|
AGRICULTURAL EXTENSION FOR SMALL FARMERS
by
Benedict Stavis
Working Paper No. 3 1979
AGRICULTURAL EXTENSION FOR SMALL FARMERS*
by
Benedict Stavis**
*This paper is part of a broader project entitled "Alternative Rural
Development Strategies," Cooperative Agreement AID/ta-CA-3 funded by
the Agency for International Development, Development Support Bureau,
Office of Rural Development and Development Administration. Obviously,
the views expressed herein do not necessarily reflect the views of the
funding agency.
**Assistant Professor, Department of Agricultural Economics,
Michigan State University, East Lansing, Michigan 48824.
1979
MSU is an Affirmative Action/Equal Opportunity Institution
MSU RURAL DEVELOPMENT WORKING PAPERS
Carl K. Eicher and Carl Liedholm, Co-editors
The MSU Rural Development Working Papers series are designed to
further the comparative analysis of rural development in Africa, Latin
America, Asia, and Near East. The papers will report research find-
ings on community development and rural development in historical perspective
as well as on contemporary rural development programs. The series will include
papers on a wide range of topics such as alternative rural development
strategies; off-farm employment and small-scale industry; marketing problems
of small farmers; agricultural extension; interrelationships between technology,
employment and income distribution; and evaluation of rural development
projects. While the papers will convey the research findings of MSU faculty
and visiting scholars, a few papers will be published by researchers and
policy-makers working with MSU scholars on cooperative research and active
programs in the field.
The papers are aimed at teachers, researchers, policy-makers, donor
agencies, and rural development practitioners. Selected papers will be
translated into French, Spanish, and Arabic. Libraries, individuals, and
institutions may obtain single copies of the MSU papers free of charge and
may request their names be placed on a mailing list for periodic notifications
of published papers by writing to:
MSU Rural Development Working Papers
Department of Agricultural Economics
206 International Center
Michigan State University
East Lansing, MI 48824
U.S.A.
TABLE OF CONTENTS
Page
A. DEFINITIONS AND BACKGROUND OF EXTENSION ACTIVITIES . . 4
B. METHODOLOGICAL PROBLEMS IN EVALUATING EXTENSION PROGRAMS 12
C. POTENTIAL PROBLEMS IN EXTENSION SYSTEMS . . . . .. 21
1. Urban Control . . . . . . . ... . 21
2. Rural Elites . . . . . . . .... ... .. .22
3. Bureaucratic Interests . . . . . . .... .27
D. ISSUES IN IMPROVING EXTENSION SERVICES . . . . .. 29
1. Centralized Management . . . . . . .... .31
2. Decentralization . . . . . . . .... . 40
3. Modes of Communication . . . . . . .... .45
4. Sources of Extension Information . . . . .... .49
E. GROUP ORGANIZATION . . . . . . . . ... . 62
F. CONCLUSION . . . . . . . ... .. .. .77
AGRICULTURAL EXTENSION FOR SMALL FARMERS
Ben Stavis*
September, 1979
Can better agricultural extension programs help reduce world
poverty? This paper answers with a very qualified yes. Agricultural
extension programs are attractive to alleviate poverty because they can
help increase food production, which is needed by the poor; they can
increase rural income by improving productivity of smaller farmers;
they can (but not always do) expand demand for farm labor and thereby
benefit landless laborers; they can be useful in a general improvement of
rural living conditions; and they can, in theory, be targeted on the
needs of specific groups in specific areas.
At the same time the benefits of extension programs are broader
than directly alleviating poverty. They can provide agricultural products
for urban needs and for export to raise foreign trade; they can increase
profits and power for rural elites, and assure stable employment for bureaucrats;
they also can generate well defined projects for foreign donors. Extension
offers to make these changes without violence and without requiring extensive
*I gratefully acknowledge the valuable suggestions which have
come from discussions with my colleagues at Michigan State University,
Carl Eicher, Akhter Hameed Khan, George Axinn, Jim Pease, and Jim Bingen,
and former colleagues at Cornell University, Milton Esman, Norman Uphoff,
and William Whyte. Extensive bibliographic assistance has been pro-
vided by Peter Riley and Alemeneh Dejene. I accept responsibility for
the formulations herein, but all my colleagues may claim credit for any
useful insights.
This paper is part of a broader project entitled Alternative
Strategies for Rural Development, funded by U.S. Agency for International
Development, Development Support Bureau, Office of Rural Development and
Development Administration. Obviously, the views expressed herein do
not necessari-ly reflect the views of the funding agency.
2
asset redistribution. For all these reasons, extension systems are
politically attractive, and will continue to receive broad support, even
if they do not do not deliver on all these promises.
This optimistic view of extension is based on a body of literature
which presumes that all (or at least most) of the innovations suggested
to a farmer are in his interest, and that the obstacles to diffusion of these
innovations are in the farmer's ignorance or psychology or in the mode of
communication.l From this presumption comes the hope that improving the mode
of communication can strongly influence the rate of adoption.
While this paper acknowledges the importance of the communication
process per se, it places more emphasis on the structural context through
which the innovations are selected and communicated. This means that the
paper emphasizes the mechanisms of control over the extension system, and
treats as an empirical issue the question of whose interests are actually
served by a proposed innovation. It is hoped that by calling attention
to the wide range of economic, political, sociological and technical
factors which influence the functioning and results of an extension program,
this paper will prove relevant in efforts to develop or modify extension
programs to better serve small farmers and rural poor.
The first section of this paper examines some definitional
problems with the concept of extension and reviews some of the historical
background. The second part explores methodological problems in making
clear-cut assessments of extension activities. The third section indicates
why extension programs do not always help the rural poor. The fourth section
Everett Rogers, with F. Floyd Shoemaker, Communication of
Innovations (New York: The Free Press, 1971); J. Paul Leagans, "Extension
Education and Modernization," in J. Paul Leagans and Charles P. Loomis, eds.,
Behavioral Change in Agriculture (Ithaca: Cornell University Press, 1971),
pp. 101-147; A.T. Mosher, An Introduction to Agricultural Extension. (New
York: Agricultural Development Council, 1978).
3
explores ways of designing extension to meet more directly the needs of
small farmers. Administrative systems which place some powers at the central
level and others at the local level, which give power to groupings of
farmers, and which assure more feedback from farmers to researchers seem
better. The fifth section highlights the role of group organization in
facilitating extension programs. A concluding section summarizes the
argument and suggests pilot projects to give the generalizations more
concrete form adapted to specific conditions.
A. Definitions and Background of Extension Activities
A wide range of extension programs have emerged. Some of the
more common will be listed here, simply to indicate the type of programs
that will be analyzed late in the paper.
1. Conventional Extension. A conventional extension program
includes personal on-farm visits by extension agents to (usually larger
progressive) farmers. These farmers may adopt suggested techniques and
provide demonstration farms; or the agents themselves may cultivate demon-
stration plots.
2. Training and Visit Extension. Extension agents may receive
regular fortnightly training, and then come to villages on a regular
schedule to give groups of "contact farmers" specific recommendations on
cultural practices.
'3. Model Farmer. Village groups can elect representatives (a
"model farmer") to attend weekly or fortnightly training programs at some
administrative center. The model farmer is then obligated (in theory) to
report back to the group what he learned at the training program.
4. Farmer Training Programs. Training programs can be developed
for various time periods. During the dry slack season it might be for a
few weeks or more. At busier times, training would be limited to a few
days. In other cases, farmer training centers can provide training through-
out a whole agricultural year to farmers who reside at campus-like centers.
5. Mass Communication. Radio programs can offer farmer information.
Demonstration can be conducted at market places, fairs, etc.
6. Models. Innovative progressive individuals and villages can be
identified and used as models. Their successes can be highlighted in media
4a
and in large meetings, and others can be transported to examine their
fields and villages. This approach highlights farmer-to-farmer exchange.
7. Market Processes of Extension. Farmers often obtain some
inputs for agricultural innovations (seeds, fertilizer, tools, chemicals,
etc.) through regular commercial networks, including stores or merchants
at periodic markets. With the inputs can come information on how to use
them. Various programs can improve the capacity of this system to diffuse
material and information.
8. Para-statal Corporation. A para-statal corporation can supply
inputs (and usually credit to buy them) through a separate bureaucracy.
Farmers may be required to sell all production of certain crops to the cor-
poration, to assure repayment of inputs and to provide state procurement
of desired commodities. Field agents responsible to the corporation can
give instructions to farmers.
9. Farmer Controlled Organization. Voluntary associations controlled
and financed by farmers have played a valuable role in many countries in
identifying innovations and inputs that fill specific farmers needs. Train-
ing can be through local meetings of members of the organization.
10. Management Education. Literacy training and basic mathematics
education can be stressed to increase a farmer's access to information, and
to improve his ability to gauge the value of any innovation. This type of
education could be incorporated in normal primary education or in shorter
training programs.
This list is not comprehensive or systematic. Various types of
programs can be adopted simultaneously and will complement each other. The
point here simply is that there is a wide range of potential formats for farmers
(and their wives) to obtain knowledge about new agricultural techniques.
The word "extension" is useful because it brings to mind organized
activities of conveying (extending) technical information to farmers and
others. At the same time, however, the word "extension" also conveys some
erroneous implications and obscures some important distinctions. The word
extension unfortunately implies that generating knowledge (research) is a
very different activity from communicating information. In fact, as will be
discussed below, the communication function is integrally related to knowledge
creation, as it shapes both the generation and evaluation of hypotheses.
Thinking of integrated "learning systems," in which scientists and farmers
help each other to learn, is more useful than thinking of separate "research"
and "extension" systems.
The word "extension" also implies that the messages to be com-
municated are selected by one set of actors, and are then conveyed to
another set of people. The audience is primarily a passive recipient, whose
main option is to accept or reject the message. In fact, however, there
A detailed semantic analysis of the term extension is available
in Paulo Freire, "Extension or Communicating" in Education for Critical
Consciousness (New York: Seaberg Press, 1973), pp. 93-97.
4c
can be different types of interactions between the audience (farmer)
and the deliverer of the message (extension system). If an agricultural
policy is designed to benefit the urban populations, the extension system
would select and deliver information, inputs, credit, and market controls
to increase production and assure extraction of commodity grain (to assure
food supply to urban populations) and cash crops for export (to provide
foreign exchange for urban consumption and investment). These needs can
frequently be met by the small portion of larger, progressive farmers. The
rest of the rural population is irrelevant, except to the extent it is
needed to provide labor on plantations, in mines, in urban factories, etc.
This approach can be called-the "colonial extension system."'
A second type of interaction occurs when the system is designed
to change the rural values and social structure in a manner deter-
mined by forces outside the rural setting. Government leaders may
want peasants, to modernize, to give up superstitions, to change
their patterns of consumption and investments, to change their con-
ditions of health and sanitation, to change the pattern of inter-
group relations. The urban leaders may feel such changes will benefit
the rural people in the long run, and this may be true; but the ulti-
mate values spring from urban cultures. In a sense this is like a
"cultural invasion" of the rural areas by a foreign value system.2
Such an approach to extension can be called a "rural stimulation
system."3
Akhter Hameed Khan, Ten Decades of Rural Development--Lessons
from India (East Lansing: Michigan State University, Department of
Agricultural Economics, 1978).
Paulo Freire, Education for Critical Consciousness, pp. 111-127.
3George Axinn, "Agricultural Research Extension Services and
Field Stations," in International Encyclopedia of Higher Education
(San Francisco: Jossey-Bass, 1978), p. 243.
5
A third form of interaction is that the communication system
would be designed and organized by farmers themselves to provide the
technical information and services for specific problems they encounter.
Such a system can be called a "rural development acquisition system."
These different types of communications systems can be seen in
the historical evolution of different extension systems.2 When extension
systems were organized in the United States, Japan, Scandinavia, Holland,
and elsewhere before the turn of the century, there was little scientific
agriculture to extend. Instead, the extension systems were principally
communication networks among local groups of farmers, systematizing the
identification of superior techniques of advanced farmers and sharing them
with other farmers. In the U.S., sometimes businessmen played an active
leadership role of these local groups, because they anticipated that agri-
cultural development would expand local markets for farm inputs. These
systems were largely controlled and funded by the farmers themselves;
extension agents submitted their work plans to farmers' committees for
approval. Sometimes agents would not come to a village until a local group
had been created.
Although this system was controlled by farmers, government
support was crucial. Legislation was needed to give farmers' organi-
zations a legal personality. Government support was also necessary
in the training of staff for these extension systems. Staff for
this type of system in the U.S. almost invariably came from local farm
families; but they were trained at the government supported land
grant college system. In this way was generated a cadre of college
graduates who were literate, educated, and technically proficient,
while at the same time fully familiar with farm activities and essen-
1George Axinn, "Agricultural Research Extension Services and Field
Stations," in International Encyclopedia of Higher Education (San Francisco:
Jossey-Ball, 1978), p. 243.
2A comparison of thirteen extension systems is available in George
Axinn and Sudhakar Thorat, Modernizing World Agriculture: A Comparative
Study of Agricultural Extension Education Systems (New York: Praeger,
1972).
tially of the same culture as their eventual clients. They had to
pretensions of superiority over the farmers.1
Governments also supported scientific research. As agricultural
science advanced, the character of these extension systems also changed.
More communications flowed from scientists to farmers. Highly special-
ized extension personnel was needed to carry these communications.
However, this new flow of information was injected to a system which
already had farmer participation and control.
In other regions of the world, extension services were created as
part of the rural programs of colonial administrators. In the British,
French, and Dutch colonies of Africa and Asia, extension programs were
designed to provide supplies of desired commodities--indigo, tea, coffee,
rubber, cocoa, peanuts, sugar, etc. To fit this need, many research and
extension systems were oriented toward a single specific commodity. In
many cases these organizations continued after independence to maximize
foreign exchange earnings. The British-American Tobacco Company which
functioned in China and other places, the Kenya Tea Authority, the Jute
Ministry in Bangladesh, the Rubber Research Institute in Malaysia and the
crop-specific "operation" or societies for cotton, peanuts, cocoa,
coffee, etc. in the Sahelian countries of Africa are all examples of
this tendency. These types of parastatal agencies normally have some
applied research to determine a suitable "package" to grow the commodity.
The extension agencies then distribute the required inputs (seeds, ferti-
lizer, insecticide) and credit to buy them, offer highly detailed field
management instructions, and then purchase from the farmers the produce,
carefully graded for quality, at a price which normally is fairly low.
K. Robert Kerr and Robert Crom, "Putting Innovative Technology to
Work in Agriculture," Mimeo, Ames, Iowa, May 1979.
These types of extension systems require intensive services. In the Kenya
Tea Development Authority, an extension agent serves only 120 farmers; in
a tobacco scheme in Tanzania, an agent served 300 to 800 farmers. Mobility
is important to deliver services, so agents need to be equipped with motor
cycles or bicycles.
