters are critical to the estimation of this
charge the number of units of extra labor and the cost of
On the quantity of extra labor needed, Villa Issa [8, p.
86-87 ], estimates that what he calls the recommended tech-
nology requires 8.7 more days of labor than does his so-
called traditional strategy in order to affect all of the opera-
tions as well as to harvest and transport the extra grain. In
[4, p. 91 ] the extra labor utilized is estimated at 10.3 days,
after eliminating labor associated with hand shelling. (Ma-
chine shelling is widely available, is inexpensive, and is
In both [4 ] and [8 ] the descriptions of the traditional
technology and of the recommended technology accord
well with the conventional and intensive technologies used
here. It will, then, be assumed that intensive technology
absorbs 8.7-10.3 days of labor more than does the conven-
tional technology. For the intermediate technology, which
requires the same operations as the intensive technology
but with some at slightly lower levels, it is estimated that
two days less labor are required than for the intensive tech-
nology, most of this at harvest time.
Table 15. Average yields, 1971-73, adjusted for the on
farm maize cost of fertilizer a for three production tech-
nologies. b Plan Puebla.
Zone Conventional Intermediate Intensive
Zone technology technology technology
1-4 1570 2668 2447
5 1887 2637 2892
a/ Yields by plot were averaged for each year and then the yearly
averages were summed and divided by three for each technology.
Maize cost of fertilizer at the farm gate was subtracted from each
yield. (Data from 1971-73 yield survey. b/Seefootnote bTable 13.
These estimates relate to days utilized in the field.
Acquiring credit, arranging for delivery of inputs, repaying
loans, and finding out about the new technology might well
take additional time. How much time is difficult to say,
first because the time required to service loans varies from
institution to institution and from farmer to farmer, second
because nearly all of the area's farmers are already using
some fertilizer on maize and a substantial portion are appar-
ently self-financed, and, third, because the acquisition of
knowledge might be unnecessary or fast. In order to recog-
nize that some charge might be necessary for these con-
cepts, it is assumed that one day per hectare is spent in
these activities beyond what is spent in similar activities by
those using the conventional technology.
This all leads to the assumption that the intermediate
technology requires 7.7-9.3 more days of labor and the in-
tensive technology requires 9.7-11.3 more days of labor
than does the conventional technology.
Turning now to the value of the extra time utilized, Villa
Issa, in a private communication, reported that there is lit-
tle seasonal change in wage rates. There are periods in
which farm labor is relatively more scarce, most evidently
at planting and harvest.
A reasonable lower bound on the daily charge for labor
is the amount farmers are paying to hire casual labor.
According to Villa Issa [8, p. 82 ]this was 41 pesos per day
in 1974. A meaningful upper bound is the estimated aver-
age wage received by those farmers working off the farm in
non-agricultural pursuits. Villa Issa [8, p. 58, 63 ] reports
data which permits estimating a rate of 65 pesos per day. In
1974 the estimated price of maize in the Puebla area was
1370 pesos per ton at the farm. Measured in maize, this im-
plies that the lower and upper bounds on the cost of labor
were 29.9 and 47.4 kg per day respectively.
Looking back at the adjusted yields in Table 15, the esti-
mated difference between the average yield for conven-
tional technology and intermediate technology is 1098
while that between intermediate and intensive technology is
- 221. On the average and employing the data for 1971-73
the intermediate technology is preferred to the intensive
technology, even without including an additional charge for
And what about the estimated difference in adjusted
yields between the conventional and intermediate technolo-
gies, was it large enough to compensate for the extra labor?
With the daily cost of labor between 29.9 and 47.4 kg of
maize the extra estimated average yield difference between
the conventional and the intermediate technologies is be-
tween 657 and 868 kg of maize after adjusting for the
maize cost of fertilizer and for the extra cost of labor. The
incorporation of a charge for labor would not dissuade the
income seeking Zones 1-4 farmer from taking up the inter-
For Zone 5, the comparable figures show that the extra
estimated difference in average yields between the conven-
tional and the intermediate and intensive technologies
ranges from 339 to 550 kg/ha of maize and from 469 to
715 kg/ha respectively, after adjusting for the extra cost of
fertilizer and for the extra cost of labor.
Comparing the intermediate and the intensive technolo-
gies the estimated two days of additional labor are associat-
ed with an estimated 255 kilos of maize (see Table 15).
