MS Inentoa Deeomn aPapers
MSU IntenationalDevelopmnt Deparment ofAgrclua cnmc
Paper ~ I No*0Mcia SaeUiest
198 Eas Laning Miciga 4882
MSU INTERNATIONAL DEVELOPMENT PAPERS
Carl K. Eicher, Carl Liedholm, and Michael T. Weber
The MSU International Development Paper series is designed to further the
comparative analysis of international development activities in Africa, Latin America,
Asia, and the Near East. The papers report research findings on historical, as well as
contemporary, international development problems. The series includes papers on a wide
range of topics, such as alternative rural development strategies; nonfarm employment
and small scale industry; housing and construction; farming and marketing systems; food
and nutrition policy analysis; economics of rice production in West Africa; technological
change, employment, and income distribution; computer techniques for farm and
marketing surveys; farming systems and food security research.
The papers are aimed at teachers, researchers, policy makers, donor agencies, and
international development practitioners. Selected papers will be translated into French
Individuals and institutions in Third World countries may receive single copies free
of charge. See inside back cover for a list of available papers and their prices. For more
information, write to:
MSU International Development Papers
Department of Agricultural Economics
Michigan State University
East Lansing, Michigan 48824-1039
MAINTAINING THE MOMENTUM IN POST-GREEN REVOLUTION
AGRICULTURE: A MICRO-LEVEL PERSPECTIVE FROM ASIA
This paper reviews from a micro-level perspective, the opportunities for increasing
agricultural productivity in the large areas of Asia in a "post-green revolution stage of
development," where modern rice and wheat varieties and moderate to high doses of
fertilizer have already been widely adopted. Productivity increases in these areas from
the spread of new varieties and increasing fertilizer use have now slowed. Continued
rapid growth in productivity to exploit the genetic potential of modern varieties depends
on more efficient use of available technology. Farmers increasingly use a wide range of
"second generation inputs" such as secondary and micro-nutrients, pesticides, and
improved cultural practices. Complexity of crop management in these areas has
increased dramatically due to interactions between a wider array of technological
components and increased location specificity of the technology. The new technology is
also more management intensive since it requires more information and skills for
efficient use. In addition, increased multiple cropping, complex crop rotation effects and
a rapidly changing technical and economic environment add further complexity to farmer
In this situation, the potential for inefficiencies in resource use is much greater if
farmers lack ready access to a continual flow of up to date technical information or have
inadequate technical and managerial skills. Evidence is presented that technical
knowledge of farmers regarding new technological components is often very poor. In
addition, in post-green revolution agriculture, technical inefficiency the difference
between farmers' production levels and the potential given existing input and resource
use generally ranges from 20 to 50 percent. Technical knowledge, extension contacts
and education are shown to be the major factors explaining differences in technical
efficiency between farmers in a given area. The Schultzian characterization of farmers
as "poor but efficient" is no longer applicable to post-green revolution agriculture in
Deficiencies in technical knowledge and skills can be traced to the performance of
rural institutions, especially adaptive research, extension, and rural schooling. Adaptive
research generally lacks a farmer and problem orientation and is often the weakest and
most neglected part of the agricultural research system. The low quality of extension
advice and rural schooling in many areas compound this problem, especially for small
farmers. Recent innovations in adaptive research (the use of a farming systems
perspective) and in extension (the Training and Visit System) are reviewed. These
institutional innovations should help to improve the capacity of the research and
extension system to increase the flow of relevant and useful information to farmers.
However, in both of these approaches there is too much emphasis on prescriptive
information or recipes for crop production at the expense of providing farmers a better
understanding of new technology and improving technical and managerial skills.
Moreover, both innovations have been instituted from the "top down" and provide little
opportunity for farmers themselves to influence the direction and performance of
research and extension.
The evidence presented strongly supports the need for increased efforts to increase
the quantity and quality of information and skill acquisition by farmers in post-green
revolution agriculture. Critical issues for investment allocation between adaptive
research, extension, and rural schooling are discussed. It is concluded that there are
limited opportunities to substitute between these three components of the formal
information and skill system and that in general they are strong complements. In some
post-green revolution areas, particularly in South Asia, the low level of formal schooling
may be a major constraint on increasing the pay-offs to the investment in adaptive
research and extension that is needed to sustain rapid increases in productivity in the
future. The private sector can also play a greater role in adaptive research and
information dissemination, especially for chemical inputs, but the public sector must
provide leadership in the foreseeable future. Finally, applied research activities, such as
plant breeding, can to some extent substitute for weaknesses in adaptive research,
extension and education by developing technologies that require less information and
managerial skills in order to be efficiently used by farmers.
MAINTAINING THE MOMENTUM IN POST-GREEN REVOLUTION
AGRICULTURE: A MICRO-LEVEL PERSPECTIVE FROM ASIA
This paper is published by the Department of Agricultural Economics, Michigan
** The author is an Economist with CIMMYT, Mexico and formerly South Asian
Regional Economist with CIMMYT based in Islamabad, Pakistan. Views expressed here
are not necessarily those of CIMMYT.
@( All rights reserved by Michigan State University, 1987.
Michigan State University agrees to and does hereby grant to the United States
Government a royalty-free, nonexclusive and irrevocable license throughout the world to
use, duplicate, disclose, or dispose of this publication in any manner and for any purpose
and to permit others to do so.
Published by the Department of Agricultural Economics, Michigan State University, East
Lansing, Michigan 48824-1039 U.S.A.
MAINTAINING THE MOMENTUM IN POST-GREEN REVOLUTION
AGRICULTURE: A MICRO-LEVEL PERSPECTIVE FROM ASIA
TABLE OF CONTENTS
EXECUTIVE SUMMARY ........................................ i
LIST OF TABLES ................................................ vii
LIST OF FIGURES.............................................. ix
ACKNOWLEDGMENTS .......................................... xi
1. INTRODUCTION ............................................... 1
II. INCREASING COMPLEXITY OF CROP MANAGEMENT IN
POST-GREEN REVOLUTION AGRICULTURE ...................... 4
III. INFORMATION, EDUCATION, AND ECONOMIC EFFICIENCY ........ 11
A. Farmers' Technical Knowledge and Input Use .................. 11
B. Technical and Allocative Efficiency in Asian Agriculture ........ 13
1. Measuring Economic Inefficiency ....................... 14
2. Empirical Estimates ................................... 16
IV. THE FORMAL INFORMATION SYSTEM INSTITUTIONAL
PERFORMANCE AND INDUCED INSTITUTIONAL INNOVATIONS ..... 24
A. Adaptive Research ........................................ 24
B. Agricultural Extension ..................................... 26
C. Recent Institutional Innovations in
Research and Extension Systems ............................ 28
1. The Farming Systems Perspective in
Adaptive Research .................................... 28
2. The Training and Visit Extension System .................. 30
D. Rural Schooling ............................................ 32
V. FARM INFORMATION AND SKILLS: IMPLICATIONS FOR
DEVELOPMENT STRATEGY ...................................... 35
A. Integrating Research, Extension,
and Rural Schooling ......................................... 35
B. The Private Sector and Information Generation
and Dissemination ..... . .............. ......... 40
C. Substituting for Farmer Information and Skills ................. 42
D. Information, Skills, and Equity .................. ........... 44
VI. CONCLUDING COMMENTS ...................................... 47
BIBLIOGRAPHY .............................................. 49
LIST OF TABLES
1. Percent of Area Planted to Modern Varieties of Wheat
and Rice in Selected Asian Countries, Early 1980s........................ 6
2. Coefficient of Variation of Traditional and Modern Inputs for
Farmers 5 to 10 and 15 to 20 Years After Introduction of Modern
Varieties ...................... .............. ............... ... 13
3. Classification of Economic Inefficiencies and
Their Policy Relevance .............................................. 15
4. Estimates of Technical Efficiency in Modernizing
Agriculture in Asia .................................................. 17
5. Summary of Two Integrated Surveys Experimental Studies of
Farm Level Productivity of Wheat and Rice.............................. 21
6. Growth Rates of Real Expenditures on Research and Extension,
South and Southeast Asia, 1959-80. ..................................... 28
7. Percent Adult Population (25+ years) Who Have Attended School
in Selected Asian Countries............................................ 33
8. Institutions and Functions Servicing the Information and Skill
Needs of Farmers ........................ ........................ 37
9. Relationship Between Extension Contacts, Education, and Farm Size,
Punjab, Pakistan, 1986 .............................................. 46
LIST OF FIGURES
Figure 1. Changes in Area under Modern Varieties and Fertilizer
Applied to Wheat in Pakistan's Punjab and Rice
in Central Luzon, Philippines ................................. 5
Figure 2. Components of Farmers' Information and Skill System ............. 36
I appreciate valuable comments on an earlier draft of this paper provided by Vernon
Ruttan, Ruben Echeverria, Bert Sundquist, Willis Peterson, Robert Tripp, Paul Heisey,
Peter Hobbs, John Flinn, Mubarik Ali, Rick Bernsten, Eric Crawford, Jock Anderson and
Carl Eicher. I would also like to thank Luis Macagno for untiring bibliographic
MAINTAINING THE MOMENTUM IN POST-GREEN REVOLUTION
AGRICULTURE: A MICRO-LEVEL PERSPECTIVE FROM ASIA
For large areas of Asia, the rapid growth of agricultural productivity associated
with the introduction of semi-dwarf wheat and rice varieties of the so-called green
revolution has entered a new phase. Most farmers in these areas now use modern
varieties and moderate to high doses of fertilizer, and productivity increases from these
two sources have slowed. Improvements in other crop management practices and in
cropping intensity play an increasing role in maintaining productivity growth and
exploiting the genetic potential of existing varieties. While much of the literature on
agricultural development continues to focus on the green revolution and its impact, this
new post-green revolution phase is already well established in many areas but has been
largely ignored by analyses of Third World agricultural development. Rather, attention
seems to have turned to more macro-level concerns, especially pricing policy and food
While agricultural development thought is notably "faddish," one of the enduring
themes over the past two decades has been Schultz's (1964) seminal contribution on the
efficiency of small farmers in traditional agriculture in allocating their resources and
responding to price incentives the so-called "poor-but-efficient" hypothesis. This laid
the theoretical justification for the high pay-off input, or science-based model of
agricultural development exemplified by the green revolution. A further implication of
the poor-but-efficient hypothesis was that there were low pay-offs in traditional
agriculture to extension and farm management efforts to encourage farmers to use their
existing technology more efficiently (Staatz and Eicher, 1984). Hence, most increases in
productivity would have to come about through introduction of new high pay-off inputs
into traditional agricultural systems.
In the two decades since Schultz's book was published, the rapid uptake of improved
varieties and fertilizer technology has transformed the agricultural landscape in many
developing regions, especially in Asia. The image of a traditional subsistence agriculture
no longer holds as small farmers use an increasing number of purchased inputs and strive
to keep pace with a dynamic technical and economic environment. The green revolution
has highlighted the role of agricultural research and especially plant breeding research in
stimulating technological change in agriculture. But the very success of the green
revolution with its emphasis on technical change may have drawn attention away from
the importance of the human element in agricultural development (Jones, 1978). In
particular, Schultz's poor-but-efficient hypothesis still widely prevails today although it
is by no means universally accepted.1 A central thesis of this paper is that general
acceptance of this hypothesis has led to a relative neglect of efforts by research,
extension and farm management to increase economic efficiency in farmers' resource
use, and insufficient emphasis on upgrading farmers' technical skills and managerial
ability.2 The very assumptions underlying the hypothesis that farmers in traditional
agriculture have evolved over a long period an efficient system through accumulation of
experiences and an intimate knowledge of their environment have become outmoded by
the rapid process of change introduced by the green revolution. Indeed, Schultz himself
has persuasively argued (largely in the context of the U.S.) that, in a dynamic
agriculture, farmers are continually in a state of disequilibrium and that there are high
returns to better information and skills to improve farmers' economic efficiency
This paper argues that two of the major sources of agricultural growth in the past
two decades in Asia the spread of modern varieties and rapid increases in fertilizer
use have already been substantially exploited. A new and more complex second
generation of inputs and management practices plays an increasing role in productivity
growth, and investments in better information and skills of farmers to improve economic
efficiency in using this wider array of inputs are needed to maintain the momentum in
post-green revolution agriculture. Furthermore, in many of these countries, increasing
food self-sufficiency, reductions in subsidies on agricultural inputs, and declining world
prices for food grains in the 1980s have led to less favorable price incentives and more
'See, for example, Shapiro (1983), Nair (1979) and for a particularly stinging
attack, Adams (1986). The publication of Schultz's book also generated considerable
controversy at the time (e.g. Lipton, 1968).
2Economic efficiency in this paper refers to both technical efficiency the
productivity of farmers' existing input mix and allocative efficiency the combination
of inputs that leads to profit maximization.
pressure to increase economic efficiency.3 However, the development of the capacity of
rural institutions to meet the needs of this changed environment has lagged the process
of technical change (Ruttan, 1978; Bonnen, 1986), although recent innovations in research
and extension are beginning to close the gap.
The paper is organized as follows. First, the increasing complexity of crop
management issues facing small farmers in post-green revolution Asia is described.
Second, evidence of economic inefficiencies in resource use in these regions and the
importance of farmers' information and skills in reducing these inefficiencies is
presented. This leads to a discussion of institutional changes in research, extension and
rural education aimed at improving information and skills of farmers and their
implications for development strategy in the post-green revolution era. The paper
focuses on those post-green revolution areas where almost all farmers use modern
varieties of rice or wheat and moderate to high levels of fertilizer of 75 kg/ha of
nutrients or more.4 A conscious effort is made throughout the paper to analyze these
issues from the vantage point of an accumulating body of farm-level research and
3Price discrimination against food grain production is probably not as large or as
widespread as commonly suggested (see Byerlee and Sain, 1986) and with declining world
prices and increased self-sufficiency, many countries have producer prices above the
world price equivalent. Herdt (1987) also documents declining economic incentives to
Philippine rice farmers in recent years.
4The issues discussed here are not exclusive to post-green revolution irrigated areas
of Asia. In many rainfed areas, improved varieties and moderate doses of fertilizer have
also been widely adopted, especially in maize in Latin America and Eastern and Southern
II. INCREASING COMPLEXITY OF CROP MANAGEMENT
IN POST-GREEN REVOLUTION AGRICULTURE
The green revolution in Asia involved widespread and rapid adoption of semi-dwarf
wheat and rice varieties, especially in the decade 1967-77, that, in turn, stimulated
adoption of two other key inputs nitrogenous fertilizer and improved supplies of
irrigation water. By the mid to late 1970s, modern varieties had been fully adopted in
many environments, although there were important exceptions for large rice growing
tracts of eastern India, Bangladesh and Thailand (See Figure I and Table 1).5 Genetic
gains in yield potential in successive generations of modern varieties have slowed and an
increasing proportion of plant breeding research in wheat and rice is now devoted to
"maintenance research" to protect yield gains against breakdown of pest resistance
(Plucknett and Smith, 1986) and to adapt semi-dwarf varieties to less favorable
A similar situation also holds for the two other major inputs nitrogenous fertilizer
and water. Fertilizer levels increased rapidly for several years after adoption of modern
wheat and rice varieties (see Figure 1) and explained much of the agricultural growth in
the 1970s in Asia (Scandizzo, 1984). Fertilizer levels have now reached fairly high levels
in many areas, are increasing much more slowly and provide lower gains at the margin
than in earlier years. For example, fertilizer use on wheat averages about 120 kg/ha of
nutrients in well irrigated areas of the Punjab, Pakistan and 160 kg/ha in the Indian
51n the past decade, modern wheat varieties have also been extensively adopted in
less favorable environments (Dalrymple, 1986a). For example, in rainfed areas of
Pakistan's Punjab, the proportion of area sown to modern varieties increased from 20
percent in 1975 to over 60 percent in 1985.
