Front Cover
 Title Page
 Panel on appropriate technologies...
 Board on science and technology...
 Table of Contents
 1. Summary
 2. Interaction of technological...
 3. Alternative criteria of appropriateness...
 4. Technological opportunities...
 5. Determinants of technological...
 6. Special features of technological...
 7. Special sectoral problems and...
 8. Policies for promoting choice...
 Back Cover

Title: Appropriate technologies for developing countries
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00082048/00001
 Material Information
Title: Appropriate technologies for developing countries
Physical Description: xiii, 140 p. : ; 23 cm.
Language: English
Creator: Eckaus, Richard S., 1926-
National Research Council (U.S.) -- Panel on Appropriate Technologies for Developing Countries
Publisher: National Academy of Sciences
Place of Publication: Washington
Publication Date: 1977
Subject: Appropriate technology -- Developing countries   ( lcsh )
Technology transfer -- Developing countries   ( lcsh )
Transferencia de tecnología
Transfert de technologie   ( rvm )
Technologie -- Pays en voie de développement   ( rvm )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Bibliography: p. 127-137.
Statement of Responsibility: prepared by Richard S. Eckaus for the Panel on Appropriate Technologies for Developing Countries, Board on Science and Technology for International Development, Commission on International Relations, National Research Council.
General Note: "Prepared ... under contract no. AID/csd-2584, task order no. 1, with the Office of Science and Technology, Bureau for Technical Assistance, Agency for International Development."
 Record Information
Bibliographic ID: UF00082048
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 02936764
lccn - 76053284
isbn - 0309026024

Table of Contents
    Front Cover
        Front Cover
    Title Page
        Page i
        Page ii
    Panel on appropriate technologies for developing countries
        Page iii
        Page iv
    Board on science and technology for international development
        Page v
        Page vi
        Page vii
        Page viii
        Page ix
        Page x
    Table of Contents
        Page xi
        Page xii
        Page xiii
        Page xiv
        Page 1
        Page 2
        Page 3
        Page 4
    1. Summary
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
    2. Interaction of technological decisions and development processes
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
    3. Alternative criteria of appropriateness of technological decisions
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
    4. Technological opportunities and transfer of technical information
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
    5. Determinants of technological decisions and their appropriateness
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
    6. Special features of technological decisions in agriculture
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
    7. Special sectoral problems and opportunities
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
    8. Policies for promoting choice of appropriate technologies
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
    Back Cover
        Back Cover
Full Text


Page 141, line 14, under the heading "Advisory
Studies and Special Reports," the cost of
Appropriate Technologies for Developing Coun-
tries should be $6.25.

Page 142, under #21. Making Aquatic Weeds Useful,
add: NTIS Accession No. PB 265-161 $7.50.

Appropriate Technologies for
Developing Countries
ISBN 0-309-2602-4

Technologies for

Prepared by
Massachusetts Institute of Technology
for the
Board on Science and Technology
for International Development
Commission on International Relations
National Research Council

Washington, D.C. 1977

NOTICE: The project that is the subject of this report was approved by the Governing
Board of the National Research Council, whose members are drawn from the Councils
of the National Academy of Sciences, the National Academy of Engineering, and the
Institute of Medicine. The members of the Committee responsible for the report were
chosen for their special competence and with regard for appropriate balance.
This report has been reviewed by a group other than the authors according to
procedures approved by a Report Review Committee consisting of members of the
National Academy of Sciences, the National Academy of Engineering, and the In-
stitute of Medicine.

This report has been prepared by Richard S. Eckaus, Professor of Economics, Mas-
sachusetts Institute of Technology, Cambridge, Massachusetts, for a special panel of the
Board on Science and Technology for International Development in cooperation with the
Office of the Foreign Secretary, National Academy of Engineering, Commission on
International Relations, National Research Council, under Contract No. AID/csd-2584,
Task Order No. 1, with the Office of Science and Technology, Bureau for Technical
Assistance, Agency for International Development, Washington, D.C.