Assuming two extra days of labor and the costs per day
given above, the extra estimated average yield difference be-
tween intermediate and the intensive technology is from
160 to 195 kilos of maize/ha.
The income seeking farmer in Zone 5 operating under
average circumstances, would, then, move towards the in-
tensive strategy even after compensating for the cost of the
extra fertilizer and the extra labor required.
It should be noted that the conclusions here do not
quite accord with those of Villa Issa. His findings are based
on two production functions, one for adopters and one for
non-adopters. Using marginal analysis, he finds that non-
adopters are using roughly that quantity of labor on their
farms such that the marginal returns to labor are equal to
the opportunity cost of labor. Non adopters would not, he
concludes, use the additional labor required by the more
labor intensive strategy.
Two elements differentiate the construction used here
from that employed by Villa Issa. The first is that the yield
differences between our conventional and intensive technol-
ogies, roughly comparable to his traditional and recom-
mended technologies, are larger. He reports a difference in
gross yields of 1208 kg [8, p. 86, 87 ]for 1973 while Table
15 shows a difference of 1630 kg for Zones 1-4 and 1577
kg for Zone 5 for 1971-73.
The second difference is in procedure. Villa Issa bases his
conclusions on production function analysis which attrib-
utes production under each technology to fertilizer and to
labor. The procedure followed here is to account for the
maize costs of the extra fertilizer and of the extra labor and
then to see if the associated yield difference is sufficient to
compensate for these extra costs. The point of departure is
that extra fertilizer requires extra labor so that the two can
be viewed as a joint input. The conclusion is that the in-
come seeking farmer with plots like the average plots re-
ported in Table 14 would find the intensive technology pre-
ferred to the conventional strategy in both Zones 1-4 and
Zone 5 but would find the intermediate strategy preferable
to the intensive technology in Zones 1-4, even after allow-
ing for the extra cost of fertilizer and labor.
Given the observations made in the previous sections, espe-
cially as they relate to yields adjusted for the cost of ferti-
lizer and the cost of labor, what can be said about the im-
pact of risk on the adoption of the recommended tech-
nology? Conclusions about the effect of risk can be closely
related to what assumption is made about how risk aversion
is manifested. A simple and intuitively appealing way to
express risk aversion holds that the farmer wishes to keep
the probability of disaster below some level. This intro-
duces two critical questions: what constitutes a disaster and
at what level should the probability be set. Even without
precise answers to these questions, there are circumstances
under which technologies can be compared and inferences
drawn about relative risk.
Data requirements for the analysis of risk are stringent.
What is needed is a set of observations representing a distri-
bution function, relating yields from a given technology to
probabilities while holding other than random factors con-
stant. Were the agronomic circumstances the same within
regions for the plots providing the data of Table 14 and
were 3 years of observations judged sufficient to represent
the area's weather, the frequency distributions of Table 13
could be used to assess the risks of the three technologies
presented. While acknowledging that these conditions are
not met, the frequency distributions of Table 13 are none-
theless suggestive about the relative risks of the three tech-
Let it be assumed for the moment that the frequency
distributions of Table 13 represent proper probability dis-
tributions. Then what do the distributions imply about
comparative risk? If the disaster level is taken to be 1200
kg/ha, then in each region the conventional technology has
a far greater probability of falling below the disaster level
than has either of the others. For Zone 1-4, a risk averter
would prefer the intermediate technology over the conven-
tional technology for any probability level of disaster but
he would select neither if that probability were below
d.055. The same would hold for the comparison between
the intermediate and the intensive technologies. Setting the
disaster level at 600 kg/ha the intermediate technology is
preferred for any probability level. In Zone 5, with other
things equal, the risk averter will tend to choose the inter-
mediate over the intensive technology if disaster is taken as
1200 kg, and is indifferent between the two if disaster is
taken as 600 kg.