6Semi-dwarf wheat varieties released in the 1960's generally yielded 30-50 percent
more than earlier taller varieties under irrigated conditions with moderate doses of
fertilizer (Nagy, 1984). Releases since then have increased yield potential by an average
of one percent per year or a total of about 20 percent. Most of this increase was due to
the crossing of spring by winter wheats. Similarly, Dalrymple (1986b) notes that no new
rice variety has out-yielded the potential in favorable environments of the variety IR8
released nearly 20 years ago, although major progress has been made in incorporating
pest resistance, stress tolerance, quality characteristics and earliness (probably at the
expenses of gains in yield potential). Recent advances in biotechnology are not expected
to change yield potential for cereals before the turn of the century.
Rice, Central Luzon Philippines (Source: Herdt, 1986)
Wheat, Punjab, Pakistan
Figure 1. Changes in area under modern varieties and fertilizer applied to wheat
in Pakistan's Punjab and rice in Central Luzon, Philippines.
Table 1. Percent of Area Planted to Modern Varieties of Wheat and Rice in Selected
Asian Countries, Early 1980s.
Bangladesh 25 96
Indonesia 82 b
India 54a 76
Nepal 36 92
Pakistan 46a 86
Philippines 85 b
Sri Lanka 87 b
Thailand 13 b
a Some important rice producing regions of India and Pakistan (e.g. Punjab and Tamil
Nadu in India and Sind in Pakistan) had adoption rates over 80%.
b Wheat is not a commercial crop.
Source: Dalrymple, 1986a, 1986b.
Punjab. Similar levels are recorded for rice in the Philippines (e.g. Lingard et al.,
1983). Likewise, introduction of semi-dwarf varieties was accompanied by rapid
expansion of irrigation facilities and improved water control. However, investment in
irrigation is also slowing in many post-green revolution areas as only the more difficult,
and hence more expensive investments remain.
Hence, the major sources of rapid growth during the green revolution era have to a
large extent been utilized, especially in better endowed irrigated areas (Vyas, 1983; Huke
et al., 1982), although steady gains will continue to be achieved through release of newer
varieties and rising fertilizer doses. Yet most observers agree that there are substantial
opportunities to increase productivity in these areas through increased yields, reduced
costs, and improved cropping systems. Most of these opportunities depend on non-
genetic gains in productivity through use of new inputs and more efficient use of existing
inputs to exploit the genetic potential of existing varieties.
Farmers now face a wide array of "second generation inputs" that offer the
opportunity to substantially increase productivity but at the same time greatly increase
the complexity of crop management. For example, for irrigated wheat in Pakistan,
farmers now commonly purchase five inputs tubewell water, nitrogenous fertilizer,
phosphatic fertilizer, tractor power and thresher services none of which were in wide
use two decades ago. In addition, an increasing number of farmers, especially in the
Indian Punjab use potash fertilizer and micro-nutrients (e.g. zinc), soil amendments, seed
treatment for disease control, improved on-farm water management methods including
precision land leveling, and more precise planting methods and spacing. Higher yields
and increased cropping intensity also lead to increased crop losses from pests (in absolute
terms) and pesticide use has become widespread, especially in rice.7 At the same time,
improved water supplies, earlier maturing wheat and rice varieties, and in some cases,
selective mechanization, have greatly expanded the opportunities for multiple cropping
which require management of complex double and triple cropping patterns that
sometimes include new and unfamiliar crops.
In these evolving production systems, crop management is generally complex.
Changes in practices required to sustain increases in productivity, while still quite
profitable, do not provide the spectacular economic returns characteristic of the first
round of inputs adopted during the green revolution. Hence, their successful adoption is
more sensitive to the efficiency with which they are used by farmers. The wider array of
technological options available and the interactions between them, requires farmers to
identify a logical stepwise sequence for adoption that fits their agro-climatic and socio-
economic circumstances. Interactions between technological components may also
require adjustment in traditional inputs. For example, irrigated wheat yields in much of
South Asia appear to be limited by low plant populations. Farmers use seed rates and
planting methods that were appropriate for low yielding conditions but which now no
longer appear to be adequate.
In multiple cropping sequences with two or three crops per year, management
complexity is also increased by the need to sometimes make simultaneous decisions on
management of each crop in the sequence. For example, in the cotton-wheat rotation of
Pakistan, the potentially positive effect of introducing earlier cotton varieties on
planting dates for wheat following cotton was largely cancelled by farmers' rapid
increase in pesticide use for cotton which increased cotton yields and delayed the cotton
harvest (Byerlee, Akhter and Hobbs, 1987). With earlier maturing varieties grown in
7Crop losses due to pests are often proportional to yield levels (Zadoks, 1985) and
pest control measures which were uneconomic at low yield levels become economic at
higher yield levels.
multiple cropping systems timeliness of operations is often a critical determinant of
system productivity and places additional burdens on management. In addition, crop
rotation in more intensive cropping systems often plays a key role in managing pest
In addition, neither the technical nor economic environment in which these
management decisions are made remains static. Increased yields and cropping intensity
tend to deplete soil potasium and micro-nutrient reserves and lead to responses to
application of these inputs. Application of phosphatic fertilizers has substantial
carryover effects that eventually allow the application of lower maintenance doses.
Control of broadleaf weeds in wheat reduces the competition for grassy weeds whose
increasing population demands control measures. Increased supplies of tubewell water
and higher cropping intensities complicate water and salinity control management in
irrigated areas of South Asia. Evolving insect and disease biotypes require the rapid
adoption of new varieties. Indeed, there is evidence that, unless farmers keep abreast of
these changes in the technical environment, productivity may decline over time due to
depletion of soil nutrients or pest build-up in intensive cropping systems.9 Along with
these changes in the technical environment, the economic environment has also been
subject to sharp adjustments in price relationships in some countries in the 1980s due to
so-called "policy reforms", especially the reduction in input subsidies. (For evidence on
the Philippines, see Herdt, 1987).
The second generation inputs are also more "management intensive" that is, they
often require more information and skills for successful adoption than the earlier
introduction of new varieties and nitrogenous fertilizer (Kahlon, 1984). While the semi-
dwarf varieties, especially for wheat, are well known for their adaptability over a wide
range of environments, the use of more complex fertilizers and pesticides often interacts
strongly with variation in soil type and with year-to-year variation in climate and pest
incidence. Hence, individual farmers need to adapt the technology to their own
requirements and frequently there are substantial returns to managing inputs for
individual fields and seasonal conditions within a farm.10
8Even the switch from bullocks to tractor power appears to have led to increased
soil compaction in some areas that requires modifications of tillage methods.
9See Winkelmann (1987) for a discussion of sustainability in intensive production
systems of the tropics.
10The highest returns to input management by fields and seasons are obtained from
deciding whether to use or not use a particular input in a specific circumstance and how
Technical skills required to efficiently use the new inputs are also much greater
than for a simple varietal change. An extreme example is use of integrated pest
management (IPM) in rice in Asia. While the practice has potentially high returns to
farmers and to society, it requires skills in identification of pest insects and beneficial
insects, quantification of pest populations and damages, knowledge of threshold levels of
pest damage, and skills in selecting the appropriate chemical, calculating the dosage and
applying the product (Goodell, 1984b). Management complexity is further increased by
the wide variety of inputs available. For example, one survey of 150 rice farmers in the
Philippines recorded 38 different insecticides being used by farmers, many of them
similar products under different names (Litsinger et al., 1980). Even the number of new
wheat varieties available to farmers has tended to increase in order to provide different
maturities to fit multiple cropping systems, and to reduce the risk of disease epidemics.
Hence, effective crop management in this post-green revolution era places heavy
demands on the information and skills of farmers. In traditional agriculture, information
is primarily generated internally by farmers through a process of learning-by-doing and
informal experimentation (Johnson, 1972; Biggs and Clay, 1981). Although management
in traditional agriculture is also often quite complex, gradual changes in the resource
base and in the external environment, especially through population growth, allowed an
evolution of farmer management to incorporate these changes based on farmers'
experiences and knowledge of their environment accumulated over generations of
farming. In the new science-based agriculture, however, the value of traditional
knowledge rapidly depreciates as new inputs and cropping patterns are introduced. Even
outside sources of information will have a built-in obsolescence (Welch, 1978).
If there is a one-shot disturbance to the equilibrium due, for example, to a new
input becoming available or a change in price ratios, farmers will eventually adjust to a
new equilibrium by their own trial and error. Yet the green revolution was not a one-
shot disturbance (as implied by Welch, 1978) but a series of increasingly complex changes
as new inputs and cropping patterns were introduced. Whereas the spread of semi-dwarf
varieties was very rapid as new seeds and information passed from farmer to farmer,
information flows from farmer to farmer for the second generation inputs are expected
to be much slower and less effective, since a much greater range and complexity of
information and skills are needed.
Management skills developed in traditional agriculture through experience in
that input is used (e.g. timing and method of application). The level of an input may
often be adjusted in a fairly wide range around its "true" optimal level without having a
large effect on productivity (Anderson, 1975).
learning-by-doing are also no longer adequate to keep pace with these changes. The
movement toward "optimal" levels and technically efficient use of a new input is a
process of continual improvement and adjustment as farmers' experience and information
grows. However, the optimum level of an input is itself continuously changing in a
technically and economically dynamic environment so that farmers can be expected to
remain in a constant state of disequilibrium as they strive to hit a "moving target."
It is now widely recognized that formalized schooling helps to develop technical
and managerial skills for a science-based agriculture.1 Schultz (1975), Welch (1978) and
others have convincingly argued that education increases the ability to acquire and
evaluate information. Or, in economic terms, education reduces the cost of obtaining a
given amount of information (e.g., through literacy skills) and increases the benefits of
this information in decision making (Ram, 1981). Education may increase farmers'
technical skills (e.g., computing dosages) as well as improve farmers' allocative ability in
adapting new technology to their own needs and adjusting to changes in the
environment. Understanding of basic scientific principles may be necessary to adopt
innovations which are sometimes counter-intuitive, or whose benefits are not
immediately obvious (e.g., use of a new wheat variety to guard against breakdown in rust
In many post-green revolution areas, decision making complexity in small-holder
agriculture is now closer to the situation of farmers in industrialized countries (the levels
of bio-chemical technologies are similar) than the image of traditional agriculture
commonly held for farmers in developing countries. Just as farmers in industrialized
countries have moved from a science-based to an information-based agriculture (Sonka,
1985), small farmers in the post-green revolution era are also entering this "information
age." However, unlike farmers in industrialized countries who have had a relatively long
period to adjust to a science-based agriculture, the increased demands on knowledge,
technical skills, and managerial capacity of post-green revolution farmers, most of whom
used practically no purchased inputs two decades ago, has been collapsed into a very
short period of less than two decades.
SI Evidence on the importance of education in Asian agriculture will be discussed in
III. INFORMATION, EDUCATION AND ECONOMIC EFFICIENCY
The green revolution in the 1960s and 1970s spawned a large number of studies of
the adoption of the new varieties and fertilizer and their equity implications 12 Until
recently, however, there were few studies on the efficiency with which the new
technology was used once adopted.
In this chapter, empirical studies of farmers' technical information and how it
affects economic efficiency are reviewed. Two types of micro-level studies are
reviewed; a) those that describe and analyze farmers' technical knowledge and its effect
on input use, and b) production function studies that estimate economic inefficiencies
and try to identify factors responsible for these inefficiencies.
A. Farmers' Technical Knowledge and Input Use
The positive effects of farmers' technical knowledge, education and extension
contacts on the adoption of a new input are well known (see reviews by Feder, Just and
Zilbermann, 1985; Herdt and Capule, 1983; Rogers, 1983). However, few studies have
examined the evolution of farmers' technical knowledge in the post-adoption stage that
characterizes post-green revolution Asia. In the mid-1970s, Bernsten (1977) surveyed
farmer's technical knowledge of 50 management practices judged by rice scientists to be
"critical for the farmer to achieve maximum input efficiency" (Bernsten, p. 191). These
included age for transplanting modern varieties, appropriate depth of standing water for
herbicide use and appropriate insecticides for given insect pests. Out of a maximum
score of 12, farmers averaged 5.7, suggesting substantial scope to increase farmers'
Recent surveys in Pakistan (Heisey et al., 1987) show that most farmers, even after
nearly two decades of experience with modern wheat varieties and fertilizers, were not
able to compute nutrient doses, especially for phosphatic and compound fertilizers. They
had inadequate knowledge of newer varieties and their characteristics, and most were
unaware of the potential breakdown in rust resistance of wheat varieties. In northwest
India, where more effective extension has been in place, farmers' knowledge appears to
be much better, but away from these areas, farmers' information scores for modern
12 For reviews, see Lipton and Longhurst, 1985; Herdt and Capule, 1983; and Ruttan
and Binswanger, 1978).
inputs are generally poor (Feder, Slade and Sundaram, 1986; Srivastava, 1976).
Deficiencies in farmers' technical knowledge generally increases with increasing
complexity of the practice (Mayani and Kumar, 1980). In particular, information on plant
protection is generally poor. Even the relatively literate farmers in the highly
commercialized agriculture of northwest Mexico have poor knowledge of plant protection
measures that they often employ, especially the appropriate product and time of
application of herbicides for grassy weeds in wheat, a major problem in the area.13
Within an area, too, there are often quite large differences between farmers in
technical knowledge (Heisey, et al., 1987; Feder and Slade, 1984). Feder and Slade(1984)
were able to relate these differences to information supply (extension contact), the cost
of information acquisition per unit area (farm size) and human capital variables
(numeracy and education).
Farm level surveys also indicate the substantial variation in use of modern inputs
even where farmers have had many years of experience with a given input. Table 2
compares the Coefficient of Variation (CV) in per hectare use of fertilizer a modern
input to the CV of traditional input usage for a number of surveys in relatively
homogeneous areas in which most farmers had adopted fertilizers. The CV is always
higher for the modern input than for the traditional input. Even after nearly 20 years of
experience of using fertilizer, farmers in Pakistan's irrigated areas still exhibit wide
variation in fertilizer doses in a given area. Taking cross-sectional data the CV has
fallen from over 100% in an area where fertilizer was only recently introduced, to 50-
60% in the irrigated Punjab. By contrast the CVs for fertilizer use on wheat in three
counties of Michigan were 30-40% in 1959-61, less than 10 years after widespread
adoption. (Calculated from Hoffnar and Johnson, 1966). Much of this variation can be
explained by differences in access to capital and irrigation water, soil variation, and crop
rotation (Byerlee et al., 1986) (which are likely to interact more strongly with modern
inputs than traditional inputs), but in part, it is due to differences in information, skills
and experiences of farmers (Heisey et al., 1987).
13Grassy weeds in wheat such as wild oats are taxonomically related to wheat and in
the early growth stages are difficult to distinguish from wheat plants. Many farmers
have difficulty comprehending that herbicides can selectively kill these weeds in a wheat
Table 2: Coefficient of Variation of Traditional and Modern Inputs for Farmers 5 to
10 and 15 to 20 Years After Introduction of Modern Varieties.