These extension systems frequently encourage monocropping of their
particular commodity. From the farmer's point of view, this may cause
problems because generally the farmer has a very complex, mixed farming
system, in which he grows several different cereals, some vegetables,
some cash crops, and livestock. The multiplicity of activities reduces
risk; moreover each commodity frequently produces byproducts which are
inputs into other rural activities. For example, cereal straw is animal
fodder; sorghum stalks are building materials; animal manure is fertilizer;
legumes build the soil and climb up maize stalks. Unless there are vast
changes in the market system, farmers frequently cannot specialize in the
particular commodity in which the colonial extension agency is interested.
Colonial extension systems normally require a staff that is literate
and fairly highly educated to handle the complex record-keeping inherent
in input supplies, credit, marketing, etc. This requirement normally
excludes local farmers, who'have no access to education, from becoming
regular staff in such an organization. Thus the staff will normally e, drawn from
the urban population, even though these people have only a limited understanding
of the rural economy and society. However, local farmers might be employed
as laborers or as communicators to give out specific information.
Colonial authorities were also concerned about general rural devel-
opment to alleviate recurrent famines, which threatened the legitimacy
1Uma Lele, The Design of Rural Development: Lessons from Africa
(Baltimore: Johns Hopkins, 1975), p. 67.
and political stability of the colonial system. Thus they showed some
interest in land tenure reforms, credit programs and in infrastructure
projects which would stabilize production and facilitate transport
both of cash crops and famine relief supplies.
Where there were numerous colonial settlers (Kenya, Algeria), the
extension services involved some farmer participation. Generally, however,
they were closely integrated with the colonial imperatives of taxation
and military/administrative control, and these requirements left little
room for decentralization or participation. Rather, the extension services
in the colonies tended to stress downward communication of information
selected to fit the needs and perceptions of colonial administrators and
to fit foreign markets.
In the Japanese colonies of Taiwan and Korea, the situation was
generally similar, with a wide range of policies to encourage export crops
(rice, sugar). However, greater efforts were made to establish farmer
participation in local farmer associations, similar to what existed in
Japan.
After World War Two, as colonies became independent and as the United
States became active in international assistance, there were some changes
in agricultural and rural policies, including extension. Countries saw
their rural needs more broadly. Not only did they want to produce cash
crops; they also wanted broadly based rural development, which would
include cultural, social, political, and economic changes. Hence, exten-
sion work was transformed to "community development" or "animation rural,"
both of which presumed widespread community participation. The village
was imagined to be a harmonious, integrated community, which could develop
holistically. Differences and conflicts between landowners, tenants,
laborers, etc. were downplayed. Some people, such as Wolf Ladejinsky,
believed that land tenure reforms in developing countries were necessary
to assure equitable and peaceful development, but his ideas were incorpor-
ated in only a few peculiar situations (post-war Japan and Taiwan) and
did not always work out (Vietnam in the 1950's).1
In both theory and practice, these different models of extension
cannot be separated from each other and must interact. For the urban
sector to get food and cash crops there must be some general social and
cultural change and farmer participation. Conversely, for farmers to solve
their problems, they need some increases in production and some access to
urban scientific and industrial products, culture, and markets. Finally,
within the rural sector, there usually are many different groups and
interests (following class, caste, or ethnic differences), so there may be
complex pressures for change within the rural sector, which can be rein-
forced by urban linkages. It is usually not the case that a "rural develop-
ment acquisition system" will be effective or equitable without some external
controls, guidelines and support from government.
Given the multiplicity of goals and the interactions that are
required, it is common that several different extension systems will be
set up in one country--one for food, one for jute, one in a specific
region, one to deal with a particular donor's project, etc. Chile in
Louis Walinsky, ed., Agrarian Reform as Unfinished Business,
The Selected Papers of Wolf Ladejinsky (New York: Oxford University
Press, 1977).
the 1960's had over a dozen separate extension agencies, and the
coordination problems were substantial.1
U.S. international programs in extension were based on a partial
adaptation of a model of organization which had emerged in the U.S.,
namely the tripartite system of agricultural research, education and
extension, all based on the land-grant college. Many observers in
the early post-war years, naively confident of the advances which had
already been made in agricultural science research, felt that agri-
cultural extension was the main bottleneck. In 1952 the Director
of the U.S. Office of Foreign Agriculture Relations stressed the role
of extension:
In these programs of agricultural technical assistance, the
main lines of endeavor are in the three fields of research,
resident education, and extension--a triumvirate long estab-
lished in this country but virtually unknown in many parts
of the world. Of2these, extension, as we know it, is usually
the missing link.
A critical aspect of the U.S. model of extension--namely con-
trol over extension by county level farmers' organizations--was not
integrated into international assistance for extension programs. The
reasons are undoubtedly complex. Perhaps the people involved in
international work did not fullyunderstand the significance of
local farmer organizations and control in the U.S. and elsewhere;
perhaps rulers of recipient countries did not want local farmer or-
ganizations for political reasons. Whatever the reason, this trun-
Marion Brown, "Agricultural 'Extension' in Chile: A Study
of Institutional Transplantation," Journal of Developing Areas 4
(January 1970).
John J. Haggerty, "The United States Farmer and the World
Around Him," Journal of Farm Economics 34:5 (December 1952), p. 601.
cated version of the U.S. system which was exported abroad was no
longer the U.S. system. It was substantially congruent with the
pre-war colonial extension systems of British, French, and Dutch
administrators in Asia and Africa in its philosophical and
structural base.
For the past thirty years, agricultural extension programs
have been important components in agricultural development work on
most continents. Major programs have been undertaken in Latin America,2
Asia,3 and selected African countries. In 1974, various countries
were spending about $1.3 billion annually for agricultural extension
(not including research or education) to support roughly 320,000
extension workers.4 (It should be noted that extension work is de-
fined differently in different countries, and some of these "ex-
tension workers" include regular field staff of ministries of
Akhter Hameed Khan, Ten Decades of Rural Development.
2Excellent reviews of agricultural development and extension
work in Latin America are Arthur Mosher, Technical Co-operation in
Latin-American Agriculture, and E.B. Rice, Extension in The Andes
(Cambridge: MIT Press, 1974).
A succinct statement of agricultural extension work in the
South Asian context is Akhter Hameed Khan, Ten Decades of Rural
Development.
James Boyce and Robert Evenson, Agricultural Research and
Extension Programs (N.Y.: Agricultural Development Council, 1975),
pp. 5, 32-36.
agriculture, and are not limited to people communicating technical
information.) About 30 percent ($89 million) of ongoing US AID
projects goes for agricultural extension activities incorporated in
larger projects.1
This is a sensible time to review some of the experiences in
agricultural extension. After several decades of development
efforts, developing countries and development assistance agencies
are now far more aware of the difficulty and necessity both to in-
crease food and agricultural production and to assure equity in
distribution of benefits. Substantial experience has been accumu-
lated in agricultural extension. The challenge now is to design
extension systems which are not extrapolations of the needs and ex-
periences of developed countries or of the interests of bureau-
cracies and urban residents of developing countries, but which re-
flect the needs, aspirations and concrete conditions of the de-
veloping countries and in particular their rural poor.
B. Methodological Problems in Evaluating Extension Programs
Over the past decades, there has been substantial develop-
ment in world agriculture. Crops and technology have changed, pro-
duction has grown, rural relationships have become commercialized.
Poverty, however, remains and may be increasing. What is tne role'
of agricultural extension activities in these changes? Do they
deliver on their promises to improve production and welfare? This
Delbert Myren, "Agricultural Extension: AID Experience,
Present Involvement and Some Unresolved Issues," AID mimeo, 1976.
13
turns out to be a very difficult question.1 Rural development requires
a complex combination of suitable conditions--seeds, fertilizer,
pest control, water management, roads, transportation, markets, price
incentives, credit, security, fair tenure conditions, consumer goods,
local organizations, effective administration, access to a fair
judicial system, a set of values and social relations that encourage
change, and technical knowledge.
Extension services can easily improve the knowledge factor,
but it is rare that this by itself (or any other single factor by
itself) can make a dramatic increase in welfare, much less production.
Moreover, the payoff of extension is closely related to the specific
qualities of the technology that is being popularized. Energeti-
cally extending technologies which are no better than traditional
ones is not beneficial and is not only a waste of money but can be
detrimental because efforts at modernization will be discredited.
Likewise extending high yield varieties does little good if fertili-
zer and pesticides (and credit to buy them) are unavailable, or if
timely irrigation cannot be provided. Nor will many farmers benefit
Excellent surveys of the issues involved in evaluating extension
programs are available in Robert Chambers, Two Frontiers in Rural
Management: Agricultural Extension and Managing the Exploitation
of Communal Natural Resources (Sussex: Institute of Development
Studies, 1975), pp. 2-4; and E.B. Rice, Extension in the Andes,
esp. pp.'161-166.
Excellent statements concerning the wide range of factors
which are necessary to "get agriculture moving" can be found in the
writings of Arthur Mosher, Technical Cooperation; Getting Agriculture
Moving (New York: Praeger, 1966); and Creating a Progressive
Rural Structure (New York: Agricultural Development Council, 1969).
if land tenure relations and/or credit and marketing patterns are
highly exploitative, as was apparent in the KADU project of Ethiopia.
Even if modern inputs are available and farmers are convinced
(or assisted) by extension agents to adopt them, adoption may not
make economic sense. -A careful analysis of India's Intensive Agri-
culture Districts Program (IADP) in the early 1960's shows that it
was effective in raising farmers' production because they used new
inputs. However, the actual factor productivity (i.e., efficiency
in using resources) of the farmers was unchanged. To increase factor
productivity, the authors of this study believe investments in re-
search systems normally have a higher payoff than investments in
extension systems.2
In gauging whether or not the new technology is superior,
attention must be paid to the question of. risk avoidance. Highly
productive agricultural technologies often are riskier. They need
the right amount and timing of rainfall./irrigation, and are more
vulnerable to insects and diseases. Because of the high cost of
inputs, their profitability is more dependent on market prices and
access, which may fluctuate. For subsistence peasants on the edge
of survival, risks of these types are extremely serious. A large
drop in production may force sale of all assets, acceptance of virtual
slavery to obtain food, dissolution of the family, and eventual
starvation. Thus the marginal peasant is quite rational when he
rejects risky technology. The definition of technical superiority
should include this question of risk avoidance.
John Cohen, "Effects of Green Revolution Strategies on Tenants and
Small-scale Landowners in the Chilalo Region of Ethiopia," The Journal of
Developing Areas 9 (April 1975), p. 335-358.
Rakesh Mohan and Robert Evenson, "The Intensive Agricultural
Districts Programme in India: A New Evaluation," Journal of Development
Studies 11:3 (April 1975), pp. 148-150.
In cases where the new technology clearly is novel and sensible
and when complementary inputs are available, extension can have a
stunning impact. This seems particularly true with hybrid seeds.
In one region of Kenya, hybrid maize was popularized in the mid-
1960's. In 1965, 5,000 demonstration plots were set up, and 28 per-
cent of the farmers visited one. Over 35 percentof farmers first heard
about4hybrid maize from extension agents. (Almost 45 percent heard from
,friends or neighbors.) -Extension programs have had success with hybrid
2 .3
maize in the U.S..,. and more recently with hybrid rice in China.
Evaluation of extension programs is complicated by the fact
that extension agents are not the only source of knowledge for
farmers, and frequently they are only a marginal source. Farmers
get information from friends and relatives, from skilled local
farmers, from merchants and salesmen, etc. This information net-
work, which might be called a "spontaneous extension system," is
often very efficient in some social environments. Historically the
diffusion of new crops around the world--including maize and potatoes
from the Western Hemisphere to Europe and Asia, early maturing varieties
of rice in Asia, numerous cash crops (sugar, indigo, opium, etc.)--has been
quite rapid, and has not required formal extension services. Early
John Gerhart, "The Diffusion of Hybrid Maize in Western
Kenya," (Mexico. D.F.:CIMMYT, 1975), p. 9.
Zvi Griliches, "Research Costs and Social Returns: Hybrid
Corn and Related Innovations," Journal of Political Economy,
October 1958.
"Some Problems on the Development of Hybrid Rice," Scientia
Agriculture Sinica, Feb., 1978. Available in JPRS 71, 717, PRC
Agriculture no. 3., p. 1-18.
4This phrase is suggested by Michel Cernea.
V. W. Ruttan and Yujiro Hayami, "Technology Transfer and
Agricultural Development, Technology and Culture 14:2 (April 1973),
p. 120.
exploration and trade missions carried seeds with them, and often had
botanists to identify useful new crops; but extension agents were not
required.
Often, however, an extension system can make a crucial initial
input into the spontaneous extension system. It can play a catalytic
role in energizing the spontaneous extension system. Moreover, in some
social environments, spontaneous extension works poorly. When a village
(or country) is divided by class, religion, caste, linguistic, factional
cleavages, or ecological factors (mountains, flooded fields), exchange of
farming information between families may be very limited. Formal extension
programs could have a major impact in such a situation. Unfortunately,
such places sometimes receive less extension contact because the extension
agents find it less comfortable to work in such a social environment.1
Spontaneous extension may also be slow in a village that has such
strong cultural homogeneity and unity that no one is willing to deviate
from the norm and try an innovation. An innovator may risk becoming
socially isolated-, and may even be accused of witchcraft.2 In Java, one
journalist writes, "What the Javanese discovered is what every anthropologist
(but almost no development economist) knows: The most potent force in
every village is not government fiat but rather it is a fear of the neighbors'
gossipy censure or 'What will people say?'"3 In such a case, spontaneous
This experience was revealed in careful studies of two Indian
villages. J.P. Hrabovsky and T.K. Moulik, "Economic and Social Factors
Associated with the Adoption of an Improved Implement: A Study of the
Olpad Thresher in India," Agricultural Development Council paper, 1967,
pp. 8-9.
Lele, p. 76.
Richard Critchfield, "More Food, Fewer Months, Java Confounds
the Doomsayers," The Christian Science Monitor, Nov. 24, 1978, p. B.12.
diffusion from a single innovator may be negligible, if a courageous
innovator exists at all. However, if such a village begins to change,
everyone may change almost simultaneously.
The problem of peasant resistance to change may be frequently
overestimated. Peasants may prefer to tell outside observers that
recommended innovations are culturally unacceptable rather than go into
the complex and controversial details of technical suitability, risk
avoidance, suitability of price, availability of consumer goods on which
to spend increased cash income, interactions within a delicately balanced
farm system, and local land tenure relations. They may not be able to
articulate their underlying fear that a particular innovation will under-
mine the long-term ecological balance of soil, fertility, animals, fish,
etc. on which their descendants will depend. They may not want to de-
scribe their interest in the maintenance of existing patron-client
social relationships, which although exploitative, also provide some
economic security and cultural stability. Nor may they want to explain
that they see diffusion of a particular innovation (along with its credit
and market linkages) to represent penetration into their community by
a distrusted political or ethnic group or an exploitative government.
They may be suspicious of extension agents who seem to come just before
election time.1 Nor may they be willing to admit resentment of those
aggressive, lucky or well-connected enough to get access to new technology,
and credit to buy it.2 It is far easier to tell an outside observer that
Marion Brown, p. 207.