Even given the strong assumptions, assumptions which
almost certainly contradict fact, needed to employ the fre-
quency distribution of Table 13 as probabilities, the data
are still mute with respect to the possible impact of risk in
Plan Puebla. This is because rankings of the technologies
made in terms of risk accord almost exactly with rankings
made in terms of average adjusted profits. In Zones 1-4, the
income seeking average farmer would choose the inter-
mediate strategy. So would the risk averter who wanted the
probability of less than 1200 kg of maize to be less than
0.056 or the probability of less than 600 kg to be less than
0.080. Both criteria indicate the same technology. So too
for Zone 5, roughly. The income seeker would choose the
intensive technology. The risk averter who wanted the
probability of less than 600 kilos of maize to be less 0.01
would do the same. A risk averter unwilling to accept a
probability greater than 0.023 of less than 1200 kg would
Table 16. Average application of nitrogen (kg/ha) to maize
on plots sampled in 1975 for smaller and larger farms by
Farmsa Zone 2 Zones 1-4 Zone 5
Smaller 94 88 108
Larger 84 86 123
a/ For each grouping farms were arrayed by size and each array
separated at the median.
choose the intermediate technology. This seems unlikely,
because this disaster level is high and the probability level is
quite low. Again, the risk averter would choose the same
technology as the income seeker.
The frequency distributions of Table 13 are not proper
probability distributions. They are only suggestive of how
risk averters might respond to the three technologies. In
considering the possible effect of risk averse behavior on
the diffusion of input intensive technology in Plan Puebla,
it can only be said that the table does not give evidence that
knowledgeable risk averters would prefer the conventional
technology. There is, of course, always the chance that
farmers' perceptions of the relationship between yields and
probabilities for a given technology are incorrect. With
experience, perceptions become better approximations of
There is other indirect evidence that risk was not playing
a prominent role in the utilization of fertilizer in 1975.
Were risk a prominent element in shaping fertilizer use, one
might expect to see a relationship between fertilizer use and
farm size within regions. The rationale is that larger farm-
ers, with more assets and income, would be less risk averse
than smaller farmer, hence more disposed to use larger
quantities of fertilizer than smaller farmers.
Simple regressions were made of nitrogen use on farm
size. For Zone 5, the estimated coefficient was positive but
with a t value of less than 1.0 indicating low statistical sig-
nificance. For Zones 1-4, and after eliminating one ultra ex-
treme observation out of 149, the estimated coefficient is
negative but with a t value of less than 1.0, again indicating
low statistical significance. In both cases, explanatory
power of the model is very low, with coefficients of deter-
mination under 0.05.
A simple comparison of fertilizer use among farms of
various sizes is presented in Table 16. First, farms were
arrayed by size for Zones 1-4, for Zone 2 and for Zone 5.
One soil type, deep soils of 'Popocatepetl is dominant in
Zone 2 and a second, soils of La Malinche, is dominant in
Zone 5. For each half of each array the average quantity of
nitrogen used on the surveyed plot was calculated. These
averages are reported in the table. They offer a mixed pic-
ture of the relationship between fertilizer use and farm size.
It is readily acknowledged that farm size and fertilizer
use might be related through factors other than risk, e.g.
larger farms have better access to credits and to information
and have lower transaction costs than have smaller farms. If
operative, these relationships will all tend to be positively
related to fertilizer use. That is to say, large farmers have
better access to credit, greater access to information, lower
transactions costs, and will tend to be less risk averse than
small farmers; each of these considerations tends to permit
or to encourage large farmers to apply more fertilizer than
small farmers. In the case at hand, as evidenced by the data
of Table 16, there is no clear evidence that any of these
factors is stimulating larger farmers to apply more fertilizer
than smaller farmers.
Has Diffusion Really Been Impeded?
Two measures of the adoption of new technology have
been presented. One of these defines adoption in terms of
farmer association with Plan connected credit agencies. On
this measure, 20 percent of the farmers with 35 percent of
the maize had 'adopted' Plan recommendations by 1975.
This measure excludes those who employ the technology
but do not appear on the official credit lists. It includes
some users of Plan connected credit who do not follow the
A second measure describes adoption in terms of what
the farmers are doing with respect to fertilizers and plant
densities. The proportion of adopters depends, of course,
on how adoption is defined. This is the measure of Table 10
or of Tables 7 and 8. The more stringent the conditions the
lower the estimated proportion of adopters. On these meas-
ures of adoption, especially the second when stringently
defined, diffusion has been slow.
The question that emerges here is whether or not either
of these measures is really relevant. Surely the first is not a
proper measure. Interest is in what farmers are doing with
fertilizer and planting densities, not in whether or not they
had official credit.