Kenya Pakistan India Pakistan Pakistan
(Vihiga) (Gilgit) (Palanpur) (Gujranwala) (Multan)
Number of years since
initial adoption of
variety and fertilizer 5-10 5-10 5-10 15-20 15-20
Coefficient of Variation (%)
- seed/ha 29 11 14 12
- number of ploughings 28 33 43 34
- organic manure/ha 70
- labor/ha 35 46
- nitrogen/ha 98 100 61b 52 53
- phosphorus/ha 85 131 60 66
aAll inputs were used by over 80 percent of farmers in the year of the survey, except
phosphorus in Gilgit (47 percent).
bTotal fertilizer applied.
Source: Vihiga Moock (1981): Gilgit Hussain (1986); Palanpur Bliss and Stern(1982);
Gujranwala and Multan original survey data.
B. Technical and Allocative Efficiency in Asian Agriculture
There is a growing body of literature that attempts to measure economic
inefficiencies within a production function framework and relate these inefficiencies to
socio-economic characteristics of farmers. These studies are not without conceptual and
methodological difficulties and some discussion of these problems is necessary before
presenting empirical results.
1. Measuring Economic Inefficiency
Economists widely distinguish between technical inefficiency and allocative or
price inefficiency, following pioneering work of Farrell (1957). Technical inefficiency
refers to failure to operate on the production frontier and is generally assumed to reflect
inefficiencies due to the timing and method of application of production inputs.
Allocative inefficiencies refer to the failure of farmers to meet the marginal conditions
for profit maximization that is, to equate the Marginal Value of Products (MVPs) of
inputs to their market prices.14 It is often useful for policy purposes to further divide
allocative errors between a) the constrained case where allocative gains are measured by
reallocating inputs within the existing expenditure level i.e. movement along the
isoquant to the expansion path, and b) the unconstrained case where allocative gains also
accrue due to movement along the expansion path until the marginal return on
expenditures is equal to the cost of capital.15 Allocative errors in the constrained case,
like technical inefficiencies, are most likely to reflect deficiencies in information and
skills and be easier to correct in the short run. Allocative errors in the unconstrained
case (i.e. scale errors) may also reflect inadequate information and skills, but they are
also likely to arise from the effects of market imperfections and risk aversion and non-
monetary goals of farmers that imply different and longer run interventions than in the
case of imperfect information. These various inefficiencies and policy interventions are
summarized in Table 3.
It is important to note at the outset that economic efficiency is only a standard for
society to judge resource productivity against its potential and is not intended to suggest
irrational decisions on the part of farmers because they do not maximize profits. The
failure of farmers to use the most efficient techniques of production because of inade-
quate information suggests that the cost to the individual farmer of acquiring better
information is greater than the benefits.16 The question arises as to whether policy
interventions can improve the "market" for information and reduce its costs to farmers.
141n multi-product firms, allocative efficiency also implies that the Marginal Rate
of Transformation between products be equal to the ratio of product prices.
15Empirical studies have used both specifications. The unconstrained case with all
factors variable implies diminishing returns to scale. An additional restriction in the
constrained case is to constrain expenditures for only variable or cash inputs.
16Other possible reasons for inefficiencies include fixed assets and vintage effects,
property rights (e.g. tenancy) as well as farmers' non-monetary values (see, for example,
Table 3. Classification of Economic Inefficiencies and Their Policy Relevance
Likely Cause of
Type of Inefficiency
1. Technical Inefficiency
Failure to operate on the
production frontier due to
errors in the timing or method
of application of inputs
2. Constrained Allocative Errors
Errors in allocating input within
existing expenditure levels -
movement to expansion path.
3. Scale Errors
Increased levels of input use
through higher expenditure
levels movement along the
2. Market failure in
3. Differential risk
effects of inputs
1. Capital constraint
2. Risk aversion
3. Inadequate information
The distinction between technical and allocative efficiency also depends on the
level of aggregation and specification of the production function. Stigler (1976), for
example, argues that if the production process is completely specified (including timing
and method of using inputs) there would be no technical inefficiency, only allocative
inefficiency. This point is particularly important given that most production functions
are specified with very aggregate categories of inputs such as land, labour, fixed capital
and variable cash inputs. If cash inputs are an aggregate of say, herbicides and nitrogen
and phosphatic fertilizers, measured "technical" efficiency will include both allocative
errors among these three inputs as well inefficiencies due to timing and method of
application of these inputs.
2. Empirical Estimates
Technical efficiency has typically been measured by estimating a frontier
production function.17 The frontier production function attributes variation from the
most efficient farm to technical inefficiency. In fact, if specified in aggregate terms of
land, labor and capital, it also captures micro-level variation in soil and land type, crop
rotation, etc., as well as sampling and measurement errors, and hence tends to
overestimate technical inefficiency. More recently, stochastic formulations of the
frontier production function have been applied to sort out the effects due to random
errors from those due to technical inefficiency (e.g. Huang and Bagi, 1984; and Kalirajan
and Flinn, 1983). However, methodological debate still arises in interpretation of the
results (Taylor and Shonikwiler, 1986; Pasour, 1981; Russell and Young, 1983).
Table 4 summarizes empirical measures of technical inefficiency for farmers in
post-green revolution Asia and within the caveats of the previous discussion, suggest that
on average, farmers could increase output by 20-50%, given existing resource use.
Comparable estimates of technical inefficiency in traditional agriculture are scarce but
suggest that the average level of inefficiency is less than 20 percent (e.g., Belbase and
Grabowski, 1985 in the Nepal hills and Mijindadi and Norman, 1984 in northern Nigeria.)
Several of these studies from post-green revolution Asia have tried to explain the
individual farmer-specific technical inefficiency in terms of farmer characteristics. In
some cases, these are related to external constraints, such as access to credit (Lingard,
et al., 1983) or irrigation water (Flinn and Ali, 1986) (see Table 4). However, a
particularly important finding is that in all studies where farmer-specific technical
efficiency was analyzed, the major factors explaining differences in efficiency were
variables measuring farmers' information and skills such as education, age, experience,
contacts with extension agents, and technical knowledge (Table 4). 18 Hence, even if the
absolute level of technical inefficiency is overestimated due to inadequate specification
of the production function (e.g., failure to include micro-soil variation), there is
17For an excellent overview of methodological issues in frontier production functions,
see Forsund, Lovell and Schmidt (1980).
18The lack of significance of some of these information and skill variables in some cases
may be due to their relative uniformity in the sample population. For example, the study
of Kalirajan (1981) found no effect of education on technical efficiency, probably
because the sample included only persons with primary school education and above
(Kalirajan and Shand, 1985).
Table 4. Estimates of Technical Efficiency in Modernizing Agriculture In Asia
1. Huang & Bagi
3. Kalirajan & Flinn
4. Lingard, et. al.
6. Peng and Chen
7. Flinn and Ali
Tamil Nadu, India
1969/70 Irrigated Wheat
1980/81 Rainfed Rice
1970-79 Irrigated Rice
1979/80 Irrigated Rice
89 Not analyzed
50 Planting method*
50 Soil type*
Timing of Inputs*
* Significant at 5% level or less
na = not available
convincing evidence that information and skills play an important role in the relative
degree of inefficiency among farmers.
A number of methods have also been used to measure allocative inefficiencies.
Many studies have compared the overall sample Marginal Value of Product of each input
to the average price of the input and tended to be highly conservative in rejecting the
null-hypothesis that farmers were efficient in allocating their resources. Even though
most authors concluded that farmers were acting in a profit maximizing manner, the
estimated marginal productivity and price relationships often suggested quite substantial
allocative inefficiencies. For example, Barnum and Squire (1978) for irrigated wheat in
India conclude that farmers were allocatively efficient even though the ratio, K =
(Marginal Value Product/Marginal Factor Cost), for "other variable inputs" (presumably
fertilizer) was 2.7. Other studies also typically find a high value for K for modern inputs
in post-green revolution settings (Bliss and Stern, 1982 (wheat); Khan and Young, 1979
(all crops); Armenia, 1983 (rice)).19 These studies also measured average allocative
errors for the total sample. Even if, on average, farmers are efficiently allocating
resources, individual farmers may exhibit substantial variation from the optimum.
The use of the profit function approach has been proposed as a means of measuring
both technical and allocative efficiency (Yotopolous and Lau, 1979), although its
application requires cross-sectional variability in prices of variable inputs. Application
of the profit function in modernizing agriculture gives somewhat conflicting results.
Junankar (1980a,1980b) rejects the hypothesis that Indian farmers in two separate
samples (wheat and rice areas, respectively) are economically efficient while Yotopolous
and Lau (1979) and Jamison and Lau (1982) accept the hypothesis for other Asian
In a combination of the above approaches, Ali and Flinn (1986) estimated a frontier
profit function for rice in Pakistan using a stochastic specification. Average economic
inefficiency was estimated at 28% with over half of farmers showing at least a 25% loss
in efficiency. Education was the dominant factor explaining differences in efficiency in
this sample of farmers.
Another approach is to include farmers' knowledge and education as variables in the
production function to measure relative economic efficiency. Jamison and Lau (1982)
review 36 such studies of the effect of education on agricultural productivity and find
that education has a statistically significant or important impact on technical efficiency
19For reviews of earlier studies, see Dillon and Anderson, 1971 and Shapiro, 1979.
in all but four of the studies. More importantly, the average increase in productivity due
to completion of basic education (4-6 years) was 9.5% in a modernizing environment
versus 1.3% in a traditional environment, thus supporting Schultz's (1975) hypothesis.
The studies from Asia showed a particularly strong and consistent effect of education
(see Phillip's (1987) comments on Jamison and Lau) and more recent evidence from Asia
further confirm these findings (Antle, 1984; Pudasaini, 1983; Butt, 1984; Jamison and
Moock, 1984). There is thus strong evidence of the importance of education in farmer
efficiency in post-green revolution settings.
There is little evidence from these studies on how schooling affects agricultural
productivity. Recently, Jamison and Moock (1984) have attempted to establish the
intermediate outputs of formal education in Nepal that have a bearing on efficiency.
These can be classified into the development of basic competencies (e.g., literacy,
numeracy and cognitive skills) and the transmission of technical information. They found
that numeracy had a large and significant effect on efficiency in wheat production,
although their results were inconclusive for other educational outputs and other crops.
Fuller (1983) concluded that literacy (in this case from adult education) increased
economic efficiency of Bangladesh rice farmers. A better understanding of the pay-offs
to the different products of schooling has an important bearing on policy interventions to
improve efficiency since some of these products can provided by alternatives to formal
schooling (e.g., extension or mass literacy programs).
Schultz (1975) and Welch (1978) have hypothesized that the main benefit of
education in a dynamic agriculture is to increase the allocative ability of farmers. The
evidence from Asia in this respect is less conclusive. Pudasaini (1983) in Nepal concluded
that the allocative effect of education was more important than the effect on technical
efficiency but Jamison and Lau (1982) in Thailand and Wu (1977) in Taiwan found no
effect of education on allocative ability.20 As yet there are no studies from developing
countries relating education to efficiency in specific management decisions (e.g. the
study of pesticide use and integrated pest management in the U.S. by Pingali and
Carlson, 1985) or in adjusting specific inputs to a rapidly changing environment (e.g.
Huffman, 1977 again for the U.S.).
Two studies (Bernsten, 1977 and Bhati, 1973) have included a measure of farmers'
technical knowledge in the production function and in both cases (for rice in the
Philippines and Malaysia, respectively) the effect on productivity was highly significant
20These results are subject to the problem of input aggregation discussed above which
leads to over-estimation of technical efficiency relative to allocative efficiency.
and strongly positive.21 Bernsten (1977) further showed that farmers' technical
knowledge was related to socio-economic characteristics of the farmer, such as age,
experience and extension contacts, although educational level did not emerge as a
The production function approach reviewed here is a blunt instrument for analyzing
complex farming systems and crop management issues, characterized by substantial
heterogeneity in resources within farms and variability in crop response over seasons.
These studies generally employ quite aggregate specifications and show little
appreciation for the complexity of technical relationships in agriculture.22 These
problems are most pronounced for farm-level as opposed to crop-specific production
functions. As Upton concludes, "the farm is a highly complex and dynamic system and
any attempt to represent such a system by a single equation is unlikely to be
operationally meaningful" (Upton, 1979). In addition, the successful application of
production function analysis with cross-sectional data depends on the existence in the
sample of substantial variability between farmers in technical and/or allocative
efficiency (Doll, 1974).
In recent years, a number of multi-disciplinary farm-level studies have been
initiated that combine the insights of agronomists and economists and integrate on-farm
survey and experimental data. These studies not only estimate the productivity "gap" but
also help identify the specific sources of inefficiencies.
Table 5 summarizes results of integrated survey-experimental research at the farm
level for irrigated wheat in Pakistan (Byerlee, et al., 1986) and rice in the Philippines
(Herdt and Mandac, 1981). A complex of factors is shown to explain variation in yields
between farmers as well as the difference between farmers' yields and what is
potentially attainable and profitable given available technology. These factors include
exogenous variables related to soil type, availability of irrigation water, agronomic
variables such as pest incidence and plant density, and production practices such as crop
21A similar approach by Jamison and Moock (1984) showed no effect of farmers'
technical knowledge on productivity. Both Bhati (1973) and Jamison and Moock (1984)
measured technical knowledge in terms of farmers' knowledge of research
recommendations. This assumes that research recommendations are in fact relevant to
farmers (see Chapter 4).
22The major exceptions to this generalization is the series of studies on rice sponsored
by the International Rice Research Institute (e.g. Kalirajan and Flinn, 1983; Lingard et
al., 1983; Flinn and Ali, 1986).
Summary of Two Integrated Surveys Experimental Studies of Farm Level
Productivity of Wheat and Rice.
Number of farmers surveyed
Number of on-farm experiments
Factors influencing yields
2. Agronomic problems
3. Production practices
Opportunities to increase productivity
1. With existing technology
2. With emerging technology
2. Weed control
with zero tillage
Age of nursery
a[(Potential yield/farmers' yield)-1]* 100
na = not available
Source: Wheat Byerlee et al. (1986); and Heisey et al. (1987). Rice Herdt and
rotation, planting date, and nutrient balance (see Table 5). The average yield gap
between potential yields that were considered profitable and feasible from on-farm
experiments and surveys, and farmers' current yields is 30-40 percent in each case. Much
of this gain can be achieved within existing expenditure levels by a better mix and timing
and method of application of inputs. Deficiencies in technical knowledge and skills of
farmers were identified in both cases as important factors determining productivity
In both of the above studies the productivity gap was expressed through a yield gap
between farmers' actual yields and economically feasible yields. In several post-green
revolution areas (e.g. parts of the Indian Punjab, Northwest Mexico, Central Luzon of the
Philippines) farmers' yields are now close to this potential and the productivity gap is
expressed in high costs of production relative to the potential (Kahlon, 1984; Byerlee and
Longmire, 1986). Here better information may substitute for high input use, for
example, in the form of lower doses of fertilizer for crops grown in multiple cropping
patterns with substantial nutrient carryover, or in integrated pest management strategies
to reduce pesticide use. For example, Kenmore (1986) estimates that 50 percent of
insecticide applications in rice in Southeast Asia are unnecessary, and that better farmer
information and skills in identifying the threshold pest population for economic
application of pesticide would help reduce this inefficiency.
In sum, the evidence from Asian farm-level research points toward substantial
opportunities to increase productivity through improved economic efficiency. The
evidence seems to indicate that technical inefficiencies occur more consistently and on a
large scale than allocative inefficiencies. However, this finding must be qualified by the
difficulties discussed above of separating the two types of inefficiencies in empirical
studies due to the problems in specifying the production function (i.e. level of
aggregation of inputs). But whatever the type of inefficiency, differences in information
and skills of farmers are usually identified as the major factor explaining the variation in
efficiency between farmers.