2For a detailed analysis of socio-political problems caused by
an attempt to modernize fishing in Java, see Donald Emerson, "Biting the
Helping Hand: Modernization and Violence in an Indonesian Fishing
Community," Land Tenure Center Newsletter, No. 5 (January-March, 1976).
the innovation simply is "culturally unacceptable." Simultaneously,
outside observers may have their own reasons for being satisfied with
an answer of cultural unacceptability, and may fail to probe additional
factors.
Research on extension systems does show that farmers who have
extensive contact with extension agents are more likely to adopt new
farming methods.1 Correlation, however, does not prove causation. The
larger market-oriented farmer can afford to be more innovative, and is
also more likely to seek extension services. The correlation between
extension contact and adoption sometimes is highest in cases where an
2
innovation is not particularly profitable. In such cases some farmers
are willing to follow (for a brief period, at least) the advice of an
extension agent perhaps because he is an effective salesman, perhaps to
endear themselves to the extension agent and obtain more profitable
favors in the future. In at least one case, farmers with more extension
contact had lower yields, although the direction of causation is not clear.
Everett Rogers, Joseph Ashcroft, and Niels Roling, Diffusion of
Innovations in Brazil, Nigeria, and India (East Lansing: Michigan State
University, 1970).
2This has been demonstrated clearly with data on short-term
adoption rates of fertilizer in different regions of Ethiopia. Bisrat
Aklilu, "Technological Change in Subsistance Agriculture: The Adoption
and Diffusion of Fertilizer in Ethiopia's Minimum Package Areas,"
Ph.D. dissertation, Boston University, 1976, pp. 233-234.
David Leonard points out that extension agents and some farmers
can have patron-client relationships. The agent needs a farmer who will
try almost anything so that he can assure superiors that he is managing
to arrange some adoption. For the farmer, doing the favor of adopting
new techniques for the agent may assure access to credits, market infor-
mation, educational opportunities, etc. which may have long-term payoffs
far greater than the losses incurred due to unprofitable innovations.
4Uma Lele, The Design of Rural Development, p. 71.
19
How should such extension success be evaluated? Must extension services
be considered effective when they result in farmer adoption of new
techniques, even if these techniques are not profitable?
This problem points to further complications in evaluation of
extension. Perhaps the goal of extension is not so much the populari-
zation of a particular agricultural technique, but rather the spread of
a particular mode of analysis so that the farmer can more accurately
evaluate his experiments. Record keeping,. rigorous cost accounting,
and statistical analysis give the farmer more accuracy in selecting
and changing continuously improved techniques over long periods of time.
Certainly this approach to analyzing farm innovation is far more useful than!
simply convincing farmers blindly to obey a suggestion to buy for single
season. It clearly is far more difficult to measure the long-term pay-
off of this new analytical approach than to estimate the profitability
of a single enlarged sugar crop. The long-term consequences of blind
obedience, however, both economically and politically, can be most severe.
The final and most serious problem of evaluating extension
systems is that the systems and their results are inevitably, intimately
and inextricably tied to the entire political, economic and social order.
The question of who benefits from an extension system is in the final
analysis shaped by patterns of distribution within the family, which are
shaped by broad cultural factors; by the interaction within villages,
which is closely related to land tenure relations; by urban-rural inter-
actions as reflected in price, tax, investment and migration patterns;
and by international market structures.
Extension systems should not be credited or blamed for political,
economic, social, or technical factors beyond their control. It is
unfair to blame the U.S. or Indian extension systems alone for the
failure of Black sharecroppers or landless laborers to benefit from
development. Likewise, it is erroneous to give excessive plaudits to
the U.S. or Kenyan extension systems without acknowledging decades of
agricultural research that generated superior techniques that could be
extended. Nor can one expect an extension service to have more than
marginal impact on prices of commodities traded internationally or
controlled domestically. Extension systems will generally reflect and
perhaps intensify the existing order of economic and social relations,
whatever thay may be.
Where society is more egalitarian, extension can reinforce this
tendency, as in China, Taiwan, and Israel and elsewhere. In China,
collective ownership of land assures that the benefits of rural develop-
ment are shared reasonably equally. Moreover, collective ownership of
land establishes local organization, which can contribute personnel,
funds, and skills to extension work. The extension system in China
works closely with rural local organizations, is substantially staffed
by farmers and financed by them and is essentially merged with the
research system. Professional specialization and status are sharply
attenuated.2 Both the role played by rural local organization and
the support from the center are common features of other
David Leonard Points out: "One of the reasons for the strength
of Kenya's farming sector has been the quality of its research complex."
Reaching the Peasant Farmer, Organization Theory and Practice in Kenya
(Chicago, University of Chicago Press, 1977), p. 246.
Benedict Stavis, "Agricultural Research and Extension Services
in China," World Development 6:5 (May 1978) pp. 631-645.
successful extension systems. However, it is clear that the consequences
of extension in China are more related to basic political factors than
to the specific characteristics of the extension system.
C. Potential Problems in Extension Systems
It is reckless to make generalizations about extension programs,
considering the wide diversity of experiences. Nevertheless, three types
of problems sometimes limit the ability of an extension system to serve
the needs of the rural poor. The first problem is that extension systems
can be directed by an urban-based political system to benefit urban
interests. Secondly, for a variety of reasons, extension agents may work
primarily with the rural elite. Thirdly, extension systems create
their own bureaucracies which can generate their own bureaucratic and
personal interests. These factors mean that the actual results of many
extension systems, regardless of the rhetoric, may be other than helping
the rural poor.
1. Urban Control. Agricultural extension systems usually are
not controlled by farmers themselves. George Axinn notes that only 13
out of 78 extension systems around the world involve some funding from
farmers or local government. Instead, extension systems are funded and
are controlled entirely by governments, which are often obligated to
give priority to urban political forces. Frequently, they have specific,
narrow goals, production of cash crops, such as cotton, peanuts, cocoa,
coffee, rubber, etc. to obtain foreign exchange for urban luxury consumer
1George Axinn, "Agricultural Research, Extension Services, and
Field Stations," in International Encyclopedia of Higher Education
(San Francisco: Jossey-Bass, 1978), p. 243.
good or for financing capital imports. Also they want higher food pro-
duction to reduce the pressure on foreign exchange caused by food
imports. Extension services may be closely related to monopolistic
marketing structures, which assure that the benefits of agricultural growth
goes to the trader or consumer, in the domestic urban areas or abroad.
This approach usually emphasizes generating marketable surpluses
of grain or cash crops by providing industrial inputs, so this type of
extension system normally concentrates its efforts in places close to
roads or other transportation infrastructure. People living far away
from roads get less attention.
Moreover, the innovation suggested by the extension system
may not fit easily into the farmer's complex farming system. The farmers
may perceive different needs. Nevertheless, if the innovation can be
used and if access to markets is adequate, farmers may welcome oppor-
tunities to sell profitable cash crops.
An extension system that is successful in spreading cash crops
will not necessarily bring direct benefits to the rural poor. Indeed,
some economic planners may believe cash crops can be spread by displac-
ing small farmers and tenants, so that larger, technically sophisticated,
financially strong farms can emerge; and so that a flexible, mobile,
quiescent rural labor force can be created for plantation labor and for
mines, and factories as well. In other cases, however, small-holder culti-
vation of cash crops is encouraged. If the case crop is labor intensive and
has a high value, cash crop cultivation may offer the small farmer (with high
labor to land ratio) his best chance for a higher income.
2. Rural Elites. These urban-rural categories of analysis
must be supplemented with the dimension of class. It is not unusual
23
for urban classes to ally with the narrow rural elites. In such a
situation, most of the benefits of development will go to the cities,
but the rural elites will also prosper. The rural poor will remain
clients of the rural elites, and are tempted to migrate into the urban
Tabor force..
Certain features in the methodology of extension work are
congruent with, and reinforce, such a political alliance. Analysis
of the activities of extension workers in many countries shows that
they tend to focus attention on the few "progressive farmers," who
have adequate resources to afford and risk new technologies, and who
are interested in selling (and buying) more. Such an empirical find-
ing is often'converted to a prescription. It is presumed (sometimes
accurately) that knowledge will "spread" from these "community
leaders" to other farmers, and that the benefits will eventually
"trickle down" to the poor, through more efficient technology, more
jobs, or lower food prices. (There is, however, a risk. The "progressive
farmer" may, for ethnic, cultural, psychological, or historical reasons,
turn out to be someone who is not trusted or respected in the community.
In such a case, there may be very little spread affect.)2
From an administrative point of view, it is convenient for
extension agents to work with large farmers. To have an innovation
utilized on 100 hectares of land, it is far easier to convince
one farmer with 100 hectares than to convince 100 farmers with only
one hectare. If the extension agent is judged by his success in
extending an innovation over a wide area, he will not waste his time
with small farmers. In addition, the extension agent, often coming
1Michael Lipton, "Towards a Theory of Land Reform," in David
Lehmann, ed., Agrarian Reform and Agrarian Reformism (London: Faber
and Faber, 1974), pp. 311-312.
2Everett M. Rogers, with F. Floyd Shoemaker, Communication of
Innovations (New York: Free Press, 1971), p. 5.
from an urban or wealthy rural background himself, will find it far
more pleasant dealing with the large farmers, who can provide food
and drink and who is likely to be literate (or have a literate son)
and to be interested in farming innovations. Moreover, due to the
political power of the larger farmers, it is likely that the very
technology developed in the research system will be better suited
to the needs of larger farmers, making it even more rational for
extension agents to orient their energies to large farmers.
In a broad sense, the village elite generally serves as the
linkage point between the village and the government for many
functions--police, tax collection, military recruitment, political
mobilization. It is normal that the rural elite be perceived by
government, extension agents, the rural poor and the rural elite
themselves as the logical way for extension agents to enter the
village economy and society. Thus, both the philosophy underlying
some extension methods, the administrative practices of most exten-
sion systems and technological factors match precisely the needs
of a small minority of "progressive farmers," who generally own more
land.
Not surprisingly, the empirical research in most countries
shows that a small minority of farmers do get the bulk of extension
services. In one survey in Kenya, progressive farmers (who con-
stitute only 10 percent of all farmers) received 57 percent of the
extension visits. The 47 percent of farmers considered non-
Rene Benalcazar R., "New Techniques, Agricultural Extension
Services and Credit Facilities as Instruments of Economic Development,
with Special Reference to Latin America," in Theoda Dams and Kenneth
Hunt, ed., Decision Making and Agriculture (Lincoln:University of
Nebraska Press, 1978), p. 521-22.
25
innovative got only 6 percent of the extension visits. Other surveys
in Kenya agree.2 Likewise, even in socialist Tanzania, with strong
rhetorical pressures toward equality, one survey showed that 59 percent
of wealthy farmers had a high level of extension contact, while
only 29 percent of poor farmers had such contacts.3 In one locality
of Northern Nigeria, the local elites had reasonably good knowledge
about extension activity, while most farmers had negligable informa-
tion.4 Thus, the normal pattern is that extension services concentrate
on services for the larger, progressive farmers.
This unequal access to extension services can have different consequences
for long term trends in income distribution and social structure, depending on the
character of institutions and on the markets. In many countries this
tendency of extension to help primarily the rural wealthy reinforces
underlying economic, political, and social trends. With profits coming
from improved technology, large progressive'farmers may buy out their
neighbors who are unable to meet the costs of commercial farming,
particularly if land ceiling restrictions are lax. Some small farmers,
and particularly their sons, may have no choice but to become land-
less laborers or migrants to cities. Thus a knowledge gap between
the rich and poor reinforces an economic gap.5 These tendencies will
David Leonard, pp. 125-177.
Joseph Ashcroft, Niels Roling, Joseph Kariuki, and Fred
Chege, Extension and the Forgotten Farmer (Wageningen: Afdelingen
voor Sociate Wetenschappen aan de Landbouwhogeschool, 1973), p. 31.
James DeVries, "Agricultural Extension and the Development
of Ujamaa Villages in Tanzania: Toward a Diological Agricultural Ex-
tension Model," Ph.D. dissertation, University of Wisconsin, Madison,
1978, p. 167.
Peter Matlon, "The Size, Distribution, Structure, and Deter-
minants of Personal Income Among Farmers in the North of Nigeria,"
Cornell Agricultural Economics Ph.D. dissertation, 1977, p. 390.
On next page.
26
be particularly severe if inputs are "lumpy" (tractors, tube wells)
and only the larger farmers can afford them or use them efficiently,
or if the total demand for the crop in question is inelastic. In
such a situation, when the early adopters of improved technology
produce more, the price will drop; the late adopters will get no
benefits, and the non-adopter may be hurt by lower prices.
If, however, inputs are divisible (fertilizer, seed) and demand
for the commodity is very high and if prices will not be depressed by increased
production, the late adopters (who generally are smaller farmers)
canget some benefits from the new technology. Indeed the smaller
farmers are likely to get most of the benefits if the innovation
requires high labor inputs per land area.. Such farmers may quickly adopt
labor intensive cash crops, vegetable cultivation, animal husbandry,
or dairy, if they get access to information, credit and markets.
In one locality of Northern Nigeria, it appears that smaller farmers
got preferential access to new peanut seeds, although there is the
possibility that village elites who made these allocations benefited
directly (by switching the seeds and giving poorer farmers bad
seeds) or indirectly (by increased political patronage) from them
also.
What has extension done for these smaller farmers who are
hurt by the expanding power of larger farmers? One subtle analysis
of extension in the U.S. maintains that its main contribution has
been not to increase their productivity but rather to facilitate
5 (from previous page)
Prakeash M. Singh and Bella Mody, The Communication Effects
Gap: A Field Experiment on Television and Agricultural Ignorance
in India," in Everett Rogers, ed., Communication and Development:
Critical Perspectives (Beverly Hills: Sage, 1976), p. 83.
1Peter Matlon, p. 383.
peaceful social change in the face of overwhelming economic forces,
Extension programs obscured for small farmers their long-term in-
ability to compete successfully with larger farmers; they were lulled
away from potentially effective political actions which might have
restructured markets. They eventually were forced out of business.
"By its sincere but futile effort to maintain a rural way of life,
Cooperative Extension helped to defuse a potential farmer revolt
in the United States. By giving farmers a false hope that adopting
new techniques of farming and farm management would preserve their
family farms, extension furthered the transformation from a rural
to an urban society in a way that avoided violence."1 A more gener-
ous analysis might say that extension agents helped nonviable farmers
make the difficult decision to leave farming. A cynical analyst
might question whether extension systems have deliberately tried
to drive small farmers out of business to maximize profits of large
agribusiness corporations.