As defined, the second does not appear to be an appro-
priate measure either. If it could be assumed that there
were some special motives for following the recommenda-
tions exactly, then progress could be measured through the
number of farmers using the recommendations. For exam-
ple, if the existing recommendations could be held to be
optimal for each of the area's farmers, then measuring in
terms of farmers following the recommendations would be
But it is unlikely that the recommendations, even the
system of 16 recommendations, fits a large proportion of
the area's farmers exactly. The heterogeneity in agronomic
circumstances (witness the variability among experimental
sites in Zone 5) and the differences in farmer circumstances
- e.g. in the opportunity cost of capital; in the opportunity
cost of labor because of farm size, off-farm work, and crop-
ping pattern differences; in the distance to markets; or in
managerial ability all combine to make it most unlikely
that one set of recommendations will fit all farmers.
What happens then is that farmers alter recommenda-
tions to the extent that they can so as to make them more
consistent with their own circumstance. Given this, meas-
ures in terms of a strict interpretation of the recommenda-
tion are excessively stringent. Many farmers, as in Puebla,
might find the recommendations too intensive for their cir-
cumstance but will, because of activities of the Plan, move
from conventional to intermediate levels. These farmers
have benefitted from Plan activities, are producing more
maize, and generating more income. The strict interpreta-
tion of adoption would exclude them.
What this suggests is that progress be measured in more
general terms, e.g. in terms of the proportion of farmers
now employing intermediate levels of fertilizer or of plant
density. The data of Tables 7 and 8 are approximations to
this sort of measure. Alternatively, the measure of progress
might be couched in terms of the average use of fertilizer.
This increased from 34 kg/ha of nitrogen in 1967 to over
80 kg/ha in 1975. A third measure is in terms of the pro-
portion of plots with more than 40 kg/ha of nitrogen which
increased from 28 percent in 1967 to 49 percent in 1971
and to 77 percent in 1975.
Each of these measures is more relevant than those de-
rived from credit lists or from those following the recom-
mendations strictly. In this case, all imply a wider diffusion
of new technology, not necessarily the recommended tech-
nology but an improved technology nonetheless.
And all of that heterogeneity alluded to earlier also has
implications for the formulation of recommendations. With
all of the differences which characterize farmers differ-
ences in natural and in economic variables it is just not
likely that average profit maximizing recommendations can
be made for large groups. As agronomic recommendations
become more and more precise, or as circumstances become
more heterogeneous,recommendations fit fewer and fewer
farmers. This suggests that the strategy should aim at for-
mulating agronomic recommendation which will: 1) pro-
mise significant increases in profits for any user but 2) be
less intensive than the optimum levels for many users.
There are two immediate advantages of this strategy.
First, general recommendations of this kind require less re-
search than do recommendations which aim to be precise.
Second, the rate of return for investment in intermediate
levels of inputs will be higher than that for more intensive
technologies, hence more appealing to risk averters and to
those with high opportunity costs. Over time, each farmer
can move towards his own optimum, guided by the knowl-
edge of his particular circumstances and his accumulating
Chapter IV focuses on considerations which might be im-
peding a more rapid diffusion of recommendations. Empha-
sis was given to institutional constraints, to profits with
attention to the cost of labor and to risk as potential fac-
tors forestalling diffusion.
In considering the possible restrictions imposed by
inadequate institutional support it was first noted that
there are sharp differences in the use of inputs among the
zones of the Plan. There is no evidence that the quality of
institutional services varies from zone to zone in a way con-
sistent with the variation in input use. It was concluded
that institutional services do not explain the differences
among zones. It is also evident that farmers had access to
credit, to inputs, and to information through non-official
channels. With institutional factors not responsible for
difference among zones and with farmer access to services
not dependent on official institutions it was concluded that
institutional factors are not responsible for slow rates of
diffusion in the area as a whole. While more effective ser-
vices would probably contribute to more rapid diffusion,
inadequate services cannot be regarded as the principal fac-
tor preventing the region's farmers from following Plan
In considering profits experimental data are marshalled
which suggest that planting densities in excess of 40,000
plants per hectare do not contribute strongly to yields, that
yields adjusted for cost of fertilizer change only slightly (if
at all) when fertilizer use is reduced from intensive to inter-
mediate levels, and that it might be possible to reduce phos-
phorus applications relative to nitrogen applications.