This does not negate the Schultzian position of small farmers of the Third World as
rational decision makers responsive to economic incentives. Rather, it suggests that
constraints in development of appropriate rural institutions to service farmers' increased
information and skill requirements, limit farmers' ability to exploit these opportunities to
improve economic efficiency and make rapid adjustments in a technically and
economically dynamic environment. This is not to say that other factors such as market
failure in input markets, capital constraints and risk aversion are not important, but even
these factors may interact closely with decisions on acquisition of information (Feder
and Slade, 1984).23 Moreover, these constraints due to capital and risk aversion have
received relatively greater attention from researchers and policy makers than
constraints due to information and skills.
23Evidence, however, suggest that risk aversion plays a relatively minor role in input
allocation decisions (Roumasset, et. al., 1987).
IV. THE FORMAL INFORMATION SYSTEM INSTITUTIONAL PERFORMANCE
AND INDUCED INSTITUTIONAL INNOVATIONS
The constraints on improving productivity analyzed above reflect inadequacies in
rural institutions agricultural research, extension and education that participate in
the development and dissemination of information and skills to farmers. This set of
institutions is often referred to as the formal information system to differentiate it from
farmers' informal learning-by-doing and experimentation.24 Institutional changes often
lag in adapting to technical change; however, these institutional changes are needed to
realize the full potential provided by new technology (Ruttan, 1978). This chapter
examines some of the problems of these rural institutions in serving a dynamic post-
green revolution agriculture and also analyzes recent institutional innovations aimed at
correcting some of these deficiencies.
A. Adaptive Research25
In most countries, the agricultural research system is a major source of improved
technical information for farmer decision making. For simplicity, agricultural research
can be categorized into science-oriented research to improve the understanding of
physical and biological processes, applied research to generate new inputs or component
technologies, and adaptive research to provide better information on crop management
at the local level, to extension and to farmers. Adaptive research should then play the
primary role in generating useful information for farmers to stimulate changes in
management and input use to increase economic efficiency. Yet according to a recent
World Bank review, adaptive research is generally "the weakest, most neglected and most
confused aspect of national research systems" (World Bank, 1985, p. 54). This situation is
in contrast to the relative strength of applied research, particularly plant breeding
research. In part stimulated by the successes of semi-dwarf varieties in the green
revolution, most Asian countries now have reasonably well-established plant breeding
24The private sector, such as input suppliers, can also be regarded as part of the
information system and will be examined later.
25 Much of the information in this section is based on the author's personal
involvement in adaptive research in several Asian countries.
programs for major food crops capable of sustaining a continuing flow of improved
The "poverty" of adaptive research efforts reflects the common approach of
conducting a series of well-controlled experiments (usually on the experiment station)
and then issuing technical information in the form of recommended "packages" of
practices for large heterogeneous groups of farmers. Typically, each discipline -
agronomy, soil fertility, weed science, entomology, water management, etc. develops
recommendations for practices related to that discipline and these are then "packaged"
without considering interactions between technological components or between
commodities in the farming system. Social scientists who might contribute to the
identification of farmer problems and relevant solutions to these problems have typically
not been included in this research process.
This approach to adaptive research has a number of problems:
1) The information is often not appropriate to farmers because: a) uniform
recommendations are made for large heterogenous groups of farmers, b)
recommendations are generated on the experiment station often under
conditions very different from farmers' fields or c) the socio-economic
circumstances of farmers are not adequately considered especially those due
to complex interactions in the farming system.
2) The recommendations promote a package of several technological components
even though there is considerable evidence that farmers adopt these
components in a stepwise manner (e.g., Byerlee and Hesse de Polanco, 1986;
Crouch, 1981; Herdt, 1987).
3) The information is condensed to simple "recipes" even though farmers require
a much broader range of information and skills to efficiently use complex
technologies and to adapt them to their own economic circumstances, fields
and seasonal conditions.
These problems reflect less the quantity of adaptive research (although
expenditures on applied research have probably expanded more rapidly than expenditures
on adaptive research) than the quality of adaptive research.26 For example, thousands of
fertilizer experiments are conducted annually on irrigated wheat and maize in Asia, yet
26 For example, classifying research on maize and wheat into technology generating
(applied research) and information generating (adaptive research) indicates that about
half of all research expenditures in Pakistan are allocated to adaptive research.
many fertilizer recommendations still lack relevance to farmers' circumstances (Eklund,
1983). Hence, in many areas, a critical weakness in the farm information system is the
inadequacy of the research system in generating a stream of relevant and useful
information for farmers.
B. Agricultural Extension
The poor quality of technical information provided by adaptive research is often
compounded by weaknesses in the quantity and quality of extension advice. Extension
has often assumed a very secondary role in the post-green revolution era. The rapid
spread of the new seeds and fertilizer from farmer to farmer with only minimal input by
extension (Lowdermilk, 1972) seemed to bear out the image of the small farmer as poor-
but-efficient and to down-play the role of extension. A good technology will "sell itself"
might be the logical conclusion from the green revolution experience. The widespread
involvement of extension agents in input delivery, credit programs, and petty rural
administration also allowed little time for their primary role information dissemination
to farmers.27 And in any event, much of the technical information to be disseminated on
cultural practices was not appropriate to the circumstances of farmers, leading to a
credibility problem for those extension agents who did become seriously involved in
The research system has also encouraged extension methods based on a "recipe"
approach to crop production whereby farmers are exhorted to use a rigid technical
package which assumes that fixed technical coefficients apply to all farmers, fields, and
seasons.28 Yet, as shown above, crop management is far too complex for a formula
approach to be used. Typically this formula or recipe has also stressed information on
types and quantities of inputs aimed at increasing yields through use of higher levels of
inputs. Opportunities to increase allocative efficiency within existing resource levels
and to improve technical efficiency have largely been neglected. Emphasis on
communicating recipes has also been at the expense of broader extension education to
improve farmers' understanding of new technology, and enhance farmers' technical and
27See Benor, Harrison, and Baxter (1984), Mohammad (1984) and Roling (1981) for
brief reviews of problems in extension systems.
28For example, in much of South Asia, an annual workshop is held to formulate a
"package of practices" for the coming crop season.
managerial skills. This understanding and these skills are needed if farmers are to adapt
prescriptive information to their own needs and improve their technical efficiency and
allocative ability. Moreover these skills should have a much lower rate of obsolescence
in a dynamic world than prescriptive-type information (Welch, 1978).
Many factors including inadequate training, inappropriate organization and lack of
incentives underlie the poor performance of many extension programs. However, it is
also widely observed that village level extension agents often lack even the basic skills
needed by farmers for effective management of modern inputs. Training courses run by
IRRI (Matheson, 1984) and CIMMYT (in Pakistan) have both observed that most extension
entrants to these courses are not able to calculate dosages correctly for even basic
inputs such as fertilizer, nor are they knowledgeable of the appropriate pesticide for a
given pest. Job incentives have not promoted a problem-solving approach to providing
farmer advice, and wide dispersal of extension agents in the villages complicates
management and supervision. In addition, the lack of linkages between research and
extension systems has often been a serious constraint on the effectiveness of both
These weaknesses in the extension system account for the rather variable findings
on the returns to extension in developing country agriculture (Perraton et al, 1983;
Huffman, 1978; and Lockheed et al., 1980) and even the suggestion that there may be
over-investment in extension (Evenson, 1986). For example, Jamison and Lau (1982) in
their review of 16 studies that analyzed the effect of extension contact on productivity,
found only eight studies with a positive extension impact. Perhaps reflecting this
variable performance, there has been a general decline in extension expenditures relative
to research. In countries of South and Southeast Asia where the green revolution had its
greatest impact, real research expenditures almost tripled from 1970 to 1980 while
expenditures on extension stagnated or even declined in some cases (Table 6). This
change in emphasis also represents a backlash against the heavy emphasis on extension in
the community development strategy of the 1950s and early 1960s. In this period, with
few viable technological improvements (due to neglect of research) for "poor-but-
efficient" farmers operating in a traditional setting, results of this strategy were
generally disappointing (Holdcroft, 1984).
Table 6: Growth Rates of Real Expenditures on Research and Extension, South and
Southeast Asia, 1959-80.
1959 1970 7.4 4.0
1970 1980 9.7 -.6
1959 1970 13.1 9.3
1970- 1980 10.2 1.4
Source: Calculated from Evenson (1986)
C. Recent Institutional Innovations in Research and Extension Systems
Without an effective "formal" information system represented by adaptive research
and extension, a large part of the burden of technology adaptation has fallen on farmers'
own informal learning-by-doing. Indeed, deficiencies in the formal information system
have sometimes led to a credibility problem among farmers and probably discouraged
them from seeking information through the formal system. Not uncommonly, farmers
are ahead of the research and extension system in technology adaptation (Biggs and Clay,
1981) although this "informal" system is not adequate to keep pace with the complexity
and dynamics of post-green revolution agriculture. However, the emergence of a
continuing stream of new technology and the opportunities to improve productivity
through increasing the technical information and skills of farmers, have led to
institutional reforms in adaptive research and extension. Two of these a) the farming
systems perspective in adaptive research and b) the Training and Visit System of
extension are briefly reviewed below.
1. The Farming Systems Perspective in Adaptive Research
A major innovation in recent years has been the farming systems approach to
adaptive research that emphasizes a strong farmer focus and problem-solving orientation
to research (Simmonds, 1986; Byerlee, et. al., 1982).29 In a farming systems approach,
explicit efforts are made to understand the complexity of interactions characteristic of
small farmer systems as a basis for planning research. Adaptive research is largely
carried out in farmers' fields with farmers' participation, using survey and experimental
methods to identify and solve constraints limiting productivity at the local level. The
research objectives call for a multi-disciplinary problem-solving approach involving both
technical scientists (e.g., agronomists) and social scientists (e.g., economists).
Adaptive research programs based on a farming systems perspective are being
tested or have been adopted in most Asian countries and promise to increase both the
quality and quantity of technical information for farmers. Onfarm experiments and a
farmer-orientation in design and analysis of experiments improves the relevance of
production recommendations to farmers. Decentralization of research by focusing on
relatively homogeneous farming systems or recommendation domains ensures that
information is tailored more specifically to farmers' needs. At the same time, there is a
move away from the package approach to providing recommendations to an approach
emphasizing a few priority stepwise changes from farmers' current practices.
The farming systems perspective in adaptive research can be viewed as a way of
combining the contributions of farmers' informal learning-by-doing and the scientific
knowledge and experimental methods of researchers. The approach calls for researchers
to integrate the knowledge and experience that farmers gain in adapting new
technologies into the design of experiments and the formulation of recommendations.
Returns to adaptive research conducted with a farming systems perspective are
potentially high although to date, there are few quantitative estimates of the pay-offs to
this approach to research.30 Returns are expected to be especially high in the irrigated
post-green revolution areas, where a substantial amount of technology is available to be
"adopted" and "adapted". Moreover, the relative uniformity of irrigated areas ensures
that information generated will be relevant to a large number of farmers. To meet the
complexity of crop management decisions, adaptive research programs increasingly
29The farming systems perspective has emphasized adaptive research although the
approach is also very relevant to setting priorities for applied research programs, such as
plant breeding (e.g. Byerlee, Akhtar and Hobbs, 1987).
30Martinez and Sain (1983) estimate high pay-offs to a pilot project in Panama.
provide recommendations conditional on specific field characteristics (e.g. land type,
crop rotation) and seasonal pest and weather conditions.31
Successful adoption of this institutional innovation in adaptive research requires
appropriate incentives for location-specific problem-solving research. This goes against
traditional centralized research planning, fragmentation of research by disciplines and
commodities, and promotion based on publications rather than solutions to farmers'
problem. The weakness of the farming systems approach is that to-date, it has often
been implemented, usually with donor support, as a project outside of the mainstream
institutional structure of agricultural research and has not really addressed the
fundamental weaknesses of research incentives (i.e. a lack of a problem-solving and
client orientation and the need for a systems perspective) that have resulted in poor
quality technical information for crop production (Hienemann and Biggs, 1985). As yet
institutional arrangements have not evolved where the major client of agricultural
research systems, the small farmer, can formally or informally pressure these systems to
address their priorities.
2. The Training and Visit Extension System
The most important recent innovation in extension systems is the Training and Visit
(T & V) System now adopted in many Asian countries usually with World Bank funding
(Benor, Harrison, and Baxter, 1984). The T & V system addresses a number of the basic
weaknesses in traditional extension systems through the following measures:
1) Non-extension duties (e.g. credit supervision, input distribution, statistical
data collection) are removed from the workload of extension agents to allow
them to focus on information dissemination.
2) Extension agents are unified under a strong management system and well
defined duties and routines are assigned to each level of the hierarchy.
3) Regular training programs are established to upgrade the skills of extension
4) The ratio of village extension agents to farmers is increased and extension
agents are required to regularly visit "contact" farmers in each village.
Improved mobility (e.g. bicycles) is often provided for this purpose.
5) Efforts are made to bridge the gap between adaptive research and extension.
31See Byerlee (1987) for a more comprehensive discussion of these issues.
Early experiences with the T & V system indicate mixed successes. Feder, Lau and
Slade (1985) estimated a 6-7% increase in wheat productivity in Haryana State of India
due to improved technical efficiency attributed to introduction of the T & V system.
This implies high returns to the investment in this institutional innovation.32
Preliminary data from some other states of India (Shingi et al., 1982; and Benor, Baxter
and Harrison, 1984) as well as from Nepal (3amison and Moock, 1984) support these
findings. However, other observers in India (Moore, 1984; Howell, 1982) note that
extension advice is still not relevant to many farmers and quantitative targets for farmer
contacts are emphasized over the quality of information disseminated. Elsewhere, the
results are less encouraging. Khan et al. (1984) found no effect of the T & V system in
Pakistan's Punjab on either productivity or knowledge of farmers. The lack of a strong
adaptive research program may partly explain the failure of the T & V system in
Pakistan.33 Overall, the experience with T & V extension is still too short to draw
definite conclusions. Institutions are notably slow to evolve and especially when an
innovation emphasizes improvements in human capital and management it may take a
decade or more before the reforms become effective.
A major shortcoming of the T & V system for farmers in post-green revolution
agriculture in Asia is its emphasis on the communication role of extension that is,
transferring specific "messages" as prescriptive information to farmers. It has yet to
meet the need for farmers to have a better understanding of new technology and
improved technical and managerial skills. In post-green revolution areas these principles
and skills include diagnostic skills on factors reducing yields, technical knowledge of
chemical inputs such as residual effects or downside risks from untimely application, as
well as specific technical skills such as calibration of knapsack sprayers or computation
of nutrient doses for compound fertilizers. Eventually, as farmers' technical skills
improve, extension efforts might shift to enhancing farmers' managerial skills, including
the ability to recognize problems and seek out additional information. This change in
32Feder, Lau, and Slade (1985) estimated that returns on investment were at least
15% annually within a 90% confidence interval. This is probably a conservative estimate
since they did not attempt to measure improvements in allocative efficiency.