3. Bureaucratic Interests
Extension systems can employ a large number of people, who
naturally have an interest in their own welfare. They view exten-
sion as a career, and naturally to a large extent plan their exten-
sion activities in a manner most suited to career advancement and
most congruent with self-perception of career status. They are
often extremely sensitive to the nuances of bureaucratic power and
status. Minor differences in procedures for computing travel
Robert Carlson, "Cooperative Extension: A Historical
Assessment," Journal of Extension, 8:3 (Fall 1970), p. 14.
allowances can have an important impact on personnel behavior. In
many circumstances, extension agents are regular government employees
with civil service and pension rights, and this inevitably shapes
their behavior. They may be attracted to government service be-
cause of its security and pension. The normal functioning of govern-
ment will influence the behavior of extension workers; if irrespon-
sibility, lack of performance, and petty corruption are not controlled
in other ministries, there is no reason to expect the extension
department to function differently. If college graduates anticipate getting
government work with a pension, it may be naive to expect them to
work for non-governmental agencies (e.g., private cooperatives) without
pension rights. All such factors inevitably affect the functioning of
an extension system and enhance or reduce its ability to serve the needs
of rural poor.
In many cases, especially when extension services are first
established, their young cadres may radiate a distinctive enthus-
iastic spirit inspired by a nationalist yearning to modernize their
country. For example, one observer described extension agents in
Peru in the early 1950s in glowing terms: "The agents...are well-
trained, honest, responsible, very much interested in their jobs
and in getting results. They show sympathy and understanding for the
people they are trying to help."l Similar comments might have been
made about the staff of Ethiopia's Extension and Project Implemen-
tation Department (EPID) in the early 1970's.
1Anibal Buitron, cited in Arthur Mosher, Technical Coopera-
tion, pp. 66-67.
29
However, in some (older?) extension services, the personnel
become somewhat elitist and isolated from farmers. These tendencies
.-have been described graphically in Venezuela:
The agents insist on being addressed as "doctor."...[They]
never leave their jeeps to visit the houses but instead
beckon members of the household to their car... The seem-
ingly foppish cleaning of town shoes with a paper hander-
chief to remove the mere suggestion of mud and the taking
of fruit without asking are other perceived manipulations
of the extension agents' sense of superiority.
Similarly, in Colombia, "agents wore ties and suits when working in
the field and refused to eat or talk with their peasant clients."2
The actual extent to which such behavior is associated with less
advice and difficulties in communication with farmers is, of course,
an empirical question meriting careful study; but in any event
there is a consensus of opinion that farmers will not trust this
3
advice, regardless of its accuracy.
D. Issues in Improving Extension Services
Despite the difficulty in evaluating extension systems, and
despite widespread problems, political support remains for extension
programs, and they are likely to be maintained and expanded. Thus
the practical challenge is not to gauge whether extension works in
an overall sense or to list the problems, but rather to make particular
R. Chesterfield and K. Ruddle, "Nondeliberate Education:
Venezuelan Campesino Perceptions of Extension Agents and their
Message," in T.J. LaBelle, ed., Educational Alternative in Latin
America (Los Angeles: University of California Latin American
Center, 1975), p. 153. Cited in Dennis Rondinelli and Kenneth
Ruddle, Urbanization and Rural Development, A Spatial Policy for
Equitable Growth (New York: Praeger, 1978), p. 92.
2Allen Jedlicka, Organization for Rural Development (New York:
Praeger, 1977), p. 23.
3Rogers argues that communication will be better between people
who are similar, and calls this the principle of homophily. Everett
Rogers and F. Floyd Shoemaker, Communication of Innovations (New York:
Free Press, 1971), p. 210.
suggestions that can maximize the value of extension systems which
will exist anyway.
Just as water obeys laws of gravity and flows downhill, agri-
cultural extension systems naturally tend to help those who control
the system. Thus the crucial question is what forces control the
agricultural extension system. Control over extension systems can
be exercised at the center, through rigid bureaucratic and personnel
regulations, or it can be deconcentrated to agents stationed in the
field but responsible to and closely monitored by the center. Both of
these approaches will be considered "centralized systems." Alternatively,
control can be decentralized to local agencies, who can hire and fund
personnel and specify program ends. This approach will be called a
"decentralized system."
For extension to help the rural poor is analogous to water
flowing uphill. Water does, of course, flow uphill, when a pump
applied pressure or suction, and when pipes confine its flow.
Similarly, it is possible for an extension service to help the poor.
It must be pushed by strong centralized management and pulled by rural
local organizations with real power. Just as pumps can be turned off
or break and pipes can clog or rust, however, administrative pressure
and local organizations may cease to be effective after a period of
time. A complex, delicate balance that somehow combines the
advantages of centralized administration with those of local control
is needed.1
From the point of the rural poor, the critical question is at
what administrative level--the central political system or the local
social system--do they have greater representation and power. It is
at that level that power over an extension system should be stressed,
if it is to best serve the poor.
1. Centralized Management
Some writers on extension in develooina nations emphasize
the potential benefits which can be achieved for more rigorous,
centralized management in which extension agents are given narrow
specific tasks and rewarded accordingly. They argue that central
control can compel extension agents to act in a desired manner.
The first step in this requires limiting the tasks that are
given to the extension agent. Because he is often a government's
primary contact with farmers, the extension agent has multiple tasks. He is
saddled with responsibilities for policing, debt collecting,
data collection, general reporting, and input supply and ration-
ing, in addition to giving technical advice. In time of national
disaster, the extension service may be given the task of distribu-
ting food or other types of relief and emergency equipment. Each
The importance of rural local organizations is highlighted
by Norman Uphoff and Milton Esman, Local Organization for Rural
Development in Asia (Ithaca: Cornell Rural Development Committee,
1974).
32
of these roles requires a different style of operation (disciplinary,
predatory, clerical, commercial, innovative); a different relationship
with farmers (hostile, passive, servicing, advisory); and different pat-
terns of movement (extensive field travel or regular presence in head-
quarters). There are, of course, economies to be secured by having a
single agent perform all these tasks. At the same time, it is unlikely
that he will be able to perform any of the tasks particularly well.
Robert Chambers points out:
The time and energy of extension staff at the lowest
levels tend to be regarded as infinitely elastic; in
fact they are, and should be treated as, finite and
scarce.
Not only will he lack time to do everything; he will find that the
work style and pattern of relationship necessary for one responsibility
will preclude effective execution of other responsibilities. Moreover,
because the demands on him are contradictory and priorities are not spelled
out, it is far easier for extension agents to shirk all responsibility
and do little work. The tendency to reauce work loads may well be rein-
forced by work group loyalties of junior staff, hostile to high level
staff with authority.3 Thus, for a variety of reasons, a clear choice
should be made concerning the prime responsibilities of the extension
agents. This seems a simple suggestion, but in reality
4
such clarity of purpose is rare in Africa, among other places.
Robert Chambers, Two Frontiers in Rural Management, p. 5.
2Ibid.
Leonard, pp. 64-80.
4H.S. Belay, "A Comparative Analysis of Agricultural Extension Systems,"
Journal of the Association for the Advancement of Agricultural Sciences in
Africa 2 (Supplement 2, Papers Presented to the Conference on Agricultural
Research and Production in Africa, September 1971) (June 1975), p. 319
33
Daniel Benor proposes that the prime (or sole) responsibility
of the extension agent be conveying agricultural information. "It is
important...that...the agents are not diverted from their task of advising
farmers by any conflicting demands to perform other services." Other
agencies should perform other tasks.1 The extension agent should be
able to focus attention exclusively on bringing information to farmers.
Arthur Mosher, however, suggests other functions are important, and that
there is an evolution of extension systems in which the function changes.
Extension agents function first as "encouraging companion" for local
innovators; then as a source of technical information; then as a contact
person to help sophisticated farmers get information from subject matter
specialists; then as a group organizer to facilitate discussions about
the politico-economic framework that affects agriculture; and finally, as
a stimulator of general development.2 Whether or not Mosher's categories
are precisely accurate and inevitable in a system being designed for the
rural poor, it is certainly correct that there are a variety of extension
needs, and these needs are always changing as development proceeds. Benor
may have accurately gauged the level of development in Turkey and India,
and may be correct that farmers there now need information. But this
observation should not exclude the possibility that other tasks may be
more relevant in other situations.
Daniel Benor and James Harrison, Agricultural Extension, The
Training and Visit System (Washington, D.C.: The World Bank, 1977),
pp. viii, 11.
A.T. Mosher, "A Note on the Evolutionary Role of Extension
Work," Land Economics 42:3 (August 1966), pp. 387-389.
34
The next issue is the selection of target groups on whom exten-
sion agents should concentrate their efforts. Indeed, the one attrac-
tive feature of a centralized system is the apparent capability to
redirect extension services from the normal beneficiaries--the large,
progressive farmer--to the small, poorer farmer. Benor proposes that
extension agents try to serve all farmers, particularly small farmers.
He suggests that each extension agent work directly with six to eight
village extension workers, who are government employees. Each of these
village extension workers will work with eight groups of farmers (each
group involving from 300 to 1,200 families depending on circumstances).
In each group, about 10 percent of the farmers (i.e. 30-120) are selected
as contact farmers. (Other experiences, mentioned below, highlight
advantages of working with smaller groups--under 30 people.) The village
extension worker would then have a rigid schedule of visits with each
group and its contact farmers once fortnightly, always at the same time
(e.g., every other Tuesday). Then, each week one day is available for
the village extension worker to receive specialized training from subject
matter specialists. Because training and visiting are set according to
rigid schedules, the system has been named the "training and visiting
system." (The principle of rigidly scheduled visits had been previously
utilized by the Kenya Tea Development Authority.)
The contact farmers are not elected by the group of farmers.
Rather, they are appointed by the government. In theory, the village
extension worker will work in consultation with village leaders to assure
that the contact farmers are progressive farmers and are representative
of various types and sizes of-farmers. Implicit in that practice is the
fear that if villagers were autonomously to elect contact farmers, they
might be wealthy farmers, not representative of the average. On the
other hand, if there has been no broad discussions in the village about
the selection of the contact farmers, then most farmers may be unaware of
who the contact farmers are, and will not know to ask them for advice.
The World Bank has utilized this T & V system in Turkey, and in
different states in India (these projects have a total cost of
$141.5 million, of which the Bank has loaned $70.7 million), and has
issued a detailed paper suggesting how to evaluate these projects.
Benor reports these projects helped increase yields over 50 percent
from 1.3-1.7 tons to 2-3 tons per hectare.
It should be mentioned that some writers on extension suggest not
trying to reach all farmers. In the Latin-American context, one writer suggests
that extension agents concentrate their efforts on medium-sized farmers.
The large farmers do not need help, and the small farmers are too numerous
and weak for the capacities of existing extension services.2 In India,
specialized agencies have been created for specific target groups, the
Small Farmer Development Agency (SFDA) and the Marginal Farmers and
Agricultural Laborers Development Agency (MFALDA). These agencies
help small farmers and landless laborers establish enterprises that
need little access to land, such as fruit and vegetable or dairy produc-
tion, veterinary services, marketing assistance, etc.
Michael Cernea and Benjamin Tepping, A System for Monitoring
and Evaluating Agricultural Extension Projects (Washington: World Bank,
1977). Details on the Bank loans for these projects are on p. 85.
2Rene Benalcazar R., "New Techniques, Extension Services ...,
p. 523.
36
In the U.S., on the other hand, in recent years some people argue
that most farmers have enough sources of technical information, and that
there is little justification for having major public programs exclusively
oriented to serving a small clientele of successful farmers. Instead,
they suggest that a new direction for extension could be general adult
nonformal education and assistance, particularly for poverty-stricken
people in both rural and urban areas.1 Others suggest that extension
agents work with small businessmen in general.
Crucial to making an extension system function effectively to
meet whatever goals are established for it are the incentives which shape
the behavior of the agents. Leonard believes that the manner in which
promotions are made within an extension service is one of the most
critical incentives.
But what criteria are to be used for making promotions? Senority?
Number of adopters? Area of adoption? Quantities of inputs distributed?
Subjective feelings of farmers about the value of the extension agent?
Moreover, how can this performance be monitored? Will it depend on internal
reporting? If so, are these internal statistics likely to be distorted?
Considerable thought must be given to structuring the career ladder and
internal incentives of the extension system so that the induced behavior
is consistent with desired policy outcomes. As general guidelines,
Leonard suggests that efforts be made to hire only those who would eventually
be capable of promotion; that at least 25 percent of employees receive
substantial promotions within 10-15 years; and that promotions and upgrading
Paul Miller, The Cooperative Extension Service: Paradoxical
Servant The Rural Precedent in Continuing Education (Syracuse:
Syracuse University Publications in Continuing Education, 1973).
37
be based primarily on job performance.
An awkward contradiction exists with regard to salary scales for
extension agents. On the one hand, higher salaries might seem to be a
logical incentive for more diligent performance and to attract better
trained personnel. On the other hand, in some situations if there are no
countervailing pressures, such an approach can result in attracting
highly educated urbanites and thereby increase the economic and cultural
gap between farmer and extension agent. This can impede communication.
There is a similar contradiction with regard to the training of
extension agents. In the absence of effective management and suitable
incentives, more education can be counter-productive. Leonard found that
there is an optimal amount of training; extension agents with more
training may expect more promotions than realistic, and become frustrated
and less effective.2 Indeed, in some projects in Africa, farmers and
farmers' sons have been given only a few weeks of training and have been
effective in communicating specific information to other farmers..
While there are advantages in a centralized extension agency,
there are inevitable problems as well. No matter how well it is managed,
it is difficult to be sensitive to the extreme diversity of natural
environments. Thus, from this point of view, the risk of the training
and visit system or of employing people with a few week's training is
that despite short-term successes, these approaches may not encourage an
inherent capacity to learn or respond to changing requirements over
space and time. This may not be a serious problem in the plains of India
Uma Lele, p. 72.
2Leonard, pp. 121-22.
38
and Bangladesh, but might be absolutely critical in the environment
of mountainous Nepal.
Moreover, it is likely that a centralized extension system may
be dominated by urban and bureaucratic political interests, which may
not coincide with farmers' interests. If promotions are determined
at the center, it is almost inevitable that better, more ambitious
agents will gravitate to the center in their orientation, work, and
eventual domicile, and will not work directly with farmers.
One factor which influences the suitability of a centralized
extension system involves the characteristics of the crops or tech-
nology in question. Centralized systems may be well suited to a crop
for which research is reasonably advanced, which is highly sensitive
to field management, and whose quality is very important for marketing
requirements. Tea and tobacco are major examples, and in different
parts of the world highly centralized, intense, expensive extension
systems are oriented towards such single crops. (In Kenya, tea growers
have one extension agent for 120 farmers, and extension services cost
$18 per farmer. ) Centralization may also be important in commodities
where timeliness in marketing and processing are critical, such as
palm oil and fresh milk.
A good example of the potential value and problems inherent in
centralized, targeted extension programs can be seen in the experience
of India. From about 1969 through the early 1970's, efforts were made
to target integrated services to small farmers through the Small
Uma Lele, p. 64-69.
39
Farmers' Development Agency (SFDA) and to landless and near landless
farmers through the Marginal Farmer and Agricultural Laborer Development
Agency (MFALDA). The Programs subsidized banks for credit services to
qualifying farmers, and provided inputs and information to expand irri-
gation, multiple cropping and animal husbandry.