Data from farmer plots for Zone 5 imply that, after
reducing yields to compensate for the farm cost of fertilizer
and for the opportunity cost of labor, the intensive technol-
ogy (one with recommended rates of nitrogen but permit-
ting plant densities down to 33,000 plants/ha at harvest) is
the most profitable technology. In Zones 1-4 an inter-
mediate technology (an average of 90 percent of the recom-
mended nitrogen, 80 percent of the recommended phos-
phorus, and with plant densities down to 35,000 plants at
harvest) is most profitable. In both regions, the conven-
tional technology offered less profits than did either the
intermediate or the intensive technology on the average
even after adjusting for the farm cost of fertilizer and the
opportunity cost of additional labor required. According to
the analysis, the cost of labor, even when valued at average
wages in non-agricultural activities, is not motivating farm-
ers to stay with the conventional technology.
A substantial amount of variation is evident in adjusted
yields for each of the technologies and for both regions.
Using frequency distributions of yields to represent prob-
ability distributions, a risk averting farmer would not
choose the conventional technology over the other technol-
ogies in either region and would be disposed to choose the
intermediate technology in Zones 1-4 and the intensive
technology in Zone 5. This is exactly consistent with selec-
tion based on profit seeking. A second look at the possible
influence of risk, in which fertilizer use is related to farm
size, showed no consistent relationship between the two
and no statistically significant relationship in either region.
On balance it was concluded that there is no evidence in the
1975 survey data that risk is influencing the adoption of
fertilizer. Risk might be inducing farmers to favor lower
planting densities, around 35,000 plants at harvest.
After examining potential profits and risks of the three
technologies in each region it was concluded that Zone 5
farmers would be more likely to use high rates of nitrogen
than would farmers in Zones 1-4, and that few farmers in
Zones 1-4 could be expected to use over 100 kg/ha nitro-
gen. This conclusion emerges from comparison of average
adjusted yields and the variability in yields. It accords
exactly with the 1975 survey data which showed only 38
percent of the Zones 1-4 plots receiving at least the average
recommendation of 110 kg/ha of nitrogen while 65 percent
of the Zone 5 plots were receiving at least the recom-
mended 100 kg/ha.
Two additional points remain. First, the farmgate maize
price of fertilizer has declined substantially since 1971,
from 5.6 to 3.2 kg of maize per kg of nitrogen in 1975.
This has almost surely encouraged profit seeking farmers to
use more nitrogen. Second, it can be asked why 38 percent
of the plots of Zones 1-4 received 110 kg or more of nitro-
gen when intermediate levels were more profitable on the
average. A myriad of explanations come to mind but it
seems likely that the most important lies in the hetero-
geneity of the area. As Table 13 shows, for some plots
amounts in excess of 110 kg/ha were more profitable than
V. REVIEWING CONCLUSIONS
This final summing up treats three themes: the first com-
bines two lessons which Plan experience helped make more
evident; the second relates to conclusions about factors
which have been important in shaping farmer compliance
with Plan recommendations; and the third deals with the
implications of the conclusions for some of the hallmarks
of the Plan.
Lessons Reinforced by Plan Experience
The experience of the Plan testifies to the willingness of
small farmers to change from traditional practices when
given the opportunity and inducement. Tradition, after all,
is nothing more than the embodied experiences of the past,
garnered in a stable environment and interpreted in the
light of the goals and circumstances of the participants.
With changes in their environment and circumstances, the
farmers of Puebla have shown their willingness to change
Plan Puebla's early demonstration of this willingness of
farmers to follow more intensive practices captured the
attention of decision makers in much of Latin America.
Now convinced that traditionalism itself is not a lasting
barrier to change, they have set out to stimulate changes
among the farmers of their own countries. Many such
efforts in Honduras, Peru, Colombia, and in Mexico are
consequences of the early activities in Plan Puebla. The
project, then, has contributed in a substantial way to the
recognition that something can be done to improve the lot
of small farmers.
Beyond this, Plan personnel demonstrated that formal
rules and working rules differed markedly for some of the
institutions charged with supporting the area's agriculture.
This realization led to changes in certain procedures,
changes which improved the services of the institutions.
This experience has sensitized other decision makers to the
possibility that attainable changes in their own working
rules could facilitate flows of information and inputs to
farmers with consequent increases in production.
Farmer Compliance with Recommendations
The preceding chapters focused on factors which might
have been influential in preventing farmers from following
the practices recommended to them. While emphasis was
given to profitability, other factors were also considered.