331n Pakistan, separate adaptive research programs were introduced with the T & V
system and these programs have yet to produce useful recommendations for farmers. In
addition, there appear to be significant lapses in extension management. In the survey by
Khan et al. (1984) only half of farmers designated by extension as "contact" farmers were
aware that they were in fact contact farmers with special obligations to disseminate
extension messages to other farmers.
emphasis from communication of crop production "recipes" to education in crop
production principles and skills recognizes the growing complexity of crop management
and the need for farmers operating different land types and crop rotations in an
uncertain and dynamic environment, to adjust technical information to their own specific
The shift in extension emphasis from a communication role to more of an
educational role requires continual upgrading of the quality of extension staff.34 One
successful example, has been training in the complex principles of integrated pest
management for rice in several countries in southeast Asia. This program has generated
a payoff of 440 percent (undiscounted) due to a reduction in pesticide use (Kenmore,
1986). It is significant that a large part of this program was devoted to field-oriented
training of both extension agents and farmers in broad principles of pest management as
well as specific skills in pest identification, subjective scoring of pest densities and
Institutional innovations in extension also suffer similar problems to adaptive
research in being imposed from the "topdown", usually with donor support. The problem
is how to maintain a client orientation to extension in the longer run. Farmer and village
level associations which can exert pressure on the performance of local extension agents
have had some success in performing this role (Lionberger and Chang, 1970; Stavis, 1979)
but are not a part of the T & V system.
D. Rural Schooling
Beyond adaptive research and extension systems, the other major source of
increasing knowledge and skills for a scientific agriculture is rural schooling.
Expenditures on rural education have been one of the fastest growing sectors in the
developing world and primary school enrollments have increased steadily. In the period
1960 to 1982 the proportion of children enrolled in primary schools increased from 55%
to 75% in South Asia and from 83% to 101% in Southeast Asia.35 Studies of economic
34This change in roles also implies changes in extension methods. Mass media
which may be appropriate for communicating messages is probably less effective for
teaching principles and skills than informal and formal training programs for individual
farmers and groups of farmers.
35 Calculated from World Bank, World Development Report, 1984, Table 25.
returns to education also show attractive pay-offs to investments in basic education,
generally (Colclough, 1982) and in agriculture, specifically (Jamison and Lau, 1982). To
the extent that these national figures reflect expenditures and returns in rural areas
where the majority of the population resides, it would be easy to conclude that
institutional changes are already underway in the educational sector to meet the growing
need to increase the general educational level of farmers. However, these data are
misleading for a number of reasons.
First, there are still large numbers of farmers in post-green revolution areas,
especially in South Asia, who lack basic numeracy and literacy skills (Table 7). Adult
literacy rates in rural areas of Pakistan, Bangladesh and Nepal average 25% or less and in
northwest India are less than 50%. Even if most of the children of these farmers attend
school and complete a basic education it will require at least another generation to
achieve a minimally educated population of farmers. Added to this is the problem that
drop-out rates even for primary schooling, are often close to 50% in rural areas.
Table 7: Percent Adult Population (25+ years) Who Have Attended School in Selected
Bangladesh 1974 18
India 1971 21
Nepal 1971 4
Pakistan 1981 18
Sri Lanka 1971 68b
Southeast and East Asia
China 1982 55
Indonesia 1980 55b
Korea 1980 80
Philippines 1975 86
Thailand 1980 79
aFigures are averages of male and female adults. Especially in South Asia, female
education is much lower than for males.
bRural population only.
Source: United Nations, Demographic Year Book, New York, 1983.
Second, there is now growing concern that the quality rather than the quantity of
education may be a major limitation (Behrman and Birdsall, 1983; Heyneman, 1983).
Educational quality is difficult to measure and even more difficult to relate to
agricultural productivity. Even evidence on returns to higher levels of education
(presumably a proxy for quality, at least for basic numeracy and literacy skills) is scanty
and conflicting for agricultural settings. Butt (1984) found a significantly higher
productivity of farmers with secondary schooling relative to primary school education in
Pakistan while Kalirajan and Shand (1985) in India found no impact of educational level
on agricultural productivity from primary school to university level. We would expect
that increasing complexity of crop management in post-green revolution agriculture
would lead to increased returns to quality and level of education. Heyneman (1983) has
expressed educational requirements for different levels of agricultural technology that
suggest a minimum of lower secondary school education for much of irrigated Asia.36
These standards are probably too rigid since extension in its educational role can partly
substitute for formal schooling, a theme to be addressed in Chapter 5.
Finally, related to the issue of educational quality is the fact that both private and
public demand for education has largely been driven by the urban employment market,
rather than agricultural employment (Todaro, 1985), presumably because returns to
education in urban employment are perceived to be higher than in farming. Given the
evidence that both social and private returns to education in a modernizing agriculture
are quite high (Jamison and Lau, 1982), this suggests a lag in adjustment of rural
household investment decisions to the new situation in the agricultural sector. It also
increases the rate of out-migration of educated youth, especially those with secondary
schooling, beyond what might be socially desirable (Todaro, 1985) and depletes the stock
of human capital needed for sustaining agricultural productivity increases.
36For irrigated post-green revolution areas, Heyneman (1983) (citing A. Harma)
gives the minimal requirements as: mathematics, independent written communications,
high reading comprehension, ability to research key words and concepts; elementary
chemistry, biology and physics. Leaf (1984) suggest somewhat similar needs for farmers
in the Indian Punjab.
V. FARM INFORMATION AND SKILLS: IMPLICATIONS
FOR DEVELOPMENT STRATEGY
The critical importance of farmer information and skills in maintaining
productivity increases in post-green revolution agriculture has far-reaching implications
for the design of agricultural development strategies. The total information and skill
system consists of public sector institutions, the private sector and farmers, as well as
policies which affect the technical and economic environment in which these institutions
and agents operate (Figure 2). The development of an appropriate strategy to create an
effective information and skill system depends on four key inter-related issues discussed
in this section.
1) Within the formal information system of the public sector, what is the
appropriate mix of adaptive research, extension and education to exploit the
complementary and substitution relationships between these different types of
investments? What institutional mechanisms are available to promote
2) What is the potential role of the private sector (input suppliers or specialized
information markets) in complementing and eventually substituting for public
sector information services?
3) To what extent can other policies substitute for the formal information and skill
system either through reducing the cost of farmers' own informal learning and
experimentation, or through applied research efforts to develop less
4) What are the equity implications of increasing returns to information and skills
in post-green revolution agriculture?
A. Integrating Research, Extension, and Rural Schooling
Farmers in post-green revolution agriculture require a farm information and skill
system that a) generates and communicates useful technical information on a continuing
basis, b) develops farmers' understanding of new technologies, c) provides basic literacy,
numeracy, and cognitive skills, and d) develops farmers' technical and managerial skills.
Table 8 summarizes the role of each institution in providing these products.
i j d Complexity of Technology
Figure 2. Components of farmers' information and skill system.
Table 8: Institutions and Functions Servicing the Information and Skill Needs of
Adaptive Research Extension Schools
Functions (** Main Role) (* Secondary Role)
information ** *b
information .a ** *
(technical & managerial) .a ** *
Developing general skills
(literacy, numeracy, etc.) **
aThrough involvement of farmers in adaptive research
bThrough farmer testing, feedback and comparisons across farmers.
A strategic policy question is the relative priority that needs to be given to each
institution and function at different stages of development, and in particular in the post-
green revolution era. Much depends on the extent to which the various elements are
regarded as substitutes or complements.
Given the management intensity of many second generation technologies, adaptive
research and extension are expected to be highly complementary. Increasing amounts of
information generated by adaptive research programs will be wasted or will diffuse too
slowly (and possibly become outdated) without a strong extension input. Likewise,
extension programs will have little impact and may lose credibility if they do not have
useful and relevant information to extend. To a limited extent, extension might operate
independently of research by utilizing experiences of successful or innovative farmers in
formulating extension advice. The relatively high degree of variation in production
techniques between farmers (as seen earlier in Table 2) suggests that this avenue could
be useful.37 However, this is a short term solution that gives rapidly diminishing returns
unless new information is continuously supplied by research.
To exploit their complementarity, adaptive research and extension must be closely
linked. One of the major problems in most countries is the lack of effective linkages
between research and extension (e.g. World Bank, 1985; Howell, 1982). Often they are
institutionally separated, with extension in the Ministry of Agriculture and research in a
parastatal organization. Differences in incentives and prestige also undermine working
relationships. Many research systems, believing that there is a wide "technological gap"
that is not being effectively addressed by existing extension systems, have established
their own technology transfer and extension activities, such as the "Lab-to-land" program
in India and "crop maximization" programs in Pakistan (Mohammad, 1984).
It is a feature of both the farming systems perspective in adaptive research and the
T & V extension system that they emphasize close research-extension linkages, especially
in verification of technology at the farm level. Researchers are also expected to play a
leading role in upgrading the knowledge and skills of extension workers through training
programs. Despite this emphasis, poor linkages between research and extension remain a
major weakness of most farm information systems (Cernea et. al., 1983; Howell, 1982).
This weakness is most apparent in feeding back information from farmers via extension
to research. A client-oriented extension system with farmers able to influence extension
performance (e.g., through farmer associations) and extension able to influence research
priorities is probably one of the most effective models for farmers to exert pressure on
research (e.g., Lionberger and Chang (1970) for Taiwan).
Whether extension and formal schooling are complements or substitutes is also a
critical policy question. The relationship between them is likely to be quite complex. To
the extent that extension emphasizes its communication function, farmers receiving the
information may use it more effectively if schooling helps them to understand the
rationale behind the recommendation. Where extension emphasizes its educational role,
it could also partly substitute for formal schooling. However, in either its
communication or educational role, the cost of extension should be significantly reduced
by competencies in literacy, numeracy and cognitive skills imparted through schooling.
37This is only true to the extent that variation in production technique reflects
differences in information and managerial skills and not micro-environmental variation
(e.g., land type, soils). Recent experiences in Pakistan indicate that variation in yields
between fields can generate hypotheses on the response to different management
practices (Hussain et al., 1986; Byerlee et al., 1986).
Hence, extension might be able to substitute for education in the early stages of
development, but diminishing returns will soon be reached unless there are advances in
basic education. In a dynamic agriculture where extension advice is subject to rapid
depreciation, the viable skills will be those of "learning to learn" imparted by formal
schooling (Welch, 1978). Sims (1985) suggests that one of the reasons for the higher
productivity of farmers in the Indian Punjab compared to the Pakistan Punjab is that
Indian farmers have better technical information, not because of better extension (which
she rates as poor in both cases), but because the higher level of education of the Indian
farmers acts as a substitute for extension. Better educated farmers can exploit a wider
range of information sources. At the extreme, the role of extension may decrease for
highly educated farmers who have access to a wide range of alternative sources of
Empirical studies of the interaction effects of education and extension on
agricultural productivity reflect these conflicting trade-offs. Some studies show that
education and extension are substitutes (e.g., Moock, 1981) while others find little effect
(e.g., Pudasaini, 1983). More research is needed to determine the extent to which
extension programs might substitute for education at different levels of agricultural
Where educational levels are very low, especially in South Asia, there is a danger
that even if recent institutional innovations in research and extension are successful, the
low level of education of most farmers will be a binding constraint on rapid increases in
productivity in the medium term. Formal schooling is a long term investment that will
only affect the productivity of the next generation of farmers. In post-green revolution
areas, dramatic changes in managerial requirements have taken place within one
generation of farmers as shown by the wide array of purchased inputs already adopted or
in the process of adoption. The question then arises on the role of adult education
programs as a means for effecting short run improvements in general educational
levels. Some limited evidence exists that adult literacy classes have positive effects on
agricultural productivity (Fuller, 1983; Jamison and Lau, 1982) but as yet too little is
known about the specific products of formal schooling that increase agricultural
productivity, and the extent that these products can be transmitted through adult
38This may represent the situation of farmers in industrialized countries that own
micro-computers and can use a wide array of information (Sonka, 1985).
Overall, the evidence suggests strong complementarities between each of the
institutions and functions bearing on information and skills in agriculture (Table 8).
Hence it will be important to identify the weak points in the system. To date, those
involved in research and extension have often ignored the role of formal schooling while
many of the studies of education and agricultural productivity have lacked an adequate
base in technical agriculture.
B. The Private Sector and Information Generation and Dissemination
In industrialized countries the private sector plays a major and increasing role in
the generation and transfer of better technical information to farmers (e.g. Ruttan,
1982; Bonnen, 1986; Turpin and Maxwell, 1976). It does this through its own adaptive
research programs, advertising, promotion and demonstration programs, dissemination of
information through input suppliers and by providing specialized information services
(e.g., magazines, consultants, soil testing services etc.).
In post-green revolution areas, the private sector has often been slow to adopt
these roles, despite the rapidly increasing use of purchased inputs and the associated
demand for better information. The most obvious opportunity for the private sector is
the dissemination of information in association with input sales, especially chemical
inputs. Some studies do note the importance of input dealers as a source of information
(e.g. Rogers and Meynen, 1975; Pontius, 1983) but they are often regarded by farmers as
a secondary source of information (Heisey et al., 1987; Feder and Slade, 1985; Litsinger
et al., 1980; Gill, 1982). A number of factors explain this anomaly. In the early stages of
the green revolution, the public sector often played a major role in the delivery of inputs
to farmers, often through the extension service. With demand for inputs and credit
established and the need to move extension back to their basic task of information
dissemination, the private sector has assumed a larger role in input dissemination.
Hence, the private sector is often a relatively recent entrant in input distribution at the
farm level, and needs to develop its own capacity and knowledge in the local use of the
inputs. This is especially the case for input dealers in direct contact with farmers who
are often local shop keepers whose major business is consumer goods.
Private firms must also establish credibility as a source of reliable information.39
Private sector promotion of inappropriate chemicals and pressure (and even bribing) of
39Claims of adulteration of chemicals are widespread by farmers in many countries,
and tests of inputs often confirm these claims (Goodell, 1984a; Hussain et al., 1985.)
public agencies to distribute a particular product through official credit programs are not
uncommon. Even deliberate misinformation is reported, such as the distribution of a
fungicide as a "growth enhancer" (Kenmore, 1986). There is also a natural tendency for
chemical companies to recommend above-optimum doses and to prefer prophylactic
treatments to treatments conditional on the specific farmer, field and seasonal
circumstances (Zadoks, 1985; Kenmore, 1986).40
Despite these problems, the role of the private sector in distributing information in
association with inputs is likely to increase, especially as specialized input dealers
assumed a greater role in input sales. Hence efforts to train local input dealers in use of
modern inputs offers a cost effective opportunity to improve information flows to
farmers. For example, in Bangladesh, the government has arranged training programs for
fertilizer distributors, while chemical companies are participating in training of pesticide
dealers (C. Pray, personal communication). The increasing trend toward private sector
involvement in applied research, especially in development of hybrids, should also provide
more incentive for private sector extension initiatives. For example, a privately owned
hybrid maize program in Pakistan's Punjab has been notably successful in raising
productivity of maize in one area through employment of its own extension agronomists
to advise farmers.
The combination of deficient public sector extension efforts and farmers'
increasing need for technical information should also provide the environment for
developing specialized information services by the private sector where farmers pay for
information services. Such institutional arrangements are relatively rare for food crop
production in developing countries. In commercial areas of northwest Mexico, Uruguay
and Argentina, private consulting services play an increasing role in advising medium and
large scale farmers. Likewise, there are examples of small farmers forming cooperative
extension services paid for by members of the cooperative (e.g., the farm extension
service for collective farmers in northwest Mexico, pest surveillance services in the
Philippines and farmer association extension services in Taiwan.) Not surprisingly, these
private extension and consulting services have a sound reputation for effectiveness.