Extensive problems were encountered.1 Subsidies to banks to
encourage them to loan to small farmers were notreally able to change
the modus operandi of the banks. Farmers sometimes did not believe that
the recommended buffalo, cattle, pineapples, etc., were suitable varieties;
animal and plant diseases took their tolls. Pettycorruption and bureaucratic
rigidities continued; bribes and gifts were needed to get eligibility
certified, to obtain application forms, to have them processed, to get
landlords to take some responsibility for their laborers, to get
veterinarians to certify the health of animals, etc. By the time these
direct expenses were added to indirect expenses of travel, an overnight stay
near a government office, and income foregone during the time spent on
such processes, the potential profit for the poor farmer from new economic
activities was seriously undermined. Administrators and local elites
2
managed to get a large share of the benefits. Finally, despite the
intentions of the program, it was not really able to concentrate on the
desired target group. Many large farmers were able to participate in
the program by dividing their land among children and thus claiming to
be small farmers.
See Biplab Dasgupta, Agrarian Change and the New Technology in
India (Geneva: U.N. Research Institute for Social Development, 1977),
p. 227-239.
Author's field observations, Andhra Pradesh, January 1976.
40
Nevertheless, the program had some successes also. Some smaller
farmers were able to buy tube wells, and many landless and near landless
were able to get into dairying, for which land ownership is not a pre-
requisite.
2. Decentralization
To guard against the dangers of centralized programs, many writers
advocate that extension systems be decentralized and controlled directly
by farmers. The logic of the decentralized system is that it will assure
that extension programs fit more accurately local needs, perceptions,
values, knowledge, and natural conditions. It is less likely to be
controlled by urban or bureaucratic needs.
There are several examples of such organizations. In most
states of the United States, county level farmers' organizations have a
major influence on the hiring, firing, and salaries of county extension
agents. In Taiwan, extension agents are hired by township farmers'.associa-
tions. Their salaries are directly related to the profitability of the
farmers' associations, and this is somewhat (but not entirely) related
to productivity of agriculture in the township.1 Thus, the agent is
presumed to have a strong incentive towards helping to increase local
farm productivity. In Finland, township councils control extension work.
In mainland China also, people who perform an extension function are hired
directly by communes and brigades. In Israel, farmers' representatives
make up an advisory board which reviews the annual work plan of the
extension system. In all these cases, there may be central guidelines
concerning salaries, educational qualifications, etc., but farmers and
Benedict Stavis, Rural Local Governance and Agricultural
Development in Taiwan (Ithaca: Cornell Rural Development Committee,
1974), p. 93.
2
Raanan Weitz and Arshalom Rokach, Agricultural Development:
Planning and Implementation, An Israeli Case Study (New York: Praeger, 1968),
pp. 391-392.
local officials can have some influence on hiring, salaries, and programs.
More important, it is presumed that social pressure will be placed on
extension personnel to serve local needs.
Decentralized systems have their own types of problems. If
extension agent's salaries are determined locally, there may be wide,
unjustified differences in salary levels, that contribute to rivalry,
jealousy, and poor morale. More important, from an equity point of
view, it is likely that the wealthy regions will pay their agents more,
while the poor regions will pay less; thus the regions which need
energetic agents most desperately may be unable to attract the best
agents. Both in Taiwan and the U.S., there are strong pressures to
reduce local salary differences of extension agents to permit more
equality and stability in the career ladder of extension agents.
Another problem with the decentralized system is maintaining
high levels of technical competence in the extension agents. Local
people may not realize the types of innovations that are possible, and
therefore choose extension agents with only limited capabilities.
Likewise, extension agents in such a situation may feel obligated
to spend all their time visiting farmers, and may fail to reserve
some time for continued technical training. In mainland China, the
extension system was under central control from 1962 until the early
1970's and it would not be surprising if centralizing influences
reemerge in the next few years as China stresses scientific and tech-
nical advance.
The most serious problem with decentralization and local
farmer control is that it does not fully or automatically solve equity
problems and may even aggravate them. If the rural sector is highly
42
inegalitarian and power is concentrated, it is likely that farmer
control will mean control of the extension system by large farmers.
Small farmers, tenants, and laborers, may not benefit. In the south-
ern U.S., for example, it is difficult to see how local control over
extension by (white) farmers could have helped (black) sharecroppers.
Land owners were eager to mechanize cotton and the sharecroppers
were forced to migrate. Given the character of local political,
judicial, marketing and credit structures, the black sharecroppers
could have been protected only with strong, countervailing central
pressures.
Analogously, males may use an extension system to extend male
control in certain parts of the economy where females had strong
claims. Particularly in Africa, woman commonly manage farms
and do much of the marketing. In Haiti, too, women are fre-
quently farm managers. However, it is also common that extension
systems are staffed and controlled by men, who give inadequate ser-
vices to the women farmers.1 Such a system is ineffective at best,
and probably reduces the relative role of women.
Hence, the crucial question is always: to whom is power to
control an extension system decentralized? Sometimes central power
is useful to control local elites (defined by wealth, age, sex,
caste, ethnicity, etc.), and to facilitate more equitable participa-
tion in local organizations of more numerous but less powerful
people, such as low caste groups and women.
1
For a detailed analysis of this situation in Kenya, see
Kathleen Staudt, "Women Farmers and Inequities in Agricultural Ser-
vices," Rural Africana 29 (Winter 1975-76), pp. 81-94.
43
Even in egalitarian situations, decentralized farmer control
cannot automatically assure-equity. It is -likely that extension
services will be used by a few aggressive farmers, will enable them
to expand, and will contribute to an inegalitarian situation. Exactly
how this happens depends on the details of technology involved; but if
there are economies of scale in profitable technology and if invest-
ments are lumpy (e.g. tube well, tractorization), tenant eviction is
likely. It is wrong to expect a decentralized extension system to
prevent this, unless special programs are undertaken. Another problem
with local control is that it does not solve the overall economic prob-
lem of advanced agriculture, namely excessive production leading to low
prices, unless it can organize farmers to limit production
voluntarily.
Of course, the ideal system would be to combine, somehow, the
strengths of a carefully managed, centralized system with the flexi-
bility and responsiveness of a decentralized system. This is com-
plicated and difficult, but it is precisely this combination which
some observers believe has been achieved in the modern American cor-
porate structure, which allows decentralized structures to make their
own operating decisions within the context of centralized strategic
and financial control.
One structural device recommended by an organization special-
ist is to have some individuals (e.g. farmer group leaders or second
level administrators) function simultaneously in two levels of
Alfred Chandler, Jr., Strategy and Structure: Chapters in
the History of American Industrial Enterprise (Cambridge: M.I.T.
Press, 1972). Cited in Hans Binswinger and Vernon Ruttan, Induced
Innovation (Baltimore: Johns Hopkins University Press, 1978), p. 333.
organization. This heightens their ability to exchange information
and feedback in both directions. An individual with such a dual role
can be considered a "linking-pin," as he links two organizations.1
In practice, the precise balance between central and local
control will change as some problems are reduced and others appear
more pressing. An example of the types of changes that are needed
can be seen in the farmers' associations of Taiwan. To assure that
they were controlled by farmers, the general manager of each local
association had been elected by the association's board of directors.
However, this election process became entangled in local factional
politics. When one faction controlled a farmers' association, it
tended to divert benefits to its members; members of other factions
were reluctant to participate. To reduce these problems the govern-
ment decided in 1974 to appoint directly the general managers of
local farmers' associations. This "solution" will, of course,
eventually recreate the original problems of excessive centraliza-
tion, and at that time, perhaps a new "solution" will be needed.
If extension systems are expected to serve the rural poor,
perhaps the most important factor in the centralization-decentrali-
zation issue is a political analysis of the balance of forces at
central and local levels. If the central political system has
important political reasons to improve the conditions of the rural
poor, and if the local political systems are highly inegalitarian,
then a more centralized system may serve the rural poor better. If
Allen Jedlicka, Organization for Rural Development, pp.
98-108.
however, the central government is more concerned with an agricultural
system geared to urban needs, and if the countryside is reasonably
egalitarian, decentralization may serve the needs of the rural poor
better. Finally, of course, where the central political system is
concerned primarily with urban needs, and where this is closely
linked to a strong rural elite, it may be unrealistic to expect
any extension system by itself to be particularly beneficial to the
rural poor.
3. Modes of Communication
Effective and efficient communication of information to
target audiences is, of course, a critical aspect of extension, and
much discussion about extension naturally has focused on this topic.
Two types of issues seem particularly important in this regard,
namely, the relationship between the content and form of the message, and
the extent to which'the form of the message generates an inter-
active process with the audience.
a. Content and Form
A wide range of communication media exists: newspapers,
radio, television, cinema, film strips, tape recordings, books,
magazines, posters, flannel boards, theater, puppet shows, group
discussions, individual discussion. Various practitioners have found
different forms particularly useful. Probably the easy mistake is to
believe that one particular form of communication is best. Undoubt-
edly, the suitability of a mode of communication is related to many
environmental factors, including the cultural context (for example,
is theater popular? are people literate?) and level of economic
development.
46
The form of effective communication is also related to the
content of the message. Obviously, different forms of communication
are required for these different functions. To use Arthur Mosher's
categories cited above, mass media may be well suited to serve as an
"encouraging companion" or as a stimulator of general development,
but individualized instruction may be needed to provide specialized
information on complicated technical questions. Another form of
communication--based on socially adept group organizers and inter-
action communication (described below)--may be needed to spark
group activities.
The need to relate the form of communication to the social
and economic environment and to the content of the message has been
confirmed in he Basic Village Education Project, in Guatemala. In
this experimental project, various combinations of mass media (radio),
farmer meetings organized by a monitor, and individual technical in-
struction were tested. The results of the experiment were:
...there is no single most effective media combination for
all situations.
The potential effectiveness of the various media combina-
tions varies with the level of development, the economic
well-being, and the present and prior exposure to mass media
and technical assistance. For an area relatively advanced...
radio alone will be immediately used as a source of new
information...In contrast, the full radio-monitor-agronomist
media combination is required...in an area rating relatively
low...In the traditional...areas, radio is capable of intro-
ducing new agricultural ideas and reducing the fear of
implementing them. However, reinforcement by agronomist
and/or monitor is needed to maximize impact of radio as an
information source...
The Basic Village Education Project, Guatemala, Final
Report (Washington, D.C.: Academy for Educational Development
1978), p. ii.
47
b. Interactive Communication
Communication can be more than the transfer of particular
information. It can catalyze an interactive process through which
the recipient, individually or in conjunction with others, develops
new thought processes and new patterns of social interaction. This
is, of course, the basic purpose of education (in contrast to
training).
48
The main purpose of group literacy programs in the style of
Paulo Freire is not simply to train people in literacy, but to educate
them to their individual and group potential. Freire calls this
process "conscientizacao." : Experiments involving
group participation in making film or video tape indicate that this
process also can contribute to the emergence of new perceptions and
patterns of social interaction.2 Similar changes can occur when people
are asked to create a play, or presumably a radio program.
There is some evidence that new social processes can be created
even by the more passive process of having groups organized to watch
collectively (or state) owned television sets, which can include agricul-
tural extension programs. Apparently, in the presence of the new technology
of television, traditional cleavages can be overlooked, at least in the
short run. In India, it has been observed: '
One of the most notable offshoots of (Satellite Instructional
Television Experiment Project) SITE has been the role of TV as
a social equalizer. It was common at the initial stages to
see different sections of the society watching the programs
while sitting in distinct groups. But slowly these disappeared
and it was no longer possible to distinguish "big" farmers from
"small" farmers, harijans from non-harijans, or educated farmers
from the illiterates."
Radio listening groups may have similar social benefits.4 Of course, in
Paulo Freire, Pedagogy of the Oppressed (New York: Seabury, 1970).
W. Anthony Williamson, "The Fogo Process in Communication," in
Training for Agriculture and Rural Development (Rome: FAO, 1975), pp. 93-98.
Lal Karamchandani, "Television for Rural Development, Indian
Experience with SITE," Training for Agriculture and Rural Development
(Rome: FAO, 1976), p. 134. See also Shingi and Mody, "The Communications
Effects Gap," p. 93.
F.M. Ragheb, "Training and the Green Revolution," in Training
for Agriculture and Rural Development (Rome: FAO, 1975), pp. 1-9.
all such matters one must wonder if traditional social patterns reassert
themselves after a while, after outside observers, who sparked the inter-
active process, leave.
For both, technical learning and broader education, practical
training and direct farmer-to-farmer interaction can play extremely
valuable roles. Farmers often can learn many things (such as use of new
machinery, identification of insects, analysis of characteristics of
new varieties) more thoroughly and rapidly in the field than they can
in the classroom, even with the best of audio-visual equipment. Often
the best, most trusted teacher is another farmer who successfully uses
a new technique. Indeed, one of the dangers of audio-visual aids is that
their use might discourage a training/education program from using field
trips and practical demonstrations. Of course, practical work can be
overdone. In China there have been reports that physical labor in the
agriculture schools (or at least one model school that is being
criticized) was so extensive that students had no time for theoretical
training.
4. Sources of Extension Information
An extension system with inaccurate or irrelevant information is
worse than a financial drain. It poisons the farmer's perception of modern-
ization and reinforces reluctance to try new techniques. Unfortunately,
it is not unusual for extension agents to offer erroneous and contradictory
information. Farmers who follow their advice find their yields may increase,
but insect losses may increased the next year. Or they may find two extension
Gillian Hart, "Labor Allocation Strategies in Rural Japanese House-
holds," PhD. Disertatibn., Cornell Agricultural Economics, 1978, Chapter 4.
agents both confidently encouraging very different planting dates for
cotton. Or they may find recommended inputs unavailable, or marketing
opportunities inadequate.
What are reliable sources of accurate, relevant information for
an extension system? There are two major sources of information: the
scientific community (both domestic and international) and the farmers
themselves. There are, however, vast problems in organizing regularized
communication with either group, particularly from the farmers. In theory,
an extension organization is supposed to be a bridge or link between
scientific researchers and farmers, providing two-way communication and
"feedback." In practice, however, most extension systems emphasize
dissemination of information from scientist to farmer, and do not adequately
carry information from the farmer to the scientist. In Kenya, for example,
Leonard has found, "The processes of feedback and technical innovation
have proved weak in the Ministry of Agriculture, particularly below the
national level."2 Upward communication is difficult because of the geographic
dispersion of extension agents, social (and sometimes racial) barriers
that exist between junior and senior staff, and an intellectual inability
and bureaucratic reluctance of junior staff to suggest changes in general
3
recommendations.
When extension agents fail to provide feedback from the smaller,
poorer farmers, a class bias emerges in the feedback process into the
research system. New technologies are "induced" by the specific requirements
Uma Lele, p. 65.
2Leonard, Reaching the Peasant Farmer, p. 160.