One of these is the adequacy of farmer access to credit, in-
puts, and information. It was argued that, while improve-
ment in these services has facilitated adoption of recom-
mendations, inadequate services are not one of the principal
barriers to more widespread use of the recommended tech-
A second factor sometimes said to be restricting the use
of Plan recommended technology is the opportunity cost of
labor. Again, while in some cases this might be an impor-
tant consideration, the evidence suggests that it is not play-
ing a significant role in limiting adoption of the recommen-
dations. Yield survey data shows that, on the average, yields
from improved technologies as compared with conventional
technologies are more than sufficient to compensate for the
extra labor required. A third factor, risk, was also consider-
ed, but with inconclusive results.
Differences between the farmers of two sub-regions in
the use of recommended technology was found to be con-
sistent with differences in profits associated with their use.
What emerged from the analysis, then, is the conviction
that profit is the principal factor influencing farmer com-
pliance with Plan recommendations.
Implications for Plan Hallmarks
Turning now to specific attributes of the Plan and potential
modifications, several characteristics are regarded as Plan
hallmarks: experimental work in farmers' fields, research
aimed at precise recommendations, extension work and
arrangements for credit organized through groups of farm-
ers, and an evaluation unit within the staff. These final
paragraphs will focus on precision in recommendations and
The project's emphasis on two-way communication with
farmers has reaffirmed the importance of on-farm research,
combining careful research with farmers' conditions. This
approach is far more likely to produce useful recommenda-
tions than research carried out on experiment stations with
little contact with farmer clients. In formulating recommen-
dations researchers aimed at a high degree of precision,
arguing that small errors can have consequences for yields
and income which, because the farmers are poor, will be
The effort to achieve precision appears to have run afoul
of the region's heterogeneity, heterogeneity which ema-
nates from differing agro-climatic circumstances and from
differing farmer circumstances. The range of values en-
countered for soils variables in Zone 5 is a graphic testi-
mony to the agrodlimatic differences. The likely differences
from farmer to farmer in opportunity cost of capital,
opportunity cost of labor, distance to markets, cropping
patterns, sensitivity to risk and so on distinguish one farm-
er's circumstances from another's. With such heterogeneity
it is unlikely that precise recommendations recommenda-
tions near the optimum for one farmer will fit many of
his neighbors. What is more likely is that a great deal of
variability in input use will emerge as farmers move towards
It is clear that a recommendation for each of the plots of
each of the farmers is not feasible and it is evident that
heterogeneity precludes formulating precise recommenda-
tions with general applicability. This suggests that inter-
mediate recommendations that are below average profit
maximizing levels but high enough to promise substantial
returns, are preferred. Certainly this is what farmers are
doing as they reduce planting densities and fertilizer levels
below recommended levels.
It can be claimed that, if farmers are already making
adjustments, the course now followed suffers no disadvan-
tages. But there are two disadvantages. The first is that pre-
cise recommendations are more expensive to generate than
are more general, less ambitious recommendations. Concen-
tration on intermediate recommendations provides substan-
tial savings, especially in the time of trained agronomists.
The second disadvantage emerges from the implications
for evaluating the project. If the project's progress is meas-
ured by farmer use of precise recommendations and if the
precise recommendations don't exactly fit all of the farm-
ers, then "adoption rates" will be low and evaluation will
be unfavorable. Certainly this is the result of a stringent
evaluation of the data from the 1975 yield survey. This
shows that adoption, defined in terms of the recommenda-
tions, occurred on less than 10 percent of the plots. Meas-
ured in terms of intermediate recommendations, however,
program progress would be more favorably viewed. And
this view would seem to accord better with what has taken
place a sharp increase in the proportion of farmers apply-
ing high levels of fertilizers.
As a final thought, then, while the farmers of Puebla are
using more fertilizer and increasing seeding densities, they
are not moving quickly to follow the Plan's recommenda-
tions exactly. The evidence cited in this monograph sug-
gests that the major factor impeding widespread diffusion
of the recommendations is that they promise little, if any,
increase in profits over intermediate levels of input use but
do require larger expenditures and might involve larger
risks. Farmers are moving toward intermediate levels of in-
put use. There are advantages in concentrating research and
recommendations in that range as well.
1. Myren, Delbert. 1969. "The Rockefeller Foundation Program
in Corn and Wheat in Mexico". Editor, Dr. Clifton R. Wharton.