These institutional arrangements are likely to expand in other areas where there is a high
payoff to better information and public extension services are not responding to this
demand. However, the high relative cost of information for small farmers (Feder and
40For example, farmers in the central plateau of Mexico have evolved an effective
dosage of 2-4, D herbicide for wheat and barley which is less than one-half the dose
recommended by the manufacturer.
Slade, 1984) limits the development of specialized markets for information for small
farmer agriculture. Finally, where specialized technical skills are needed, such as in
pesticide application, the private sector by offering contractual services for input
application can also play a role in increasing technical efficiency. Specialized
contractors with more information and experience, as well as more specialized
equipment, should be able to substitute for farmers' lack of technical skills.
C. Substituting for Farmer Information and Skills
While it seems inevitable that increasing complexity and commercialization of
agriculture will place greater demands on farmers' information and skills, there may be
ways to partly alleviate these demands. Applied research and particularly plant breeding
research can, and often is, aimed at the limited managerial capacity of small
farmers.41 For example, plant breeding programs for small farmers generally give much
higher weight to pest resistance in varietal selection than similar programs aimed at
commercial agriculture. This is partly to reduce expenditures on pesticides but it also
substitutes for extension resources and the managerial capacity of small farmers.42
However, genetic resistance to many pests erodes over time (due to genetic
adaptation by the pest population) so that a continual stream of new varieties is required
to maintain pest protection. This wider selection of varieties tends to increase farmers'
management complexity.43 More importantly, for farmers who do not understand this
breakdown in pest resistances, the challenge to plant breeders is even greater, since
succeeding generations of varieties must have other superior traits, especially higher
yields, to encourage rapid farmer adoption. 44
41Conversely in developed agriculture, plant breeding might aim at relatively high
managerial capacity. For example, maize hybrids appear to be superior to open-
pollinated varieties under high input and management but this difference is reduced at
lower levels of management.
42For example, both CIMMYT and IRRI devote a large share of their crop
improvement programs to genetic resistance to insects and diseases.
431n irrigated wheat, the expected effective life of a variety is about five years.
44Severe rust attacks in wheat occur only sporadically so that breakdown in rust
resistance may not be obvious for some years. In Pakistan only 25% of farmers in one
recent survey understood the potential breakdown of disease resistance and the need to
continually update varieties (Heisey et al., 1987).
Broad adaptability of varieties is another trait which can partly substitute for
extension and managerial capacity of small farmers. A broadly adapted variety that does
well over a range of conditions reduces the complexity to extension and to farmers of
recommending a number of individual varieties for specific conditions. For example,
farmers often plant wheat in irrigated areas over a range of planting dates depending on
the crop rotation in specific fields. Wheat breeders have traditionally developed
separate varieties for normal and late planting. However, recognizing the managerial
complexity of this strategy, breeders are now screening for single varieties that do well
over a range of planting dates. Incorporating these breeding objectives to accommodate
limited managerial capacity of farmers will, of course, be at the expense of more rapid
growth in yield potential. However, where there is a wide gap between farmers' yields
and potential yields of existing varieties this does not seem a high price to pay in the
Similar principles can be applied to chemical inputs. Pesticides which have a broad
spectrum of application to several pests are easier for farmers with limited information
to apply than narrow-spectrum pesticides that require accurate identification of the
major pests and sometimes the mixing of two or more pesticides.45 Likewise,
"management neutral" pesticides that are effective over a range of dosages and times of
application will require less technical skills for successful use.46 Research on herbicides
that can be applied in granulated form in irrigation water, and on slow-release nitrogen
fertilizers are other examples of attempts to increase technical efficiency of input use,
through reducing "management sensitivity". In general, these types of applied research
activities aimed at reducing demands on extension services and farmers' management
have, with the exception of genetic resistance to pests, received less emphasis than they
45This is particularly true for weedicides. Broad-spectrum insecticides are likely to
have the disadvantage of killing predator insects as well.
46The sensitivity to dosage is important not only to allow for errors in calculating
dosage and mixing the product, but also in ensuring a uniform response within the field.
The latter is often a major constraint on technical efficiency in fertilizer and pesticide
use. Simple methods of application, such as hand spreaders for broadcasting chemicals
also have much potential for reducing technical inefficiencies by helping to provide a
more uniform application of the product in the field.
471n many cases, this may reflect the fact that most research and development in
agricultural chemicals is undertaken by the private sector for farmers in industrialized
countries with quite different levels of technical skills and better access to technical
Another alternative to complement the formal information system is to reduce the
cost of farmers' own learning-by-doing through subsidized input prices and credit
programs. Subsidies on chemical inputs have been a widespread policy response. For
example, fertilizer subsidies in Pakistan are strongly biased toward phosphatic fertilizers
in the belief (debatable) that this is the limiting nutrient for most crops and regions.
While these subsidies can help reduce subjective risk and speed early adoption, they are
politically difficult to reduce and often account for a high proportion of total
government expenditures for the agricultural sector. They may also eventually lead to
input use above the social optimum (Stoneman and David, 1986) as has occurred in
irrigated wheat in northwest Mexico (Byerlee and Longmire, 1986) and with pesticide use
on rice in Indonesia. Hence, the high cost of these subsidies must also be evaluated
against investments in adaptive research and extension to provide better information to
In addition to subsidies, governments frequently try to "force" the use of a
technological package through "tied credit". In these programs farmers are required to
use the recommended package as a condition for loans from official credit banks, usually
at low or negative real interest rates. This system if enforced, allows little opportunity
for farmers to adapt technology to their own circumstances and frequently leads to
inefficient input use due to the inappropriateness of the recommended package to
individual farmers. More importantly, by fixing the technological coefficients, farmers
are discouraged from developing their own knowledge of the technology through informal
experimentation on different levels and combinations of inputs under their own
conditions (Scobie and Franklin, 1977).
D. Information, Skills, and Equity
The green revolution provoked major controversies on the equity effects of
technological change. The accumulated body of evidence now indicates that small
farmers quickly followed large farmers in using the seed and fertilizer technology, and
that the technology was essentially scale neutral (Ruttan and Binswanger, 1978; Anderson
et al., 1986; Lipton and Longhurst, 1985). There is substantial evidence that input and
water supply systems and credit were initially biased toward large farmers, but in most
cases this bias has been reduced over time, allowing small farmers to benefit equally (on
a relative basis) from green revolution technologies. However, differential access to
information and education is a continuing problem in post-green agriculture and increases
the potential for growing inequities in the future. This is especially so in South Asia
where large differences in farm size persist in many areas. Adaptive research that
conducts most experiments on research stations is more likely to produce information of
greater relevance (or less irrelevance) to large farmers. Likewise, extension programs in
many countries have long been criticized for their emphasis on large farms (e.g., Roling
et al., 1981). The diffusion model of innovations by identifying the "progressive" farmer
as the innovator, has led to a deliberate bias of extension systems toward large farmers
(Roling et al., 1981).48
There is also considerable evidence that access to education in rural areas is
closely related to income, wealth and social status of rural households (Psacharopoulos
and Woodall, 1985).49 As the value of education in agricultural production increases with
technological change, inequalities in schooling have potentially long run implications for
Recent evidence from Pakistan demonstrates these biases (Table 9). Although
education and literacy levels are generally low and extension services deficient, there is
a marked bias in these services towards large farmers, most of whom had received some
basic education and had some contact with extension.50 But even most large farmers
had not completed lower secondary schooling, which is hypothesized to be the minimal
requirement for efficiently adapting post-green revolution technologies.
The payoff to better information and skills is also hypothesized to increase with
farm size. Larger farmers have greater incentives to seek additional information, and
this in part explains the earlier adoption of green-revolution technology (Feder and Slade,
1984). Farm size will also increase the returns to allocative ability and hence education
Recent institutional innovations in adaptive research and extension can potentially
reduce inequities in access to information and skills. The farming systems perspective in
adaptive research emphasizes an understanding of small farmer circumstances as a basis
for designing technologies. Hence the probability that research recommendations are
48Most studies of recipients of extension advice provide empirical support for this
bias (Garforth, 1982).
49For example, the school enrollment ratio for boys in Gujrat and Maharastra States
in India for the wealthiest 10% of rural households (55%) is double that for the poorest
10% of households (23%). The differences is even larger for girls (Psacharopoulos and
50These differences in access to information and education are rarely considered in
the voluminous literature for South Asia on the efficiency of small versus large farms.
Table 9: Relationship Between Extension Contacts, Education, and Farm Size, Punjab,
Less than 5 to More than
5 ha 10 ha 10 ha
Percent farmers with
extension contact in
past year 13 24 61
Percent farmers literate 32 50 71
Percent farmers some
secondary schooling 14 25 32
Source: Original data from survey of 300 farmers in 1986.
appropriate to small farmers is enhanced. The T & V extension system by increasing the
mobility of extension workers and the number of field visits should allow greater access
by small farmers to extension services. A key element in the T & V system, however, is
the selection of "contact farmers" who receive the extension messages with the
understanding that they will pass them along to other farmers. Some controversy exists
as to whether contact farmers are representative and, if not, the extent to which
extension advice is monopolized by large farmers (see Moore, 1984 and Feder, Slade and
Sundaram, 1986). In a stratified rural society, differences in social and economic
circumstances between farmers are likely to be major barriers to the transfer of
information from large to small farmers.
VI. CONCLUDING COMMENTS
The changes initiated by the green revolution have revolutionized the technology of
rice and wheat production in much of Asia and has had a profound effect on the
managerial complexity of small farmer agriculture. A major premise of this paper is
that in many of these post-green revolution areas, knowledge and skills of farmers have
become critically limiting factors in maintaining increases in productivity. Investments
to increase the quantity of technical information and develop the technical and
managerial skills of farmers have not kept pace with investments in developing new
technology. More importantly, institutional changes in research, extension and rural
schooling needed to improve the quality of information and skill development have
limited the opportunity to exploit the potential of the new technology, resulting in
substantial technical and allocative inefficiencies in post-green revolution agriculture.
The endurance of Schultzs' "poor but efficient" hypothesis in development thought,
aid agencies and national policy makers has maintained emphasis on the "high pay-off
input" strategy to development and slowed the shift in priorities toward investment in
information generation and transfer, and skill development for farmers.51 Indeed, the
pendulum seems to have swung from viewing small farmers as ignorant and tradition
bound to a situation where they are looked to as an example of rational decision making
and a store of knowledge from which scientists should learn. The challenge is to combine
the knowledge and insights of farmers of their environment with the information and
skills generated by research, extension and formal schooling that are needed for
effective management of science-based agricultural technology.
The increased emphasis on farmer-oriented adaptive research (the farming systems
perspective) and extension reform in the 1980s represents the beginnings of a process to
alter the balance. Appropriate institutional arrangements to accommodate these
changes are still evolving. Moreover, investment in adaptive research, extension and
rural education is still inadequate in many areas. Unless these imbalances are corrected
there is a danger of further increasing inequalities in the agricultural sector between
small and large farmers due to differential access to knowledge and skills.
Agricultural economics and other social science research is needed to guide critical
decision making in investment and institutional change. These research needs include the
51This is particularly ironic given Schultz's championship of the role of human capital in
agricultural change (e.g., Schultz, 1975).
relatively neglected topics of managerial processes of small farmers, appropriate
institutional arrangements for inducing a problem-oriented research and extension
system and the more general but critical questions relating to the complementarity and
substitutability between applied research, adaptive research, extension communication,
extension education, rural schooling and adult education.
Adams, 3. "Peasant Rationality: Individuals, Groups, Cultures," World Dev. 14 (1986):
Ali, M. and J.C. Flinn, "Profit Efficiency Among Basmati Rice Producers in Pakistan
Punjab." Paper presented to National Rice Seminar, Islamabad, April, 1986.
Anderson, 3.R. "International Agricultural Research Centers: Achievements and
Potential," Draft Report, Consultative Group on International Agricultural
Research, Washington, D.C., 1985.
Anderson, 3.R. "One more or Less Cheer for Optimality". J. Aus. Inst. Agric. Sci.
Antle, 3. M. "Human Capital, Infrastructure and the Productivity of Indian Rice
Farmers," 3. Dev. Econ. 14 (1984): 163-81.
Armenia, P. "Resource Productivity Estimates on Lowland Rice Farms in Baybay,
Leyte." Ann. Trop. Res. 5(1983): 6-12.
Barnum, H. N. and L. Squire. "Technology and Relative Economic Efficiency." Oxford
Econ. Papers 30 (1978): 184-198.
Behrman, J. D. and N. Birdsall. "The Quality of Schooling: Quantity Alone is Misleading,"
Amer. Econ. Rev. 73 (1983): 928-46.
Belbase, k. and R. Grabowski. "Technical Efficiency in Nepalese Agriculture." J. Dev.
Areas. 19(1985): 515-26.
Benor, D., J. Q. Harrison and M. Baxter. Agricultural Extension: The Training and Visit
System, Washington D.C.:World Bank, 1984.
Bernsten, R.H. "Constraints to Higher Rice Yields in the Philippines." Unpublished Ph.D.
Thesis, University of Illinois, Champaign, Urbana, 1977.
Biggs, S. D. and E. J. Clay. "Sources of Innovation in Agricultural Technology," World
Dev, 9 (1981):321-36.
Bhati, U.N. "Farmers' Technical Knowledge and Income; A Case Study of Padi Farmers of
West Malaysia", Malaysian Econ. Rev. 18(1973): 36-47.
Bliss, G. 3. and N. H. Stern. Palanpur: The Economy of an Indian Village. Oxford:
Clarendon Press, 1982.
Bonnen, 3.T. "A Century of Science in Agriculture: Lessons for Science Policy." Amer.
3. Agric. Econ. 68(1986): 1065-1080.
Butt, M. S. "Education and Farm Productivity in Pakistan," Pakistan 3. Applied Econ.,
Byerlee, D. "From Adaptive Research to Farmer Recommendations and Extension
Advice." Agric. Admin., (forthcoming, 1987).
Byerlee, D., M. R. Akhter and P. R. Hobbs. "Reconciling Conflicts in Multiple Cropping
Systems through Plant Breeding Research: The Example of Cotton and Wheat in
Pakistan's Punjab," Agric. Systems (forthcoming, 1987).
Byerlee, D., L. R. Harrington and D. L. Winkelmann, "Farming Systems Research: Issues
in Research Strategy and Technology Design," Amer. 3. Agric. Econ. 64 (1982): 897-
Byerlee, D. and E. Hesse de Polanco. "Farmers' Stepwise Adoption of a Technological
Package: Evidence from the Mexican Altiplano." Amer. 3. Agric. Econ., 68(1986):
Byerlee, D., P. Hobbs, M.R. Akter and A. Majid. "Development of Improved Crop
Technology within the Context of Pakistan's Multiple Cropping Systems." Pak. 3.
Agric. Soc. Sciences., I (1986); 1-28.
Byerlee, D. and 3. Longmire, Comparative Advantage and Policy Incentives for Rainfed
and Irrigated Wheat in Mexico. Economics Working Paper 86/1, CIMMYT, Mexico,
Byerlee, D. and G. Sain. "Food Pricing Policy in Developing Countries: Bias Against
Agriculture or For Urban Consumers?" Amer. 3. Agric. Econ., 68(1986); 961-69,
Carter, M.R. "The Definition and Estimation of Technical Efficiency Using Frontier and
Other Omitted Variable Estimators," Unpublished paper, Department of Economics,
University of Wisconsin, Madison, 1982.
Cernea, M. M., 3. K. Coulter and 3. F. A. Russell (eds) Agricultural Extension by
Training and Visit: The Asian Experience. A World Bank and UNDP Symposium.
Washington D.C.: World Bank, 1983.