3bid., p. 162
51
of the economy. However, this induction is not automatic. Induction
requires institutions, of which an extension system is one, to convey
specific technological needs of various producers to the researchers.1
If the extension system does not provide feedback about the needs of
small, subsistence farmers, it is unlikely that appropriate technologies
will be induced for them. In the absence of such pressures, the innovations
that are induced are more likely to be suited to the needs of large
aggressive farmers, who can influence the research system directly by
going to a university to talk with scientists or indirectly, through a
Ministry of Agriculture to which they have access.2 The farmers with
cash crops crucial for foreign exchange earnings (rubber in Malaysia,
cocoa in Ghana, coffee, sugar and coconuts in the other countries) seem to
have little problem conveying their needs to research institutes. But
the perspectives and needs of the poor, limited-resource, subsistence
farmer, unable to purchase machinery and chemicals, rarely inform the
work of research units. Thus, any extension system, and especially one
designed to serve the needs of rural poor, needs to emphasize feedback,
particularly from small farmers.
The bias towards overemphasizing communication to the farmer,
instead of from the farmer, has several sources. First, it reflects
the structure of control within the extension system. If the
extension system is controlled and financed by farmers, the
agents usually are responsive to interest and needs of farmers,
1Alain de Janvry, "Social Structure and Biased Technical Change in
Argentine Agriculture," in Hans Binswanger and Vernon Ruttan, Induced
Innovations (Baltimore: Johns Hopkins University Press, 1978), p. 311.
2Rene Benalcazar R., "New Techniques, Agricultural Extension Services."
and are quick to communicate upward the farmer's concerns. If, however,
the extension system is controlled by the central government to develop
specific crops for urban consumption or export, the extension system will
see its task is to convince farmers to try specific practices, and not to
understand the farmers' perspectives.
As extension services become specialized, they are staffed with
educated people. In many countries, very few rural residents have access
to formal education, so extension services may become staffed with people
with urban backgrounds. There is a tendency for the urban educated
people to presume an ignorance and hostility to innovation among the back-
ward, traditional farmer.1 This justifies a program designed to instruct
and lead the farmer. This tendency is reinforced by general bureaucratic
compulsion to justify rather high salaries of government employees
relative to farmers. Of course, this bureaucratic interest merges with
broader social forces eager to justify the privileges of the educated
elite.
Ironically, some of these negative tendencies of an extension
system may be obscured and even reinforced by a highly energetic, patriotic
spirit within the system. However, at issue is not the dedication,
probity, or aspirations of extension agents--which can be highly variable
and very important--but rather their underlying attitude to farmers and
their conception of their roles in relation to farmers.
In the U.S. these types of problems were minimized by the recruit-
ment procedures of extension programs, which essentially require that
Rene Dumont, "Training for Rural Development, the Gulf between
Farm and Town," in Training for Agriculture and Rural Development (Rome:
FAO, 1976), pp. 15-17
extension agents came from farm families. This ensured some underlying
attitude of respect to the farmers, and reduced the likelihood of a social
gap between agent and farmer.
The normal training programs for extension agents can easily
reinforce the notion that extension agents teach farmers. They are taught
how to convey information to the farmers--how to use demonstration plots,
local fairs, mass communication, audio-visual aids, felt boards, tape
recorders, etc., and review research on effective communication techniques.
They also learn how to select "informal leaders," who can pass information
on to others. These various communication techniques are important, and
proper selection among them can do much to assure that information reaches
the poor. However, only in a few places are extension agents taught how
to learn from farmers and how to convey information from farmers to scientists.
In reality, some farmers have a great deal of technical knowledge,
vast experience, and keen insight into agricultural questions, as many
extension agents have discovered. Moreover, in many cases, "spontaneous
extension" systems function very effectively, and profitable technologies
spread rapidly as friends and relatives exchange information, and as merchants
and salesmen buy produce and sell inputs.1 These characteristics of a
rural community render the conventional extension role of transfer of infor-
mation from scientist to farmer substantially superfluous in many cases--a
fact extension systems sometimes realize but must obscure to protect their
In the U.S., with a far better developed commercial infrastructure
than most developing countries, the formal extension service is the first
source of information for farmers only 15-40 percent of the time. In one
survey only 14 percent of farmers considered extension to be the most
reliable source of information. Win M. Lawson and Howard M. Dail, "Sources
of Information for Farmers," Journal of Cooperative Extension 4:3
(Fall 1966), pp. 163-168.
own claims to be helping development and therefore deserving of more
resources.
These facts highlight alternative (or additional) roles for exten-
sion systems. If extension agents can learn superior techniques from the
most advanced farmers, they can then help spread this information to other
farmers and to researchers. Effective feedback from farmers can certainly
indicate which innovations from researchers are relevant to farmers'
needs and what sorts of further improvements are needed. Feedback is
essential to researchers to learn how an innovation 'fits -into the farmers'
overall farming system and seasonal labor constraints and opportunities.
Moreover, specific superior varieties and cultural practices of advanced
farmers can often be brought directly into research programs. One obser-
ver sums up this perspective:
Extension workers learn from progressive farmers what to tell
others. In fact, much agricultural development in such
countries as Holland can be explained by this mechanism of
locally originated innovation rather than by the utilization
of agricultural research station finding.1
Similar observations could be made about the U.S., where farm bureau
agents originally saw their task as facilitating exchange of ideas among
farmers, and not conveying new ideas from research stations to farmers.
This would suggest that the early U.S. and European information exchange
systems might have relevant lessons for current needs elsewhere. Feedback
is also needed to designers of extension communications, so they will
know how to plan effective radio programs or other formats for communication.
Niels Roling, Joseph Ascroft, and Fred Wa Chege, "The Diffusion
of Innovations and the Issue of Equity in Rural Development," in Everett
Rogers, ed., Communication and Development: Critical Perspectives (Beverly
Hills: Sage, 1976), p. 68, citing observations of Professor A.W. van den
Ban, Agricultural University, Wegeningen.
55
Sometimes a preliminary requirement for effective feedback is
that farmers become aware of their own knowledge and confident in express-
ing themselves in the presence of government officials. In many cases,
rural people have been so badly repressed for generations that this emer-
gence of consciousness and confidence is a difficult process. Paulo Freire
recommends that this process can be facilitated in a small discussion group
where people give reassurances to each other. A skilled leader can use
a dialogical process to help people discover and express what they know.
Freire suggests that adult literacy programs are particularly effective
vehicles for this process.1
Other types of activities can serve a similar function. In a
pilot project conducted in a Tanzanian village, a dialogical process with a
group of villagers uncovered how much they knew about methods for grain
storage. When various traditional techniques were combined with some
modern ones, low-cost but effective grain.storage systems were developed.
In principle, once villagers have discovered that they have the power
to solve one set of problems, such as grain storage, they may apply this
knowledge to other problems. Thus, it would be most interesting to do
follow-up studies in the villages where this pilot project was conducted,
to see if farmers there are using group dialogical processes to solve
other community problems. A report of a somewhat similar project in
Ecuador, also based on Freire's principles, indicates a widespread expansion
of community activities, such as schools, bus service, rural electrification,
Ipaulo Freire, Pedagogy of the Oppressed.
2Appropriate Technology for Grain Storage. (New Haven: Economic
Development Bureau, 1977).
roads, bridges, night guards, running water, drainage, etc.1
One of the most deliberate, systematic attempts to incorporate
the knowledge of advanced farmers into the extension system was developed
in Japan in the 1870's when the government wanted to modernize agriculture
but realized that the large-scale farming practices of the U.S. and
England, utilizing large machinery, were irrelevant.2 In 1878, the govern-
ment appointed one or two leading veteran farmers in each prefecture to
serve as an Agricultural Correspondence System. These veteran farmers
would gather detailed reports about local agricultural techniques and
conditions, send them to the government, receive suggestions from the
government, and organize local Agricultural Discussion Societies to enable
all farmers to share the information. In 1881, a national organization
for veteran farmers'was established, the Agricultural Society of Japan.
Four year later, an "itinerant instructor system" was organized, utilizing
both graduates of agricultural colleges and veteran farmers. Veteran
farmers were crucial in staffing the system until 1889, when sufficient
trained people were available to staff the system. A few years later,
the itinerant instructor system was merged with the newly formed Prefec-
tural Experiment Stations, and in 1899 many extension services were
incorporated within compulsory farmers' associations, which received
Edgardo Rothkegel Ortuzan, "The Ecuador Non-Formal Education
Project, in Richard Niehoff, ed., Non-Forward Education and the Rural Poor
(East Lansing: Michigan State University College of Education, 1977),
p.p. 111-120.
2Takekazu Ogura, Agricultural Development in Modern Japan (Tokyo:
Fuji Publishing Company, 1967), pp. 301-303: Toruzo Tatsuno and Reiichi
Kaneko, Agricultural Extension Work in Japan (Tokyo: Agriculture, Forestry,
and Fisheries Productivity Converence, 1959); and Ron Aqua, Local Institu-
tions and Rural Development in Japan (Ithaca: Cornell Rural Development
Committee, 1974).
some state subsidy. From 1911 to 1935, the agricultural associations
hired 5,200 to 14,000 technicians. After 1948, extension services were
consolidated under direct government sponsorship.
In China, also, major efforts have been made to assure that the
research and extension system absorbs the insight of ordinary farmers.l
Advanced farmers have travelled to other localities to explain their
techniques and have joined research stations to demonstrate and test their
methods for high production. At the same time scientists have been
posted periodically to work directly with farmers in local testing
stations which are sponsored by communes, brigades and teams. These
local stations are staffed with a combination of young school graduates
and mature, advanced farmers.
While these two experiences suggest some ways of maximizing farmer
input for extension services, there are certainly far more approaches which
can be tried. In the Basic Village Education Project in Guatemala, the
monitors who organized group meetings to discuss radio broadcasts provided
weekly feedback reports to the producers of the radio programs. It is
unlikely that the radio broadcasts alone, without this organized feedback
to assure relevance, would have been so effective.
Another method of providing feedback is being tried in Chile,
in which various extension agencies will compete to provide extension
services. The government gives vouchers to farmers, which they use for
the extension agency of their choice--the regular government system or
Benedict Stavis, Making Green Revolutions: The Politics of
Agricultural Development in China (Ithaca: Cornell Rural Development
Committee, 1974), pp. 172-89, and "Agricultural Research and Extension
Services in China," World Development 6:5 (May 1978), pp. 631-45.
2Nesman, "Basic Village Education," p. 124-125.
an extension system attached to a seed or fertilizer company, to a bank,
etc. The agency can turn the voucher in to the government for cash.
Another feature of the system is that the extension systems are required
to prepare detailed financial analyses of farm accounts to identify
profitable farms and farm activities. In Guatemala, a careful review
of farmer tests is being conducted by the Institute of Agricultural Science
and Technology to assure feedback.
While farmers are knowledgeable, they are not omniscient; nor
are scientists and highly trained subject matter specialists irrelevant.
Indeed, precisely because farmers are knowledgeable, the information
that they demand is often highly technical and very specialized. Sometimes
farmers ignore extension agents not because they are unwilling to change,
but because they suspect that the extension agent does not have suitably
specialized technical information. In the Indian Punjab, for example,
progressive farmers go directly to researchers at the agricultural university,
In Finland farmers demand personalized suggestions from animal nutrition-
ist and forestry management specialists.
Thus, an extension system that learns from farmers must also
include highly trained subject matter specialists. In both Japan and
China, the policies of expanding farmers' inputs have not prevented the
development of professional extension systems using academically trained
personnel. In Japan, by 1889 professional extension agents had replaced
veteran farmers as "itinerant instructors." In China, ideological
pressures against professionalization have been strong, particularly in
certain years (e.g., 1958), but academic training in agriculture has
continued, and graduates have been constantly placed in research and
extension units. China's new policies adopted after 1977 stressing the
importance of science, technology, the value of rigorous formal training,
and the benefits of farm management will most likely lead to a profes-
sionalization of the extension system. China's experience, in particular,
not only offers innovative approaches to assuring farmer input. It also
highlights the difficulties of doing this while developing a professional
extension system capable of dealing with very sophisticated, specialized,
information. A similar lesson emerges from the Guatemala project, where
better results were obtained by backstopping the monitor with an
agronomist.
Communication between extension and scientific researchers is
not the responsibility of extension personnel alone. The organization
and values of the agricultural science research units--which are shaped
forces that control research--influence how .mucIJscientists are receptive
to feedback. Research and extension systems have different goals, and are
subject to the control of different ministries. Sterling Wortman warns:
Too often, scientists as well as extension leaders consider the
activities leading to adoption by farmers are not the responsi-
bility of the research establishment. Until th s erroneous
idea is overcome, progress will be slow indeed.
The difficulties inherent in this problem should not be minimized.
Scientists often prefer to work in laboratories or nearby experimental
1Edgar G. Nesman, "The Basic Village Education Project: Guatemala,"
pp. 121-131.
2Sterling Wortman, "The Technical Basis for Intensified Agricul-
ture," Agricultural Development, Proceedings of a conference sponsored
by the Rockefeller Foundation, April 23-25, Bellagio, Italy (New York:
The Rockefeller Foundation, 1969). Cited in Burton Swanson, Coordin-
ating Research, Training, and Extension," in Training for Agriculture
and Rural Development (Rome: FAO, 1976), p. 9
plots, developing insights of interest to their professional colleagues
in universities and international research centers. This seems the most
promising strategy for higher professional, social, and economic status.
In contrast, conducting extensive on-farm trials, spending much time traveling
in the field, and talking with extension agents and farmers is less attrac-
tive to scientists and may present a serious role conflict.1
Coordination and feedback between research and extension is also
related to the level of basic physical and institutional infrastructure.
Agricultural scientists often feel their job is done when they have
developed a superior seed variety, or ascertained that a particular
chemical can deal with particular pest. Extension agents may be eager
to popularize the seed or chemical. But thetaskof multiplying the seeds
or manufacturing or importing the chemicals is not done by either the research
or extension organization. It can easily be overlooked or mismanaged, with
the result that researchers are annoyed that their suggestions are not being
adopted, while extension agents are frustrated that they are not getting
inputs from research.
There is much room for experimentation and innovation with regard
to policies that can maximize scientists' receptivity to feedback. A
general value system and specific incentives which support and reward
scientists who orient themselves towards farmers may be useful. The
challenge is to do this without cutting a country off from international
scientific exchanges--which tend to stimulate and reward scientific
excellence and simultaneously break feedback loops with farmers.
Swanson, "Coordinating Reasearch . .," p. 10.
60
Structural innovations may be relevant also. China has taken the
radical step of essentially merging their research and extension systems
into a multi-level network of experimental stations, which are staffed to
varying degrees by scientists, local educated youth, administrators and farmers.
These stations conduct experiments at all levels, from genetics at central
stations to field tests at local levels. Extension personnel are not
specialists in conveying-information, but are links and participants
in the experimental and learning process between and with farmers and
scientists.1 In the Puebla Project in Mexico, research and extension
personnel, although retaining distinct functions, worked closely together
on a day-to-day basis, and this was considered crucial to the project.2
In the U.S., many individuals in land-grant universities and elsewhere
have one portion of their salaries and responsibiJities earmarked for
research activities and a separate portion specified for extension
activities. This also contributes to close linkages between research
and extension.