Library of Congress Cataloging in Publication Data

Eckaus, Richard S 1926-
Appropriate technologies for developing countries.

"Prepared . under contract no. AID/csd-2584, task order no. 1, with the Of-
fice of Science and Technology, Bureau for Technical Assistance, Agency for
International Development."
Bibliography: p.
1. Underdeveloped areas-Technology. 2. Technology transfer. I. National Re-
search Council. Panel on Appropriate Technologies for Developing Countries.
II. Title.
HC59.7.E25 338.91'172'4 76-53284
ISBN 0-309-02602-4

Printed in the United States of America


BRUCE S. OLD, Vice President, Arthur D. Little, Inc., Cambridge,
Massachusetts, Chairman
JAMES P. GRANT, Director, Overseas Development Council,
Washington, D.C.
CHRISTIAN KRISTOFF, Manager, Corporate Product Planning
Group, General Motors Corporation, Detroit, Michigan
FREDERICK T. MOORE, Economic Advisor, Industrial Projects De-
partment, World Bank, Washington, D.C.
FRANCISCO SAGASTI, Science and Technology Policy Instruments
(STPI) Project, Lima, Peru
JOSEPH STEPANEK, Consultant, Checchi and Company,
Washington, D.C.
SIMON TEITEL, Consultant, Office of the Program Advisor, Inter-
American Development Bank, Washington, D.C.

Consultant and Report Author
RICHARD S. ECKAUS, Professor, Department of Economics, Mas-
sachusetts Institute of Technology, Cambridge, Massachusetts

JAY J. DAVENPORT, Board on Science and Technology for Interna-
tional Development, Commission on International Relations, Na-
tional Research Council, Professional Associate
HUGH H. MILLER, Office of the Foreign Secretary, National
Academy of Engineering, Executive Secretary
JULIEN ENGEL, Board on Science and Technology for International
Development, Commission on International Relations, National Re-
search Council, Head, Special Studies


DAVID PIMENTEL, Professor, Department of Entomology and Sec-
tion of Ecology and Systematics, Cornell University, Ithaca, New
York, Chairman
CARL DJERASSI, Department of Chemistry, Stanford University,
Stanford, California, Chairman, 1973-1975


GEORGE S. HAMMOND, Foreign Secretary, National Academy of
RUTH ADAMS, American Academy of Arts and Sciences, Boston,
EDWARD S. AYENSU, Director, Department of Botany, Smithsonian
Institution, Washington, D.C.
DWIGHT S. BROTHERS, International Agricultural Development
Service, The Rockefeller Foundation, New York, New York
GEORGE BUGLIARELLO, President, Polytechnic Institute of New
York, Brooklyn, New York
CHARLES DEN N I SON, Former Vice President, International Minerals
and Chemicals Corporation, New York, New York
WILLIAM A. W. KREBS, Vice President, Arthur D. Little, Inc.,
Cambridge, Massachusetts
FRANKLIN A. LONG, Program on Science, Technology, and Society,
Cornell University, Ithaca, New York
FREDERICK T. MOORE, Economic Advisor, International Bank for
Reconstruction and Development, Washington, D.C.
JOSEPH PETTIT, President, Georgia Institute of Technology, Atlanta,
JOSEPH B. PLATT, President, Claremont University Center, Claremont,
HELEN RANNEY, Chairman, Department of Medicine, University
Hospital, San Diego, California
ROBERT M. WALKER, McDonnell Professor of Physics, Director,
Laboratory of Space Physics, Washington University, St. Louis,


MICHAEL G. C. MCDONALD DOW, Deputy Director, Head, Over-
seas Programs
JULIEN ENGEL, Deputy Director, Head, Special Studies