Colclough, C. "The Impact of Primary Schooling on Economic Development: A Review of
the Evidence." World Dev. 10(1982): 167-85.
Crouch, B. R. "Innovation and Farm Development: A Multidimensional Model." In
Extension Education and Rural Development. B. R. Crouch and S. Chamala (eds),
New York: John Wiley & Sons, 1981.
Dalrymple, D. G. Development and Spread of High Yielding Wheat Varieties in
Developing Countries, Bureau of Science and Technology, USAID, Washington, D. C.:
Dalrymple, D. G. Development and Spread of High Yielding Rice Varieties in Developing
Countries. Bureau of Science and Technology, USAID, Washington, D.C. 1986b.
Dillon, 3. L. and 3. R. Anderson. "Allocative Efficiency, Traditional Agriculture and
Risk." Amer. 3. Agric. Econ. 53 (1971): 26-32.
Doll, 3. P. "On Exact Multi-Collinearity and the Estimation of the Cobb-Douglas
Production Function." Amer. 3. Agric. Econ. 56 (1974): 556-63.
Eklund, P. "Technology Development and Adoption Rates." Food Pol. 8 (1983): 141-53.
Evenson, R. E. "The IARCs: Evidence of Impact on National Research and Extension and
on Productivity," Draft paper, Yale Univ., New Haven, Conn. 1986.
Farrell, M. 3. "The Measurement of Productive Efficiency." 3. Royal Stat. Soc.,
Feder G., R.E. Just and D. Zilberman, "Adoption of Agricultural Innovations in
Developing Countries; A Survey". Econ. Dev. Cult. Change 33 (1985): 255-98.
Feder, G. and R. Slade. 'The Acquisition of Information and the Adoption of New
Technology." Amer. 3. Agric. Econ. 66 (1984): 312-20.
Feder, G., L. 3. Lau and R. H. Slade. The Impact of Agricultural Extension: A Case
Study of the Training and Visit System in Haryana, India, Staff Working Paper No.
756, Washington, D.C.: World Bank, 1985.
Feder, G. and R. Slade. "The Role of Public Policy in the Diffusion of Improved
Agricultural Technology." Amer. 3. Agric. Econ. 67 (1985): 423-28.
Feder, G., R. H. Slade and A. K. Sundaram. "The Training and Visit Extension System:
An Analysis of Operations and Effects." Agric. Admin., 21 (1986): 33-60.
Flinn, 3. C. and M. Ali. "Technical Efficiency in Basmati Rice Production." Draft
paper. Los Banos: International Rice Research Institute, 1986.
Forsund, F.R., C.A. Lovell and P. Schmidt, "A Survey of Frontier Production Functions
and Their Relationship to Efficiency Measurement", 3. Econometrics, 13 (1980): 5-
Fuller, R. H. "Education, Productivity and the Social Organization of Production in the
Rural Sector: A Case Study of Rice Farmers in Northeast Bangladesh." Unpublished
Ph.D. Dissertation, Florida State Univ., 1983.
Garforth, C. "Reaching the Rural Poor: A Review of Extension Strategies and Methods."
In Progress in Rural Extension and Community Development, Vol. 1, G. E. Jones and
M. 3. Rolls (eds), Chichester; N.Y: John Wiley, 1982.
Gill, G. J. "Pilot Survey of Farmers' Sources of Information". Draft paper, Dakka:
Bangladesh Agricultural Research Council, 1982.
Goodell, G. E. "Bugs, Bunds, Banks, and Bottlenecks: Organizational Contradictions in
the New Rice Technology." Econ. Dev. Cult. Change 33 (1984a): 23-41.
Goodell, G. E. "Challenges to International Pest Management Research and Extension in
the Third World: Do We Really Want IPM to Work?" Bull. Entomol. Soc. of Amer., 30
Heisey, P. et al. "Farmers' Knowledge and Perceptions of Improved Wheat Technology."
Draft paper, PARC/CIMMYT Collaborative Research Program, Islamabad, Pakistan,
Heinemann, E. and S.D. Biggs. "Farming Systems Research: An Evolutionary Approach to
Implementation." 3. Agric. Econ. 36(1985): 59-65.
Herdt, R. W. and C. Capule. Adoption, Spread and Production Impact of Modern Rice
Varieties in Asia. Los Banos, Philippines: IRRI, 1983.
Herdt, R. W. and A. M. Mandac. "Modern Technology and Economic Efficiency of
Philippine Rice Farmers," Econ. Dev. Cult. Change, 29 (1981): 375-400.
Heyneman, S. P. "Improving the Quality of Education in Developing Countries," Finance
and Devel. 20 (1983): 18-21.
Hoffnar, B.R. and G.L. 3ohnson, Summary and Evaluation of the Cooperative Agronomic-
Economic Experimentation at Michigan State University, Exp. Stat. Bul. 11,
Michigan State University, East Lansing, 1966.
Holdcroft, L.E. "The Rise and Fall of Community Development, 1950-65: A Critical
Assessment," In Agricultural Development in the Third World, C.K. Eicher and 3.M.
Staatz (eds), Baltimore; John Hopkins University Press, 1984.
Howell, 3. Managing Agricultural Extension: The T and V System in Practice.
Discussion Paper No. 8, Agricultural Administration Unit, Overseas Development
Institute, London, 1982.
Huang, C. 3. and F. S. Bagi. "Technical Efficiency of Individual Farmers in Northwest
India," Southern Econ. 3., 5 (1984): 108- 15.
Huffman, W. E. "Allocative Efficiency: The Role of Human Capital," Quart. 3. Econ., 91
Huffman, W. E. "Assessing Returns to Agricultural Extension," Amer. 3. Agric. Econ. 60
Huke, R.E., V. Cordova and S. Sardido. San Bartolome: Beyond the Green Revolution.
Research Rep. No. 74, IRRI, Los Banos, 1982.
Hussain, S.S. et al, Wheat in the Irrigated Farming Systems of Mardan District:
Implications for Research and Extension, Research Report, Pakistan Agricultural
Research Council, Islamabad, 1985.
Hussain, T. Wheat in the High Mountain Valleys of Gilgit. Report No. 2, Aga Khan Rural
Support Prog., Gilgit (Pak.), 1986.
Jamison, D. T. and L. 3. Lau. Farmer Education and Farmer Efficiency. Baltimore: John
Hopkins Univ. Press, 1982.
3amison, D. T. and P. R. Moock. "Farmer Education and Farm Efficiency in Nepal: The
Role of Schooling, Extension Services and Cognitive Skills," World Dev. 12 (1984):
Johnson, A. W. "Individuality and Experimentation in Traditional Agriculture," Human
Ecol. 1 (1972): 149-59.
Jones, W. 0. "Turnips, The Seventh Day Adventist Principle, and Management Bias."
Food Research Inst. Studies 16 (1977-78): 141-157.
3unankar, P. N. "Do Indian Farmers Maximize Profits?" 3. Dev. Stud. 17 (1980a): 48-61.
3unankar, P. N. "Tests of the Profit Maximizing Hypothesis: A Study of Indian
Agriculture," 3. Dev. Stud. 16 (1980b): 186- 203.
Kahlon, A.S. Modernization of Punjab Agriculture, New Delhi; Allied Publishers, 1984.
Kalirajan, K. "An Econometric Analysis of Yield Variability in Paddy Production." Cdn.
3. Agric. Ec. 29 (1981): 283-294.
Kalirajan, K. "Farm Specific Technical Efficiencies and Developmental Policies." 3.
Econ. Stud. 11 (1984): 3-13.
Kalirajan, K. P. and 3. C. Flinn. "The Measurement of Farm Specific Technical
Efficiency," Pakistan 3. Applied Econ. 11 (1983): 167-80.
Kalirajan, K. P. and R. T. Shand. "Types of Education and Agricultural Productivity: A
Quantitative Analysis of Tamil Nadu Rice Farming." 3. Dev. Stud. 21 (1985): 232-43.
Kenmore, P. E. "Crop Loss Assessment in a Practical Integrated Pest Control Program
for Tropical Asian Rice," In P. S. Teng (ed), Crop Loss Assessment and Pest
Management, St. Paul, Min; American Phytopath. Society, 1986.
Khan, M.3., M. Sharif and M. Sarwar, Monitoring and Evaluation of Training and Visit
System of Agricultural Extension in Punjab, Pakistan, Publication No. 212, Punjab
Economic Research Institute, Lahore, Pakistan, 1984.
Khan, M.3. and R.A. Young. "Farm Resource Productivities, Allocative Efficiencies and
Development Policy in the Indus Basin, Pakistan." Land Econ. 55(1979): 388-96.
Leaf, M. Song of Hope: The Green Revolution in a Punjab Village. New Brunswick:
Rutgers Univ. Press, 1984.
Lionberger, H. F. and H. C. Chang. Farm Information for Modernizing Agriculture: The
Taiwan Experience. New York: Praeger, 1970.
Lingard, 3., L. Castillo and S. 3ayasuriya. "Comparative Efficiency of Rice Farms in
Central Luzon, The Philippines." 3. Agric. Econ. 34(1983): 163-173.
Lipton, M. and R. Longhurst. Modern Varieties, International Agricultural Research, and
the Poor. CGIAR Study Paper No. 2, Washington, D.C.: World Bank, 1985.
Lipton, M. "Theory of the Optimizing Peasant." 3. Dev. Stud. 4(1968): 325-51.
Litsinger, 3. A., E. C. Price and R. T. Herrera. "Small Farmer Pest Control Practice for
Rainfed Rice, Corn and Grain Legumes in Three Philippine Provinces." Phillip.
Entom. 4 (1980): 65-86.
Lockheed, M. E., D. 3. Jamison and L. 3. Lau. "Farmer Education and Farm Efficiency:
A Survey." Econ. Dev. Cult. Change 29 (1980): 37-76.
Lowdermilk, M.K. "Diffusion of Dwarf Wheat Production Technology in Pakistan's
Punjab". Ph.D. Thesis, Cornell University, Ithaca, 1972.
Matheson, P. C., M. A. Altieri and W. C. Gagni. "Modification of Small Farmer Practices
for Better Pest Management," Ann. Rev. Entomol. 29 (1984): 383-402.
Martinez, 3.C. and G. Sain, The Economic Returns to Institutional Innovations in National
Agricultural Research: On-farm Research in IDIAP, Panama. Economics Working
Paper 04/83, CIMMYT, Mexico, 1983.
Mayani, V. V. and K. Kumar. "Existing Knowledge Based Training Needs of Small
Farmers." Gujrat Agric. Res. 3. 5 (1980): 25- 30.
Mijindadi, N.B. and D.W. Norman. "Efficiency Differentials in the Traditional
Agriculture in Northern Nigeria." Agric. Sys 14(1984): 213-218.
Mohammad, A. Approaches to the Transfer of Crop Production Technology in Developing
Countries, AGPC/MISC/81, Rome: Food and Agricultural Organization, 1984.
Moock, P. R. "Education and Technical Efficiency in Small Farm Production," Econ.
Dev. Cult. Change, 29 (1981): 723-740.
Moore, M. "Institutional Development: The World Bank and India's New Agricultural
Extension Program," 3. Devel. Stud. 20 (1984): 303-317.
Nagy, 3.G. "The Pakistan Agricultural Development Model: An Economic Evaluation of
Agricultural Research and Extension Expenditures." Ph.D. Thesis, University of
Minnesota, St. Paul, 1984.
Nair, K. In Defence of the Irrational Peasant: Indian Agriculture after the Green
Revolution. Chicago; Univ. of Chicago Press, 1979.
Pasour, E. C. 3r. "A Further Note on the Measurement of Efficiency and Economies of
Farm Size." 3. Agric. Econ. 32(1981): 135-146.
Peng, T. K. and Chen, M. V. "The Measurement of Technical Efficiency of Maize
Production in Taiwan," 3. Agric. Econ. (Taiwan) 38 (1985): 1-23.
Perraton, H. et al. Basic Education and Agricultural Extension; Costs, Effects and
Alternatives. Staff Working Paper No. 564, Washington, D. C.: World Bank, 1983.
Phillips, 3. M. "A Comment on Farmer Education and Farm Efficiency: A Survey."
Econ. Dev. Cult. Change, 35(1987): 637-44.
Pingali, P. L., and G. A. Carlson. "Human Capital, Adjustments in Subjective Probability
and the Demand for Pest Controls," Amer. 3. Agric. Econ. 67 (1985): 854-61.
Plucknett, D. L. and N. 3. H. Smith. "Sustaining Agricultural Yields," BioScience 36
Pontius, S. K. "The Communication Process of Adoption: Agriculture in Thailand," 3.
Dev. Areas. 18 (1983): 93-118.
Psacharopolus, G. and M. Woodall. Education for Development: An Analysis of
Investment Choices. New York: Oxford Univ. Press, 1985.
Pudasaini, S. P. "The Effects of Education in Agriculture: Evidence from Nepal," Amer.
3. Agric. Econ. 65 (1983): 509- 15.
Ram, R. "Role of Education in Production: A Slightly New Approach," Quart. 3. Econ. 95
Rogers, E.M. Diffusion of Innovations, Glencoe, New York: Free Press, 1983.
Rogers, E. M., and W. L. Meynen. "Communication Sources for 2- 4,D Weed Spray
Among Columbian Peasants," Rur. Sociol. 30 (1975): 213-19.
Roling, N. G., 3. Ascroft and F. W. Chege. "The Diffusion of Innovations and the Issue of
Equity in Rural Development." In Extension Education and Rural Development, B.R.
Crouch and S. Chamala (eds.), New York: John Wiley & Sons, 1981.
Roumasset, J.A., M.W. Rosegrant, U.N. Chakravorty and 3.R. Anderson. "Fertilizer and
Crop Yield Variability: A Review". Draft paper, University of Hawaii, Honolulu,
Russell, N. P. and T. Young. "Frontier Production Functions and the Measurement of
Technical Efficiency," 3. Agric. Econ. 34(1983): 139-149.
Ruttan, V. W. Agricultural Research Policy. Minneapolis: Univ. of Minnesota Press,
Ruttan, V. W. "Induced Institutional Innovation." In Induced Innovation: Technology,
Institutions and Development. (H. P. Binswanger, V. W. Ruttan et al., eds.),
Baltimore: John Hopkins Univ. Press, 1978.
Ruttan, V. W. and H. P. Binswanger. "Induced Innovation and the Green Revolution." In
Induced Innovation: Technology, Institutions and Development. (Edited by H. P.
Binswanger, V. W. Ruttan et al.), eds., Baltimore: John Hopkins Univ. Press, 1978.
Scandizzo, P. L. Agricultural Growth and Factor Productivity in Developing Countries.
Econ. and Social Dev. Paper 42, Rome: Food and Agricultural Organization, 1984.
Schultz, T. W. Transforming Traditional Agriculture. Chicago: Univ. of Chicago Press,
Schultz, T. W. "The Value of the Ability to Deal with Disequilibria," 3. Econ. Lit. 13
Scobie, G. M. and D. L. Franklin. "The Impact of Supervised Credit Programmes on
Technological Change in Developing Agriculture." Austral. 3. Agric. Econ. 21 (1977):
Shapiro, K.H. "Efficiency Differentials in Peasant Agriculture and their Implications for
Development Policies." 3. Dev. Studies. 19(1983): 179-90.
Shingi, P. M., S. Wadwalker and G. Kaur. Management of Agricultural Extension:
Training and Visit System in Rajasthan. Ahmedabad: Indian Inst. of Manag., 1982.