Naturally, if extension and research systems emphasize feedback
from farmers, there are important implications for agricultural education
programs. A closer link between farm and school is needed. Farm workers
might be admitted directly into higher education (after appropriate literacy
training); students might be given plots of land to work at the beginning
of their studies, and grades could take into account general attitudes
1Stavis, "Agricultural Research ..."
2Swanson, "Coordinating Research ...," p. 10.
61
towards farming and helping farmers.1 Some steps are being taken in
some places (Nepal, Malawi, Malaysia, etc.,) to do this.
H.K.F. Hoffmann, "University-level Education in Agriculture,
A perspective for the Year 2000," in Training for Agriculture and Rural
Development Rome: (FAO, 1976), pp. 50-51.
E. Group Organization
Throughout the discussion of centralized organization, decentralized
control, modes of communication and feedback lies the question of the degree
to which extension systems should help farmers bring about group action. In
some philosophies of extension, the notion of group action seems useless.
If an extension agent is going to work with progressive individual farm
managers and help them develop specific farm plans, then he must work with
individuals; groupings of farmers are irrelevant. Similarly, it would
appear, if the extension system emphasizes mass communication (e.g. radio,
newspaper) which reaches each individual directly, local organization would
seem unnecessary.
On the other hand, organizing farmers into groups can greatly sim-
plify many of the tasks in agricultural extension and development. Working
with a group enables the extension agent to reduce the equity problems
inherent in working with the large progressive farmers. He need not risk
the administrative problems of working individually with large number of small
farmers. He can also get around the problem of working with progressive
people who are socially isolated in the village and whose innovations
will be shared by others.
Group approaches to extension serve a broader purpose also. There
is a wide range of innovations in agriculture which require social interaction.
There are three reasons for this. The first is the issue of economies of
scale. Many activities, such as marketing, have important economies of scale.
The second issue is the free rider problem, i.e., how to make sure that every-
one contributes his share to building or maintaining collective goods, such
as roads or irrigation canals. The third is the external diseconomy pro-
blem, e.g., how to make sure that one person's rational actions do not harm
other people.
In the latter two situations, some coercion and/or special indivi-
dual incentives are required to enforce contribution and participation, or
to prevent some action. Typically, coercion is applied through the police
Mancur Olson, Jr., The Logic of Collective Action (New York:
Schocken, 1968), p. 2.
and courts of the state apparatus. In many cases, however, the state does
not have these resources, or they are used in an arbitrary, patronizing, and
corrupt manner. There can be great advantages if the coercion and incen-
tives required by development can emerge voluntarily as social pressure from
organized social groupings, and at least partially substitute for state
power.
To give some concrete examples, groups can generate social pressures
that encourage repayment of credit.1 In some villages in Java, village so-
cial pressures have been successfully utilized to encourage adoption of
family planning. Extensive publicity about family planning generated a mood
of acceptance, and publicly listing non-acceptors created some social stigma.
2
With such community pressure and gossip, most people decided to participate.
Similar experiences are reported in China.3 Such social pressures can equally
apply to agricultural innovations.
Group action is useful in many other activities. If extension acti-
vity is concerned with improved public works, maintenance of irrigation
systems, or distribution of irrigation water, some group action by farmers
is essential. Group organization can be very useful in creating new marketing
channels. They can both save farmers the time and energy inherent in in-
dividual marketing, and can offer competition to traders. In many situations,
marketing cooperatives are the easiest to set up because these benefits are
Dale Adams, "The Economics of Loans to Informal Groups of Small
Farmers in Low Income Countries," Department of Agricultural Economics
and Rural Sociology. Ohio State University, mimeo, 1978.
Richard Critchfield, "More Food..." p. B 12-13.
Leonard Chu, Planned Birth Campaign in China, 1949-1976
(Honolulu: East West Center Communication Institute, 1977), p. 41.
so obvious. If tractors, threshers or other machines are sensible, group
purchase and utilization can assure that benefits are widely shared and
can reduce the likelihood that the first individual purchaser of machines
will use his new profits to buy out his neighbors, who can no longer com-
pete with him. If improvements in livestock management are suggested --
such as regulated grazing and breeding or control of communicable livestock
diseases -- cooperative action by many farmers is needed. Group action by
farmers may also be useful in controlling plant diseases and pests, in-
cluding rodents. For improvements in general sanitation, which are crucial
for improved human health, some collective actions (or at least restraints)
are needed. For these types of activities and others, group organization
can be very helpful, and extension work can both be facilitated by groups
and help form the groups.
At a broader, more political level, when farmers are organized in
groups, they are more likely to exercise power over the personnel and
policies of the extension system.
When organized, the farmers are better able to fund local extension
activities and to participate in local experiments by contributing land
and their own knowledge.
At the broadest, philosophical level, the animation rurale projects
in former French colonies are "based on the belief that man is by nature a
social animal who finds individual fulfillment through participating in
activities which lead to the development of his community." This view
has its roots in both Catholic humanism and African socialism.
Jeanne Marie Moulton, Animation Rurale; Education for Rural
Development (Amherst: University of Massachusetts Center for International
Education, 1977), p. 20-21.
One of the most important potential benefits of organizing farmers
into groups is that it becomes feasible to train individual farmers who are
selected by and responsible to a group of farmers. Such persons can be
extremely effective in interpreting new ideas to farmers and in bringing
farmers' problems and observations to the research system. Unlike govern-
ment-employed extension agents, the trained farmer stays in the village,
is not seeking promotion to urban areas, and is accessible at all times
to villagers. If this person is selected by a group of farmers, it is
likely that social pressures will increase his likelihood to share quickly
and fully information about modern technology with his neighbors.
From a financial point of view there are important advantages. Such
a person can be paid a salary relative to a farmer's income, not relative to
bureaucrats' salaries. Moreover, part or all of the local person's salary
can be financed by the farmers' group, perhaps through the profits of credit,
marketing, or grain processing activities, as the extension agents of Taiwan's
farmers' association are paid. In Finland, specialized voluntary associations
of farmers have been created to provide technical assistance in a wide range
of areas -- marketing and purchasing, cattle raising, forestry, management,
etc. Farmers pay for individualized specific services.l Both the lower
salaries and local, self-financing which are possible through this approach
can greatly reduce recurrent government expenditures on an extension system
/ and make it less likely that financial constraints will prevent expansion of
an effective system.
The training of representatives of groups of farmers has, in fact,
been the cornerstone of several extension programs. The extension program
1Nils Westermarck, Finnish Agriculture (Helsinki: Kirjayhtyna,
1969), p. 57-71.
at Comilla, devised by Akhter Hameed Khan, was based on this idea. At
Comilla, a Thana Training and Development Center was established to offer
a wide range of instructions to representatives (cooperative managers and
model farmers) elected by local groups of farmers. These representatives
came for instruction for a whole day every week or two, and then reported
back to weekly group meetings in the Villages.1
In animation rurale projects in former French colonies in Africa,
the same principle of having villagers select representatives to receive
special training has been utilized. These individuals are called "ani-
mateurs."2 In the Puebla Project in Mexico, paraprofessionals helped
maintain a link between extension agent and farmer.3 The World Bank's
T & V system is a bit different because the contact farmer -
is selected by the government, not elected by the villagers.
In any of these approaches, group organization can provide valu-
able economies of scale. It is far more efficient for an extension agent
to brief a group of farmers directly than to work with individuals.4 For
example, in Bangladesh,. if one hundred model farmers hear an explanation of
a Thana Extension Officer and then each one conveys this information to
thirty members of his village society, then the extension agent has reached
3,000 farmers in just one week. Likewise in the T & V System, a Village
Akhter Hameed Khan, Reflections on the Comilla Rural Development
Projects (Washington: American Council on Education Overseas Liaison
Committee, 1974), p. 17.
2jean Fauchon, "Integrated Rural Development and Planning for Rural
Communities," in Training for Agriculture and Rural Development (Rome:
FAO, 1975), pp. 79-85.
Swanson, "Coordinating Research...," p. 11.
4Leonard emphasizes the economy of communication to groups. p. 203.
See also Rene Benalcazar R.,"New Techniques,Ag. Exten. Services..," p. 523.
Extension Worker (VEW) can be working with eight groups of farmers, each
with 40-150 farmers, or a total of 300-1,200 farmers. A fully trained
Agricultural Extension officer would be responsible for eight VEW, so
would be overseeing extension activities of 2,500-10,000 farmers. In
the absence of group organization, it would be difficult for an agent to
reach effectively a fraction of these numbers. In the intensive cash-
crop schemes, an agent normally serves only a few hundred farmers.
Even in projects stressing agricultural extension via radio
broadcasts, it appears that supplementing the radio programs with weekly
group forums organized by a "monitor" added to the impact. When the
monitor is backstopped by an agronomist, the effectiveness seems even
higher.1
As suggested earlier, groups can also be utilized in the process
of helping people to discover their own knowledge and power.2 The group
may be for literacy training or grain storage (as suggested above) or may
be shaped around general agricultural technology needs; it might be radio
listening group or film making group.
There are, of course, problems inherent in the notion of organizing
groups. The theory of using groups generally assumes that the groups are
characterized by an internal cohesion, a sense of mutual obligation, and a
rough equality. It is presumed that because group actions are rational, it
is rational for every individual to participate in the group. These pre-
sumptions are frequently erroneous.
1Edgar G. Nesman, "The Basic Village Education Project: Guatemala,"
pp. 121-131.
2Paulo Freire, Peadogogy of the Oppressed.
It is remarkably easy to establish a group that lacks internal
cohesion. A group may be composed of people living in a particular
locality or along a river, or road. The size of the group may be ar-
bitrarily fixed according to the number of people likely to be served by
a warehouse, or according to the ideal economies of scale for a cattle
dip, tube well, or for farm credit. In such a case, it may be that the
indigenous patterns of social interaction and trust -- kinship, friendship, temple
membership, tool sharing groups, voluntary credit or funeral societies,
political factions, etc. -- may not be congruent with, and may be com-
pletely unrelated to the lines of delineation which seem rational to ad-
ministrators concerned with fertilizer, irrigation, range land management,
or animal health. They may accidently throw together into a group people
who are strangers, or who have been feuding for generations. Naturally,
such a group will not function in the desired manner.
Another problem occurs in hierarchical societies characterized
by extreme inequality and strong patron-client relationships. In such
a situation the social inequalities will probably shape the character of
the group. The powerful patrons or their representative will dominate the
group and use its resources to reinforce their positions. Information,
credit, inputs, and other resources expected to spread throughout the group
are likely to be commandeered by the rich and powerful. The group becomes
a new locus of profit, control, and corruption. These problems are, of
course, reinforced by the general tendency of power to concentrate in the
hands of leaders and organizers of groups -- the "iron law of oligarchy"
described by Michels. When all these factors are combined it is not in-
frequent that groups really become tools of exploitation for the rich,
and that the poor prefer to avoid the groups, whenever possible.
Finally, it may not be rational for an individual to participate
in group activities which will benefit him. Indeed, if he can avoid the
cost and obligations of group activities but share in the benefits with-
out undermining group functioning, he obviously is rational.
Whether and how these problems can be overcome depends on the
character of the group in question. Some successful group activities are
based on the indigenous voluntary groups which exist in rural societies --
kinship, religion, labor exchange, credit associations, etc. Size is im-
portant; particularly if the group is small enough so that each individual's
actions perceptibly influence the group's fortunes, the internal social
pressures for cohesive action can be strong.2 But the group needs to be large
enough to generate economics of scale. Perhaps the optimal size for such
voluntary interaction is around 30 35 people -- the size of the Chinese
work team,3 of successful midwestern food cooperatives,4 or of groups in
Mexico's Puebla Project.5
These social pressures were utilized at Comilla, where the credit
associations were required to have weekly meetings to maximize the face-
Mancur Olson, The Logic.
2Mancur Olson, The Logic..., pp. 53-57.
Carl Riskin, "Maoism and Motivation: Work Incentives in China,"
in Victor Nee and James Peck, eds., China Uninterrupted Revolutions From
1840-to the Present (New York: Pantheon, 1973), p. 431.
Ron Cotterill, The Social Economics of Participatory Consumer
Cooperatives (East Lansing: Michigan Sate University, Department of
Agricultural Economics, Agricultural Economics Report No. 369, December 1979),
p. 31.
5Allen Jedlicka, Organization for Rural Development, p. 74
to-face relations and social pressures they would generate. As supervision
at Comilla became relaxed and corrupt, as basic level groups were expanded
and reformed so that face-to-face relations in the group setting became
less important, the programs at Comilla became more controlled by the
elite farmers.1
It is often the case, however, that a larger organization is re-
quired, either to obtain potential economies of scale (e.g., for purchasing
trucks, installing electricity systems) or to include all potential "free
riders" (e.g., a large irrigation system). As the group gets large enough
so that an individual's actions no longer perceptibly affect the group's
fortunes, and as group social pressures are reduced, some combination of
personal incentives, threats of coercion, and external supervision seem
necessary to assure participation.2
Equally important, some organizational format is needed to tie to-
gether small face-to-face groups of farmers into larger units that can
provide appropriate economies of scale. In Taiwan, for example, the farmers'
associations are organized from smaller village-based groups. Moreover, they
have responsibility for tax collection and monopoly over fertilizer sales.
Thus, all farmers were compelled to deal with the farmers' associations, even
if they did not join. In addition, they had to join to take advantage of
credit, extension, and marketing services.
1Harry Blair, The Elusiveness of Equity: Institutional Approaches
to Rural Development in Bangladesh (Ithaca: Cornell Rural Development
Committee, 1974), pp. 45-61.
2Mancur Olson, The Logic.
3Benedict Stavis, Rural Local Governance and Agricultural Development
in Taiwan (Ithaca: Cornell Rural Development Committee, 1974).
Similarly, in Bangladesh, the two-tier cooperation system is being
established. In China, the commune has three levels to provide different
appropriate scales of operation; membership is compulsory. In the United
States, the system of having farmers' organizations sharing in the hiring
and control of extension agents involved a combination of federal subsidies
to the Farm Bureau organization and personal incentives. At first, farmers
were required to join the Farm Bureau to receive individualized extension
services and to participate in the insurance and marketing programs of
corporations controlled by the Farm Bureau. In general, the U.S. system
does not have smaller groups below the county level, but historically in
some places township organization has played an important role.
Frequently groups are set up to administer credit programs. Go-
vernments make continuation of loan programs to group members contingent
on repayment of past loans, and hope that the social pressures generated
by those who want future loans will force potential defaulters to repay.
This approach is used in Bangladesh.
An interesting variant is being tried in a World Bank credit pro-
ject in Malawi. To obtain a loan, group members are required to deposit
assets valued at 20 percent of the loans in a blocked saving account.
Shortfalls in repayment are deducted from this account, and then the funds
are returned with interest.2 Such a program presumably generates strong
social pressures for loan repayment, but may constrain small farmer parti-
cipation, as they may lack assets to deposit.
Mancur Olson, pp. 148-159.
Dale Adams, "Economy of Loans...," p. 5.
In Bangladesh, however, it seems that this approach is not adequate.