Since the 1960's, science and technology have been increasingly rec-
ognized as powerful influences in economic growth processes. This is
nowhere more evident than in the developing countries, which have in
some sectors gained enormously from the use of the technological
knowledge accumulated in the industrialized countries. Agricultural
and industrial production, the discovery and exploitation of natural
resources, and transportation, communications, health care, educa-
tion, and many other fields have benefited from technical innovations.
Yet the transfer of technology to the developing countries has been
accompanied by new economic and social problems. While some of
these may be inherent in the processes of modernization, others appear
to be specifically associated with the characteristics of the technologies
transferred from industrialized countries.
By the late 1960's, there was growing concern about apparent
incongruities between the goals of the developing countries, their labor
supply conditions and other resource endowments, and the tech-
nologies these countries were importing. Advanced technologies were
characterized by the use of less labor per unit of output, meaning
that they not only failed to expand employment as fast as output but, in
some instances, may actually have displaced labor and contributed to
already substantial unemployment. The new technologies also often
seemed to require large plants and equipment and thus appeared to be
relatively capital intensive, a strain in countries where capital is unusu-
ally scarce. Further, the imported technologies appeared to require the

viii Preface
creation of large establishments and to demand new and higher levels
of labor skills, to make new social demands on the countries, and to
change their economic and social structures. Foreign exchange seemed
to become increasingly absorbed in the payment of royalties for
licenses for foreign technologies, thereby fostering dependency rela-
The effects we are describing are said by some development
specialists to lack verification; others would differ with this view.
There has, of course, been a growing preoccupation within developing
countries themselves with the adaptation of technologies to their
particular economic and political objectives and conditions. This con-
cern is also manifested by national and international lending and
assistance agencies and among many development specialists. Atten-
tion has turned to the possibility of finding new technologies more
suitable than those now available. However, the difficulties in specify-
ing precisely the characteristics that would make new technologies
more suitable are apparent in the variety of names they have been
given in anticipation of their discovery. These names include: "labor
intensive," "progressive," "capital saving," "village level," and "in-
termediate." The term "appropriate" has come to be most widely
used, perhaps in recognition that there are many conditions determin-
ing the degree of suitability of a particular technology to any environ-
ment: the political and economic objectives of each country, its social
structure and functioning, and the availability and quality of its produc-
tive resources.
In the early 1970's, the National Academy of Sciences (NAS)-
National Academy of Engineering (NAE) through the Board on Science
and Technology for International Development (BOSTID) joined the
discussion of scholars and practitioners who were caught up in the task
of analyzing criteria for technological choices for development. The
U.S. Agency for International Development (AID) requested BOSTID to
study the issues inherent in the concept of "appropriate" technologies.
In accordance with its usual practice, BOSTID convened a panel
broadly representative of current knowledge and practice on problems
of technological choice, utilization, adaptation, and transfer to guide
the study and produce a report. This was preceded by meetings of
several ad hoc groups brought together to chart a course of action.
These groups were composed of economists, engineers, scientists,
development administrators, and others experienced in the problems of
technology transfer, utilization, and adaptation. To limit an inquiry that
threatened to become as broad as the development process itself, the
panel confined its attention to the industrial sector (with a brief look at