Sims, H. "The State and Rural Development: Agricultural Policy and Performance in the
Indian and Pakistan Punjabs", Unpublished Ph.D. Dissertation, University of
California, Berkley, 1985.
Simmonds, N. W. "A Short Review of Farming Systems Research in the Tropics." Expl.
Agric. 22 (1986): 1-13.
Sonka, S. T. "Information Management in Farm Production," Computers and Electronics
in Agriculture, 1 (1985): 75-85.
Srivastava, S. K. "Plant Protection in India: Problems and Prospects," PANS 22 (1976):
Staatz, 3. M. and C. K. Eicher. "Agricultural Development in Historical Perspective." In
Agricultural Development in the Third World. C. K. Eicher and 3. M. Staatz (eds),
Baltimore: John Hopkins Univ. Press, 1984.
Stavis, B.N. "Agricultural Extension for Small Farmers", Rural Dev. Working Paper No. 3,
Michigan State University, East Lansing, 1979.
Stigler, G.3. "The X-Istence of X-Efficiency", Amer. Econ. Rev. 66(1976): 213-216.
Stoneman, P. L. and P. A. David. "Adoption Subsidies vs Information Provision as
Instruments of Technology Policy," Econ. 3. 96 (1986): 142-51.
Taylor, T. G. and 3. S. Shonikwiler. "Alternative Stochastic Specifications of the
Frontier Production Function in the Analyses of Agricultural Credit Programs and
Technological Efficiency," 3. Dev. Econ. 21 (1986): 149-60.
Todaro, M.P. Economic Development in the Third World, New York: Longman, 1985.
Turpin F. T. and 3. D. Maxwell. "Decision Making Related to Use of Soil Insecticides by
Indiana Corn Farmers," 3. Econ. Entomol. 69 (1976): 359-62.
Upton, M. "The Unproductive Production Function," 3. Agric. Econ. 30 (1979): 179-191.
Vyas, V.S. "Asian Agriculture; Achievements and Challenges". Asian Dev. Rev. 1(1983):
Welch, F. "The Role of Investment in Human Capital in Agriculture." In T. W. Schultz
(ed), Distortions of Agricultural Incentives. Bloomington: Indiana Univ. Press, 1978.
Winkelmann, D.L., "Diversification, Sustainability and Economics". In Sustainability
Issues in Agricultural Development, Proc. of the Seventh Agricultural Sector
Symposium, Washington, D.C.: World Bank, 1987.
Winkelmann, D.L. "Toward a Sustainable Agriculture." CIMMYT, Mexico (forthcoming).
World Bank. Agricultural Research and Extension: An Evaluation of World Bank
Experience. Washington, D. C.: World Bank, 1985.
Wu, C.C. "Education and Farm Production; The Case of Taiwan", Amer. 3. Agric. Econ.
Yotopoulous, P. A. and L. 3. Lau (eds). "Resource Use in Agriculture: Applications of the
Profit Function to Selected Countries," Food Research Institute Studies 17 (1979): 1-
Zadocks, 3. C. "On the Conceptual Basis of Crop Loss Assessment: The Threshold
Theory," Ann. Rev. Phytopathol. 23 (1985): 455- 73.
MSU INTERNATIONAL DEVELOPMENT PAPERS Price
IDP No. 1 Carl K. Eicher and Doyle C. Baker, "Research on Agricultural Development in Sub- $ 8.00
Saharan Africa: A Critical Survey," 1982 (346 pp.).
IDP No. IF Carl K. Eicher et Doyle C. Baker, "Etude critique de la recherche sur le $10.00
developpement agricole en Afrique subsaharienne," 1985, (435 pp.).
IDP No. 2 Eric W. Crawford, "A Simulation Study of Constraints on Traditional Farming Systems $ 5.00
in Northern Nigeria," 1982 (136 pp.).
IDP No. 3 M.P. Collinson, "Farming Systems Research in Eastern Africa: The Experience of $ 4.00
CIMMYT and Some National Agricultural Research Services, 1976-81," 1982 (67 pp.).
IDP No. 4 Vincent Barrett, Gregory Lassiter, David Wilcock, Doyle Baker, and Eric Crawford, $ 5.00
"Animal Traction in Eastern Upper Volta: A Technical, Economic and Institutional
Analysis," 1982 (132 pp.).
IDP No. 5 John Strauss, "Socio-Economic Determinants of Food Consumption and Production in Out of Print
Rural Sierra Leone: Application of an Agricultural Household Model with Several
Commodities," 1983 (91 pp.).
IDP No. 6 Beverly Fleisher and Lindon 3. Robison, "Applications of Decision Theory and the $ 5.00
Measurement of Attitudes Towards Risk in Farm Management Research in Indus-
trialized and Third World Settings," 1985 (106 pp.).
IDP No. 7 C. Peter Timmer, "Private Decisions and Public Policy: The Price Dilemma in Food $ 5.00
Systems of Developing Countries," 1986 (58 pp.).
IDP No. 8 Michael L. Morris, "Rice Marketing in the Senegal River Valley: Research Findings $ 5.00
and Policy Reform Options," 1987 (89 pp.).
IDP No. 9 Carl Liedholm and Donald Mead, "Small Scale Industries in Developing Countries: $ 6.00
Empirical Evidence and Policy Implications," 1987 (141 pp.).
IDP No. 10 Derek Byerlee, "Maintaining the Momentum in Post-Green Revolution Agriculture: A $ 5.00
Micro-Level Perspective from Asia," 1987 (57 pp.).
MSU INTERNATIONAL DEVELOPMENT WORKING PAPERS
WP No. 1 Daniel Gait, Alvaro Diaz, Mario Contreras, Frank Peairs, Joshua Posner and Franklin Out of Print
Rosales, "Farming Systems Research (FSR) in Honduras, 1977-81: A Case Study," 1982
WP No. 2 Edouard K. Tapsoba, "Credit Agricole et Credit Informel dans le Region Orientale Out of Print
de Haute-Volta: Analyse Economique, Performance Institutionnelle et Implications
en Matiere de Politique de Developpement Agricole," 1982 (125 pp.).
WP No. 3 W.P. Strassmann, "Employment and Construction: Multicountry Estimates of Costs and Out of Print
Substitution Elasticities for Small Dwellings," 1982 (48 pp.).
WP No. 4 Donald C. Mead, "Sub-contracting in Rural Areas of Thailand," 1982 (52 pp.). Out of Print
WP No. 5 Michael T. Weber, 3ames Pease, Warren Vincent, Eric W. Crawford and Thomas Stilwell, $ 5.00
"Microcomputers and Programmable Calculators for Agricultural Research in Developing
Countries," 1983 (113 pp.).
WP No. 6 Thomas Stilwell, "Periodicals for Microcomputers: An Annotated Bibliography," 1983 See IDWP #21
WP No. 7 W. Paul Strassmann, "Employment and Housing in Lima, Peru," 1983 (96 pp.). Out of Print
WP No. 8 Carl K. Eicher, "Faire Face a la Crise Alimentaire de l'Afrique," 1983 (29 pp.). Free
WP No. 9 Thomas C. Stilwell, "Software Directories for Microcomputers: An Annotated See IDWP #22
Bibliography," 1983 (14 pp.).
WP No. 10 Ralph E. Hepp, "Instructional Aids for Teaching How to Use the TI-59 Programmable Out of Print
Calculator," 1983 (133 pp.).
MSU INTERNATIONAL DEVELOPMENT WORKING PAPERS CONTINUED
WP No. 11 Michael L. Morris and Michael T. Weber, "Programmable Calculator (TI-59) Programs Out of Print
for Marketing and Price Analysis in Third World Countries," 1983 (105 pp.).
WP No. 12 Valerie Kelly, Robert D. Stevens, Thomas Stilwell, and Michael T. Weber, "An Annotated $ 7.00
Directory of Statistical and Related Microcomputer Software for Socioeconomic Data
Analysis," 1983 (165 pp.).
WP No. 13 Chris Wolf, "Guidelines for Selection of Microcomputer Hardware," 1983 (90 pp.). $ 5.00
WP No. 14 Eric W. Crawford, Ting-Ing Ho, and A. Allan Schmid, "User's Guide to BENCOS-- $ 3.00
SuperCalc Template for Benefit-Cost Analysis," 1984 (35 pp.).
Copy of BENCOS Template in IBM PC-DOS 1.1 Format, on single sided double density $15.00
diskette (readable on most MS-DOS systems).
WP No. 15 James W. Pease and Raoul Lepage with Valerie Kelly, Rita Laker-Ojok, Brian Thelen, $ 7.00
and Paul Wolberg, "An Evaluation of Selected Microcomputer Statistical Programs,"
1984 (187 pp.).
WP No. 16 Stephen Davies, James Seale, Donald C. Mead, Mahmoud Badr, Nadia El Sheikh, Out of Print
and Abdel Rahman Saidi, "Small Enterprises in Egypt: A Study of Two Governorates,"
1984 (100 pp.).
WP No. 17 Thomas C. Stilwell, "Microcomputer Statistical Packages for Agricultural Research," $ 3.00
1984 (23 pp.).
WP No. 18 Thomas C. Stilwell and P. Jordan Smith, "An Annotated Directory of Citation Database, $ 3.00
Educational, System Diagnostics and Other Miscellaneous Microcomputer Software of
Potential Use to Agricultural Scientists in Developing Countries," 1984 (34 pp.).
WP No. 19 Amalia Rinaldi, "Irrigation in Southern Africa: An Annotated Bibliography," 1985 $ 4.00
WP No. 20 Daniel C. Goodman, 3r., Thomas C. Stilwell, and P. Jordan Smith, "A Microcomputer $ 5.00
Based Planning and Budgeting System for Agricultural Research Programs," 1985
WP No. 21 Thomas C. Stilwell, "Periodicals for Microcomputers: An Annotated Bibliography," $ 5.00
Second Edition, 1985 (89 pp.).
WP No. 22 Thomas C. Stilwell, "Software Directories for Microcomputers: An Annotated $ 3.00
Bibliography," Second Edition, 1985 (21 pp.).
WP No. 23 Alan Hrapsky with Michael Weber and Harold Riley, "A Diagnostic Prescriptive $ 5.00
Assessment of the Production and Marketing System for Mangoes in the Eastern
Caribbean," 1985 (106 pp.).
WP No. 24 Donald C. Mead, "Subcontracting Systems and Assistance Programs: Opportunities Out of Print
for Intervention," 1985 (32 pp.).
WP No. 25 Carl Liedholm, "Small Scale Enterprise Credit Schemes: Administrative Costs and Out of Print
the Role of Inventory Norms," 1985 (23 pp.).
WP No. 26 James 3. Boomgard, Stephen P. Davies, Steve Haggblade, and Donald C. Mead, Out of Print
"Subsector Analysis: Its Nature, Conduct and Potential Contribution to Small
Enterprise Development," 1986 (57 pp.).
WP No. 27 Steve Haggblade, Carl Liedholm, and Donald C. Mead, "The Effect of Policy and Policy Out of Print
Reforms on Non-Agricultural Enterprises and Employment in Developing Countries: A
Review of Past Experiences," 1986 (133 pp.).
WP No. 28 John T. Milimo and Yacob Fisseha, "Rural Small Scale Enterprises in Zambia: Results Out of Print
of a 1985 Country-Wide Survey," 1986 (76 pp.).
WP No. 29 Stephan Goetz and Michael T. Weber, "Fundamentals of Price Analysis in Developing $ 7.00
Countries' Food Systems: A Training Manual to Accompany the Microcomputer Software
Program 'MSTAT,'" 1986 (148 pp.).
WP No. 30 John S. Holtzman, "Rapid Reconnaissance Guidelines for Agricultural Marketing and Food $ 5.00
System Research in Developing Countries," 1986 (75 pp.).
MSU INTERNATIONAL DEVELOPMENT WORKING PAPERS CONTINUED
WP No. 31 Nicholas William Minot, "Contract Farming and Its Effect on Small Farmers in $ 5.00
Less Developed Countries," 1986 (86 pp.).
MSU INTERNATIONAL DEVELOPMENT REPRINT PAPERS
RP No. 1 Carl Liedholm, "The Private Sector Connection to Development," 1986 (19 pp.). $ 3.00
RP No. 2 James D. Shaffer with Michael Weber, Harold Riley and John Staatz, "Influencing $ 3.00
the Design of Marketing Systems to Promote Development in Third World Countries
RP No. 3 Carl K. Eicher, "Famine Prevention in Africa: The Long View," 1987 (18 pp.). $ 3.00
RP No. 4 Michael L. Morris, "Cereals Marketing in the Senegal River Valley (1985)," $ 6.00
1987 (126 pp.).
RP No. 5 Mandivamba Rukuni and Carl K. Eicher, "The Food Security Equation in Southern $ 3.00
Africa," 1987 (32 pp.).
RP No. 6F Eric Crawford et Mulumba Kamuanga, "L'Analyse Economique des Essais Agronomiques $ 3.00
Pour la Formulation des Recommandations aux Paysans," 1987 (33 pp.).
RP No. 7F Eric Crawford, "L'Analyse Economique des Essais Zootechniques," 1987 (36 pp.). $ 3.00
RP No. 8 Eric Crawford and Valerie Kelly, "A Field Study of Fertilizer Distribution and Use $ 3.00
in Senegal, 1984: Summary Report," 1987 (32 pp.).
RP No. 9 Kelly Harrison, Donald Henley, Harold Riley and James Shaffer, "Improving Food $ 5.00
Marketing Systems in Developing Countries: Experiences from Latin America,"
1987 (135 pp.).
RP No. 10 Mark Newman, Eric Crawford and Jacques Faye, "Policy Relevant Research on the $ 3.00
Food and Agricultural System in Senegal," 1987 (30 pp.).
RP No. 10F Mark Newman, Eric Crawford et Jacques Faye, "Orientations et Programmes de $ 3.00
Research Macro-Economiques sur le Systeme Agro-Alimentaire Senegalais," 1987
RP No. 11 Eric Crawford, Curtis Jolly, Valerie Kelly, Philippe Lambrecht, Makhona Mbaye, $ 6.00
and Matar Gaye, "A Field Study of Fertilizer Distribution and Use in Senegal,
1984: Final Report," 1987 (111 pp.).
RP No. 11IF Eric Crawford, Curtis Jolly, Valerie Kelly, Philippe Lambrecht, Makhona Mbaye, $ 6.00
et Matar Gaye, "Enquete sur la Distribution et l'Utilisation de I'Engrais au
Senegal, 1984: Rapport Final," 1987 (106 pp.).
RP No. 12 Mark D. Newman, P. Alassane Sow and Ousseynou NDoye, "Private and Public $ 3.00
Sectors in Developing Country Grain Markets: Organization Issues and Options
in Senegal," 1987 (14 pp.).
RP No. 13F R. 3ames Bingen et Jacques Faye, "La Liaison Recherche-Developpement en Afrique $ 3.00
de l'Ouest Francophone: L'Experience du Senegal," 1987 (32 pp.).
RP No 14 Mark D. Newman, "Grain Marketing in Senegal's Peanut Basin: 1984/85 Situation $ 3.00
and Issues," 1987 (16 pp.).
Copies may be obtained from: MSU International Development Papers, Department of Agricultural Economics, 7
Agriculture Hall, Michigan State University, East Lansing, Michigan 48824-1039, U.S.A. All orders must be prepaid
in United States currency. Please do not send cash. Make checks or money orders payable to Michigan State
University. There is a 10% discount on all orders of 10 or more sale copies. Individuals and institutions in the Third
World and USAID officials may receive single copies free of charge.