Frequently, the defaulters are the larger, powerful farmers who are immune
to village social pressures and may, in fact, have a vested interest in
the collapse of institutional credit, which might compete with their own
money-lending activities. In such cases, it would seem that the application
of police and judicial power, involving the seizing of mortgaged properties,
is necessary to supplement social pressures. Alternatively, different groups
might be set up of small, medium, or large farmers, so that members have
roughly equal status. In general, Adams reports that groups established for
the sole purpose of obtaining credit lack viability.
For landless laborers, group activities in the form of labor
unions can be extremely useful in improving not only wage levels but also
working conditions (including stability of employment, dignity of treatment,
etc.) This has been amply demonstrated in Kerela, India, among other places.
In some cases local groups will form spontaneously, as expansions
of previously existing voluntary organizations. This is most likely to
happen where land ownership is reasonably egalitarian and where government
is tolerant of local organization. The Grange movement in the U.S. and
the cooperative movement in Denmark are two examples.
In most cases, however, active government support for all these forms
of group activities is needed. Careful, thorough supervision and checking
from higher levels and the effective utilization of police and judicial
power seem necessary to nip in the bud the first indications of corruption.
Training programs are also useful to assure competence in organizational
management as well as technology (e.g., bookkeeping, etc.). Government regulations
that require groups to be composed of homogeneous farmers (i.e., all small,
medium, or large) can be helpful in preventing patron-client relationships
from being replicated within new groups. Active, powerful local groups do
not displace government roles, but rather place new, extensive demands on
government.
Local schools can play a useful role in assisting groups of farmers.
A teacher can provide some technical input; the school grounds can some-
times provide a site for a test plot, school children can become better edu-
cated for productive rural lives. The "peasant universities" of Scandinavia
are an excellent example. (The risk of special rural schools is the possi-
bility of institutionalizing a second-class education for rural residents.)
In addition, adult literacy programs can be encouraged because they strengthen
mass control over groups; an educated, literate populace will be more
capable of understanding the finances of the group, more confident about
participation, and more able to request government intervention to punish
corruption. Sometimes an interactive communication process, in which
people form groups to make plays, movies, radio programs, etc., can be
helpful in creating and reinforcing group dynamics.
Nourishing group activities is a critical and delicate task. Small,
effective groups can easily be smothered and crushed by rigid government
activities. At the same time, government encouragement, support, and
supervision are needed to assure a suitable balance of personal and group
incentives and to regulate the use of coercion. Large groups and organi-
zations cannot function effectively in a vacuum, and a government which
ignores groups will destroy them as effectively as a government which ri-
gidly controls them.
An example of a sophisticated, complex combination of policies to
ensure group functioning can be seen in rural China. Work teams are rea-
sonably small enough for each person's efforts to have noticeable impact.
Dumont, "Training for Rural Development...," p. 19.
2Rolland Paulston, Folk Schools in Social Change (Pittsburgh:
University of Pittsburgh Center for International Studies, 1974).
They are cohesive, often closely related to kinship groupings and other
pre-existing voluntary organization. Individual material incentives are
used to encourage participation. At the same time, collective incentives
exist and numerous political and ideological meetings reinforce group con-
sciousness. Government supervision and training is extensive, but when
government controls were too rigid, group cohesion suffered. Finally,
coercion is in the background with regard to people who contemplate excessive
private economic activities or migration away from the group.
The other problem with groups -- for which the above solutions will
not be relevant -- is that they will be opposed by some political forces
precisely for the reasons groups are advocated. Bureaucrats may not want
groups, which can place demands on their activities. The locus of power
may be moved away from easily satisfied bureaucratic supervisors to very
demanding farmers. The rural elite may not want to see an organization of
poor people which might be used to support demands for land reform and/or
higher wages. Merchants and moneylenders may fear that groups of farmers
may undermine their monopolies in marketing systems. Foreign economic in-
terests may be worried. Urban political actors may fear that when farmers
are organized, they will develop the political power to redress urban biases
in economic plans.
Such factors were behind the weakening of animation rurale activities
in Senegal after 1963.1 Such problems have also been noted in Niger,2 and
indeed may have virtually blocked the participatory aspects of the program
Moulton, Animation Rurale, p. 84-86.
2Dominique Gentil, "Les Cooperatives Nigerieness Traditions Vil-
lageoises et Modernization Cooperative," 1'Ecole Practique de Hautes Etudes
doctoral dissertation, 1971, and "Methodologie de 1'Implantation de Nouveau
Systems Cooperatif on Niger," Development et Civilizations 52-53, (April-
September 1973), cited in Moulton, p. 140-143.
from the beginning. In many countries of Latin America (such as Brazil
and Honduras) rural, urban, and bureaucratic elites have been highly sus-
picious of any groups in the countryside, especially if they are controlled
by poor farmers. Often these political instincts have a sound basis. In
many cases opposition political groups, including Marxist groups and
Christian Democratic Parties have, in fact, had vigorous campaigns to or-
ganize rural groups. They hoped these groups would support their political
movements in elections, in political demonstrations, or in other ways. To
organize such groups, political movements may promise land reform, higher
prices, or other benefits. Naturally this can frighten rural and urban
elites. Far-sighted elites may in fact see some reason for reform, but
not all will. Some will fear that any extension program that encourages
or tolerates groups controlled by poor farmers will be too dangerous be-
cause it could be used by political opponents.
Obviously these issues applied historically in the colonial systems.
There are sound political reasons for the failure of the colonial extension
systems in South Asia and Africa to allow real power to devolve to farmers'
groups. Conversely, it is difficult to imagine how the farmer acquisition
systems in the United States and Finland could have evolved if British or
Swedish and Russian rule had continued.
It is inevitable that the establishment of group activities
among farmers, which seems necessary for many facets of extension work,
Robert Charlick, "Power and Participation in the Modernization of
Rural Hausa Communities," University of California at Los Angeles Ph. D.
dissertation, 1974.
76
will involve complex political issues. Political leadership is needed at
the center to help urban, rural, and bureaucratic elites to see their own
long-term interests in increases in farm production and in the broad dis-
tribution of benefits.
Likewise at the local level, leadership is needed. In most cases
groups will not emerge spontaneously.. The animation rurale programs
recognized this need for leadership and had extensive government staff at
several levels, vehicles, training centers, etc., to provide a cadre of
organizers. Setting up effective groups can be a very difficult challenge
requiring dedicated, sensitive, patient, humble leaders and organizers. It
is easy and disastrous to smother peasant initiative and reinforce the notion
that the group is simply a tool for bureaucratic control. In the Tanzania
pilot project on grain storage, the team of organizers was sensitive to this problem:
First of all it was necessary to convince the villagers that
the outside team did not have a preconceived idea which it
had "up its sleeve" all the time just waiting for the little
drama of village democracy to play itself out... It was only
after having carried a certain line of design (the Nigerian
crib) forward in discussion for several weeks only to drop
it when the villagers brought up serious contentions, that
the same credibility was finally established. It was then
clear that the tea did not have a vested interest in any
particular design.
From where do such organizers come? Are they part of the indigenous
rural society? Alternatively can they really be trained, as animation
rurale presumed? This is often the
Moulton, Animation Rurale, p. 26.
2Appropriate Technology for Grain Storage, p. 41.
critical question, for if gifted organizers are not available, and if typical
bureaucrats do the organizing, it is likely that the groups will simply be
reflections of the bureaucracy and the dominant social forces in which it
operates; extensive participation is unlikely. Being a good organizer re-
quires not only training but also particular commitment and values. It is
frequently the case that people with these characteristics emerge from
political, religions, and student movements, which have strong value orien-
tations. If such organizers exist, will farmers be so suspicious of their
motives that they will be reluctant to join organizations? Will elites allcw
them access to the poor? Thusa. crucial aspect of group formation is whether or
not a government maintains a political and ideological position that engenders
support of potential group organizers, and acceptance by the people to be organized.
As with other issues, no uniform recommendation can be made about
using groups. Clearly they can be tremendously useful, but there are many
problems inherent in a group approach. Flexibility and sensitivity to
changing Relations is needed to find the best way of incorporating group
activities into extension programs.
F. Conclusions
In trying to use an extension system for equity purposes, the prob-
lems are complicated and many of the suggestions are mutually contradictory;
policy makers are not spared the burden of making difficult choices.
Moreover, the needs of countries and extension systems continually change,
as one set of problems is resolved and others become more salient.
In planning and improving extension systems, perhaps the first
step is to have realistic expectations of what is possible. Extension sys-
tems can make important contributions to development, but they cannot work
miracles or do everything. They can help people learn rigorous analytical
techniques to evaluate experiments, they can help provide inputs and deliver
credit. Of course, extension may not be able to do all these tasks simu-
ltaneously; a choice must be made about the priorities of these and other
potential roles.
However, extension programs by themselves in the absence of land
tenure reforms and vigorous, egalitarian.input supply programs, should not be
expected
/to reverse the trend towards concentration of assets in the rural society,
or to save the small, poor, or inefficient farmer. They can however,
assure that the small farmer is not disadvantaged with regard to access
to information. In some cases the extension system may have no choice
butto urge poor farmers to quit the farming profession. Nor can exten-
sion do much to strengthen the whole rural sector vis-a-vis the urban
sector. Such factors are deeply entwined with the whole political struc-
ture, and extension can affect these matters only marginally, and only when
conscious concerted efforts are made to do.so. It must be accepted that
the energy available in extension systems is usually very small compared
to the momentum of existing economic and social changes. If the energy
is focused in space and time (for example, on a small demonstration area)
some impact can be visible, some problems will become more apparent, and
some people will become better educated in these issues. But realism al-
lows only limited expectations about the social changes that can be im-
posed by an extension system; there is no point in criticizing small,
underfunded, inadequately staffed, politically weak extension systems for
inability to make major changes in social and economic structure.
The internal management of an extension system can be manipulated
for both growth and equity purposes. There are inevitably different
types of inequities at both central and local levels; someone concerned
with equitable development will try to design an interaction of central
and local forces which maximizes tendencies towards equity. But ultimately
a difficult choice will have to be made as to which level is least
inequitable, and the preponderance of control must be vested at that level.
Hence careful institutional analysis of the local and national polities is
essential in designing or modifying extension systems.
Another set of issues in improving extension systems is the
utilization of groupings of farmers. It is increasingly clear that
extension programs that reach individual, progressive farmers will have
high costs per farmer reached, and may aggravate problems of equity. If
extension systems organize the work with groups, the cost of information
dissemination can be much reduced; in addition, certain technical problems,
such as irrigation, can be handled more efficiently. There is a chance
that a larger portion of the rural population can benefit by getting
more equitable access to information, credit, markets, and machinery.
Most important, groups improve the potential for farmer control over the
extension system.
It is not, however, easy to set up groups. Sometimes they can
draw strength from existing voluntary organization; but if local culture
is static and highly inegalitarian, it is most difficult for local groups
to escape these tendencies. It is important that there are individual
financial and legal incentives to participate in group activities; some form
of penalties or coercion may be needed. A multi-tiered organization may be
needed to provide groups that are small enough for face-to-face interaction,
while providing the foundation for other organizations large enough to
provide suitable economies of scale. Innovative organizational strategies
are needed to deal with these dilemmas. Moreover, astute political
leadership will be needed to overcome opposition to having farmer groups.
At the same time, individualized instruction should not be discarded
entirely. As farmers get more advanced, they will need specialized,
individual assistance in making complex managerial decisions.
Special attention can be given to make sure the extension
system has the capacity to learn useful ideas about agricultural technology
from advanced, experienced farmers. Particularly when research is not well
advanced, this can be a useful source of relevant ideas which can be trans-
ferred to other farmers, and can be brought into more formal research
programs. In these situations, agents need special training on how to
learn from farmers, but this should not undermine the need for well-trained
agents who can give the farmers the technical sophistication they demand.
Undoubtedly, the correct combination of policies will vary from place to
place and can be ascertained best by trial and error.
Finally, it must be remembered that the formal extension system
is only one of many sources of information for farmers. The informal,
spontaneous extension process--in which farmers get information from friends,
relatives, merchants, etc.--is always very important. Thought should be
given to how the formal extension system can reinforce and take advantage
of these informal networks. How to utilize merchants is a particular
challenge. The interests of farmers and merchants do not coincide completely;
but merchants often provide much information to farmers about input supplies
and product markets.
It should be clear that many questions on agricultural extension
must be geared to the specific local environment. Mechanical transfer of
any country's extension system (whether that of the U.S., Japan, Taiwan,
etc.,) is filled with danger. Moreover, a country's needs change as some
problems are solved and other problems become salient. Closely monitored
pilot projects can be useful to generate accurate information about how
general ideas work out in particular locations. Making sure that the
pilot project is replicable and utilized properly is a major problem, but
there is no substitute for practical experience in situ.
MSU RURAL DEVELOPMENT PAPERS
RDP No. 1*
RDP No. 2*
Akhter Hameed Khan, "Ten Decades of Rural Development:
Lessons from India," 1978.
Lane E. Holdcroft, "The Rise and Fall of
ment in Developing Countries, 1950-1965:
and an annotated Bibliography," 1978.
Community Develop-
A Critical Analysis
RDP No. 3*
RDP No. 4
RDP No. 5
James E. Kocher and Beverly Fleisher, "A Bibliography on
Rural Development in Tanzania, 1979.
Enyinna Chuta and Carl Liedholm, "Rural Non-Farm
A Review of the State of the Art," 1979.
David W. Norman, "The Farming Systems Approach:
the Small Farmer," 1980.
Employment:
Relevancy for
MSU RURAL DEVELOPMENT WORKING PAPERS
RDWP No. 1
RDWP No. 2*
RDWP No. 3*
RDWP Np. 4
RDWP No. 5
RDWP fo. 6
RDWP No. 7
RDWP No. 8
Benedict Stavis, "Turning Point in China's Agricultural
Policy," 1979.
Kathryn M. Kolasa, "The Nutritional Situation in Sierra Leone,"
1979.
Benedict Stavis, "Agricultural Extension for Small Farmers,"
1979.
Steve Haggblade, Jacques pefay, and Bob Pitman, "Small
Manufacturing and Repair Enterprises in Haiti: Survey
1979.
Results,"
Peter Riley and Michael T. Weber, "Food and Agricultural
Marketing in Developing Countries: An Annotated Bibliography
of Doctoral Research in the Social Sciences, 1969-79," 1979.
Harold M. Riley and Michael T. Weber, "Marketing in Developing
Countries," 1979.
Victor E. Smith, Sarah Lynch, William Whelan, John Strauss and
Doyle Baker, "Household Food Consumption in Rural Sierra Leone,"
1979.
Omar Davies, Yacob Fisseha and Claremont Kirton, "The Small-
Scale Non-Farm Sector in Jamaica...Initial Survey Results,"
1980.
Single copies of the MSU Rural Development Papers and MSU Rural Development
Working Papers may be obtained.free .by writing to: MSU-Rural Development
Department of Agricultural Economics, 206 International Center, Michigan State
University, East Lansing, Michigan 48824; U.S.A.
*Out of print.
|