Preface ix
agriculture) and to making recommendations bearing on the long-term
industrial development of the less-developed countries.
The characterization of technology strategies measured against
criteria of appropriateness proved to be more formidable than was
initially realized. In the end, BOSTID turned to Richard S. Eckaus,
Professor of Economics at Massachusetts Institute of Technology, for
the demanding task of writing a report that would analyze the interrela-
tionships between technological choices and economic, social, and
political aspects of the development process. This report, which has as
its primary aim to help decision makers and others become more aware
of the complexities and constraints inherent in technological choice,
was written by Professor Eckaus on the basis of his own extended
study and several rounds of discussion and review with the panel and
its members. The report is not meant to be a scholarly treatise or a
comprehensive review of the literature. It is an exposition of the
considered views of Professor Eckaus discussing the interaction among
the many factors essential for the rational selection of a technology.
The views finally expressed in the document are those of Professor
Eckaus, with broad concurrence by members of the panel. It should be
noted that one member of the panel (see Appendix) and some experts
in the larger interested community do disagree, both on some details
and on some principles. We feel the report will make a useful contribu-
tion if it stimulates further discussion and thereby helps bring about the
recognition that technological choices are always made within a politi-
cal and social context and that, whatever a society's policies or goals,
all criteria for making choices bear a price. When costs and benefits are
better understood, more rational technological choices are likely to
As Chairman of the Panel on Appropriate Technologies, I take
pleasure in thanking each of the panel members for his patience and
cooperation in this endeavor over many months of effort. Special
thanks and recognition are owed Professor Eckaus, not only for the
scholarship he brought to the preparation of the panel's report but also
for his forbearance and dedication.
On behalf of the panel I also wish to thank those who prepared
background materials for the report. Although they are not reproduced
here, the informal case studies and the panel's discussions with their
authors were of great help in clarifying the issues. Contributors to this
part of the work included Jose Giral, of the National Autonomous
University of Mexico, on appropriate chemical technologies; W. Paul
Strassmann and John S. McConnaughey, of Michigan State Univer-
sity, on housing and residential construction; Gerard K. Boon, of the

x Preface

College of Mexico, on economic technological behavior in the metal-
working industries; Charles Kusik, of Arthur D. Little, Inc., on iron
and steel making; Vernon W. Ruttan and Hans Binswanger, of the
Agricultural Development Council, on technology transfer and re-
search in agriculture; and Simon Teitel, Mauricio Thomae, Hugh
Schwartz, and Jose Villavicencio, of the Inter-American Development
Bank, on acquisition, adaptation, and technology-development experi-
ences of entrepreneurs and decision makers in developing countries.
We are indebted to Gustav Ranis, Economic Growth Center, Yale
University, who joined us in our discussions and commented in detail
on an early draft of the report. The panel also extends thanks for
thoughtful written comments contributed by Nazli Choucri, Depart-
ment of Political Science, Massachusetts Institute of Technology;
Walter Falcon, Director of the Food Research Institute, Stanford Uni-
versity; and Larry E. Westphal, Development Economics Department,
World Bank.
Finally, I wish to thank Mercedes Rodriguez-Dickinson, who pa-
tiently typed the manuscript for printing; Elizabeth McClure, our
helpful editor; and Karen Hladek and Diosdada De Leva for biblio-
graphical assistance.




Introduction, 5
Development Processes and Technological Decisions, 9
Criteria of Technological Appropriateness, 10
Technological Alternatives and Information Transfer, 12
Determinants of Technological Decisions, 12
Technological Decisions in Agriculture, 13
Technical Decisions in Small Enterprise and the Service
Sector, 14
Policies for Improving Technical Decisions, 16

Introduction, 19
Efficiency and Technological Change as Sources of
Growth, 20
Capital Accumulation and Saving, 24
Growth and Improvement in Quality of the Labor Force, 26
Changes in Sectoral Composition, 29
Changes in Economic Organization, 32
Social and Political Development, 34

xii Contents

Introduction, 37
Maximization of Net National Output and Income, 38
Maximization of Availability of Consumption Goods, 39
Maximization of Rate of Economic Growth, 40
Reduction of Unemployment, 41
Redistribution of Income and Wealth, 43
Regional Development, 43
Balance of Payments Relief, 44
Promotion of Political Development and National Political
Goals, 46
Improvement in the "Quality of Life," 47
Complementarity and Competitiveness of Alternative Criteria
for Appropriate Technologies, 50

Introduction, 53
Range of Technological Alternatives from which Choices
Can Be Made, 53
Sources and Costs, 61
Role of Engineering Education in Generation and Transfer
of Technology, 64

Introduction, 67
Technological Decisions by National Private Enterprise, 67
Influence of Government Policies on Choice of
Appropriate Technologies, 72
International Private Enterprise and Market Influence on
Choice of Appropriate Technologies, 75
Influence of Bilateral and Multilateral Official Institutions
on Choice of Appropriate Technologies, 79