Subsistence Agriculture and Economic Development Aldine
Publishing Co., Chicago.
2. CIMMYT. 1969. "The Puebla Project 1967-69. Progress Report
of a Program to Rapidly Increase Corn Yields On Small Hold-
ings". Centro Internacional de Mejoramiento de Maiz y Trigo.
3. Winkelmapn Donald. 1975. "Plan Puebla After Six Years". in
Small Farmer Agricultural Study in Developing Nations. Sta-
tion, Bulletin No. 101, Department of Agricultural Economics.
Agricultural Experiment Station, Purdue University, West La-
4. CIMMYT. 1974. "The Puebla Project: Seven Years of Expe-
rience: 1967-1973". Centro Internacional de Mejoramiento de
Maiz y Trigo, El Batin, M6xico.
5. Villa Issa, Luis. 1974. "Adopci6n de Tecnologia Nueva en
Zonas de Temporal: El Efecto del Factor Incertidumbre". T6sis
de Maestr(a, Colegio de Postgraduados, Escuela Nacional de
Agriculture, Chapingo, M6xico.
6. D(az, Heliodoro. 1974. "An Institutional Analysis of a Rural
Development Project: The case of The Puebla Project in
Mexico". Ph.D. Thesis, University of Wisconsin.
7. Moscardi, Edgardo. 1976. "A Behavioral Model for Decision
under Risk among Small Holding Farmers". Ph.D. Thesis, Uni-
versity of California, Berkeley, California.
8 Villa Issa, Manuel. 1976. "The Effect of The Labor Market in
The Adoption of New Production Technology in a Rural Devel-
opment Project. The case of Plan Puebla, M6xico". Ph.D.
Thesis, Purdue University, West Lafayette, Indiana.
9. Foster, George. 1962. Traditional Cultures and The Impact of
Technological Change. Harper and Row, New York.
10. Moscardi, Edgardo. 1972. "Riesgo y Transferencia de Tecnolo-
gia: Estudio para el caso del Plan Puebla". TBsis de Maestr(a,
Colegio de Postgraduados, Escuela Nacional de Agricultura,
11. Kennedy, J. O. S. 1976. "Indifference Curves for Expected and
Focusloss Income: Measurement and Applications in Agricul-
ture. Paper presented at the ADC/CIMMYT Risk and Uncer-
tainty Conference. Centro Internacional de Mejoramiento de
Maiz y Trigo, El Batan, Mexico. (Mimeo).
12. Scandizzo, P. L. and J. L. Dillon, 1976. "Peasant Agriculture
and Risk Preferences in Northeast Brazil: A Statistical Sampling
Approach". Paper presented at The ADC/CIMMYT Risk and
Uncertainty Conference. Centro Internacional de Mejoramiento
de Maiz y Trigo, El Batan, M6xico. (Mimeo).
13. Peters, Jeffrey. 1973. "Factors Hindering More Widespread
Farmer Participation in Plan Puebla and Some Concrete Recom-
mendations for Increasing Participation". Puebla, M6xico.
14. Avila Dorantes, Jose A. 1973. "Evaluaci6n de Factores que Ac-
t6an Como Restricciones en la Aplicaci6n de las Recomenda-
clones del Plan Puebla". Tesis de Maestria, Colegio de Postgra-
duados, Escuela Nacional de Agricultura, Chapingo, M6xico.
15. Hernndez S., Ramiro. 1972. "El Modelo Aproximativo y la
Matriz Experimental Como Factores que Influyen sobre el Ses-
go al Aproximar Superficies de Respuesta a Dos Factores".
T6sis de Maestria, Colegio de Postgraduados, Escuela Nacional
de Agriculture, Chapingo, M6xico.
16. Perrin, R. K., D. L. Winkelmann, E. R. Moscardi, and J. R.
Anderson 1976. "From Agronomic Data to Farmer Recommen-
dations: An Economic Training Manual". Centro Internacional
de Mejoramiento de Maiz y Trigo, Mexico City.
"iMM T T&ENTL..DP'OF f9A C.rUA CNW
S a e G ft
'freo e oiiio Th T ns,
T h e /fsi nm y b i 6 W i ,* r K P
V.. ya ;
it. ho t 4 01 7 2l,'
D o a d ikIm n V
Th dpino eamietd wo-