Contents xiii

Introduction, 83
Diversity in Agriculture, 84
Sources of Knowledge and Technological Change in
Agriculture, 87
Policies for Implementation of Technological Decisions
in Agriculture, 95

Introduction, 99
Characteristics of Small-Scale Enterprise and Choice of
Appropriate Technology, 101
Service Sectors, with Special Attention to Construction, 105

Introduction, 112
Economic Policies to Improve Choice of Technologies, 114
Policies for Technological Development, 115
Institutional and Political Policies for Choice of
Appropriate Technologies, 120
Investigation of Potential for Improvements in Licensing
Procedures, 121


Development Processes
The burgeoning interest in finding and implementing "appropriate
technologies" reflects a recognition of the essential role of technology
in development. Technological decisions and the pace of technical
change affect all development processes-economic, political, and
social-and, in turn, are affected by those processes. However, there
is great scope for variation in the relations among technological deci-
sions and development processes and little basis for a belief in
technological determinism.
Economic growth is generated from several different sources and in
different patterns. In general, the following economic changes are
combined in successful development:

Increases in the amount of resources available;
Improvements in economic efficiency with which resources are
Technological change that expands the potential productivity of
Changes in the relative weights in the economy of agriculture,
industry, and other sectors; and
Changes in organizational methods, with a declining role for
informal and family enterprise and a widening scope for specialization
in production and exchange.


Of the many problems that emerge in the course of development, the
growth of open unemployment is among the more intractable and
socially disruptive. It has its sources in rapid population growth and the
displacement of labor. The fundamental economic means for resolving
the problems are the acceleration and improvement in growth pro-
cesses, including the use of technologies now available for the inten-
sive and efficient use of labor. There is little evidence to suggest that
major research efforts to find efficient "intermediate" technologies for
small-scale village-level production would either be markedly success-
ful or contribute substantially to development.

Criteria of "Appropriateness" of Technology
The use of any particular technology is not an end in itself. The
criterion for an "appropriate" technological choice must be found in
the essential goals and processes of development. A number of differ-
ent criteria have been proposed, either implicitly or explicitly. These
include the maximization of output, maximization of the availability of
consumption goods, maximization of the rate of economic growth,
reduction in unemployment, regional development, reduction in bal-
ance of payments deficits, greater equity in the distribution of income,
promotion of political development (including national self-reliance),
and improvement in the quality of life. This last criterion has recently
been interpreted as development by means of relatively self-sufficient
village or rural activities.
These criteria are not only alternative in the sense of attracting
different degrees of support but, in many circumstances, can also be
competing. In addition, understanding of the relationships between
them and technological decisions is incomplete, especially for the last
three criteria.

Sources of Technological Information and Decision Processes
There are a number of different sources of technological information
and means of dissemination. These include new research and develop-
ment, transfer of existing knowledge through technical literature,
specialized education, technical consulting and equipment sales, and,
finally, adaptation of known methods. The available information does
not demonstrate that industrial research and development organiza-
tions in developing countries have been markedly successful compared
with agricultural research institutions. However, there is insufficient

Overview 3

data to judge the relative significance, benefits, and costs of the
alternative methods of generating and transferring technological
Technological decisions are made under different circumstances and
with different motivations that affect the degree to which they achieve
any of the criteria of appropriateness. Private enterprises may choose
their technologies to maximize profits, but this choice will be consis-
tent with the maximization of total national output and income only if
there are perfect markets for resources and products. In actuality,
markets have many imperfections, including some degree of market
control through monopoly power, the segmentation of markets, and
government interference through regulations, taxes, and subsidies. In
developing countries, informally organized family enterprise continues
to be important and, in general, it does not try to maximize profits in a
conventional sense. Government and public enterprise can be directed
to pursue any of the criteria of appropriateness, and even private
enterprise can be controlled for such purposes. Yet, unless
technologies that maximize profits or minimize costs are chosen,
continuing subsidization from the government budget is likely to be
required. The conflicting goals often posed for government and public
enterprises obstruct their achievement of any particular objective.

Technological Choices in Agriculture, Service, and Small-Scale
Agriculture and the choice of agricultural technology have a particu-
larly critical role in development given the size of this sector in most
developing countries. Much of the agriculture in these countries is
organized in family farms whose goals are not the conventional
profit-maximizing ones and which, because of their small size, cannot
engage in technological experimentation. Nonetheless, peasant
farmers have been willing to adopt agricultural innovations when they
are demonstrated to be profitable. There have been important innova-
tions in agriculture in high-yielding seed varieties, intensive use of
fertilizer, and mechanization. But conflicting reports of the effects of
these innovations on employment and income distribution make it
impossible to conclude whether, on balance, they displace labor,
although in some circumstances that may well happen.
Small-scale and service enterprise appear to be relatively successful
in absorbing labor. The technologies used in the latter sector seem to
have greater potential for efficient use of labor-intensive techniques.

This may also be true of most small-scale enterprise in general;
however, it is difficult to determine whether the labor-intensive
methods used are actually efficient.

Conclusions and Policy Recommendations
Unfortunately, little detailed, quantifiable information about the
characteristics of technology is readily available in a form necessary
for policy formation. Qualitative and verbal descriptions are in-
adequate. This lack of information means that there is little reason to
believe that simple and sweeping proposals for the adoption of particu-
lar types of technology will be "appropriate" by important develop-
ment criteria mentioned earlier. Moreover, no panaceas or quick
remedies for the problems of development are to be found in the choice
of particular types of "intermediate" technologies.
It is clear from existing studies that there are efficient technological
possibilities for using more labor-intensive methods in a number of
sectors in the developing countries. These are typically not fully
exploited, in part because of distortions in the prices of labor, capital,
and other resources sometimes created by misguided government
policies to stimulate development. Thus, the first recommendation is to
encourage the formation of resource and product price policies that will
encourage the use of efficient labor-intensive methods.
While only modest expectations and resource allocations are now
warranted, engineering research to attempt to extend the range of
efficient technological choices would be worthwhile to help the de-
veloping countries benefit from their particular resources. The second
recommendation, therefore, is for technological and economic re-
search that would start by first identifying the particular products
and/or processes for which such research is likely to be successful.
The third recommendation is to investigate the conditions that
appear to have stimulated the adoption of efficient labor-intensive
technologies and the methods for effective dissemination of informa-
tion about such methods. In particular, the role of industrial and trade
associations, which in some countries have enabled small-scale enter-
prise to compete successfully using relatively labor-intensive tech-
niques, should be studied.
In negotiations for the transfer of technology through licensing
agreements, buyers and sellers have unequal knowledge and bargaining
power. Since these negotiations may, therefore, result in distorted
choices and charges, the means for improving the outcome of such
licensing negotiations deserve particular attention.

1 Summary

Technological decisions and the pace of technical change affect all
development processes and, in turn, are affected by them. The combi-
nations and proportions in which labor, material resources, and capital
are used influence not only the type and quantity of goods and services
produced, but also the distribution of their benefits and the prospects
for overall growth. The significance of technological choices made in
the course of development extends beyond economics to social struc-
ture and political processes as well. The new products and new
methods, which are the central features of the technological transfor-
mations in the developing countries, represent especially sharp breaks
with the past and thus have especially profound consequences. Be-
cause they interact extensively and intricately with development pro-
cesses, technological decisions can also be used as conscious instru-
ments of development policy.
The growing interest in finding and implementing "appropriate"
technologies reflects an implicit, if not always explicit, recognition of
the essential role of technology in development. This report examines
the role of technology in developing countries to determine the content
and methods of appropriate technological decisions. Decisions about
technology are always specific; they are choices of particular products
and production methods. The objective of this report, however, is to
arrive at generalizations about the character and consequences of such
microeconomic decisions.


Development is a complex process that varies among countries,
reflecting differences in their social structures and goals, population
patterns, natural resource endowments, and capital accumulation. In
all cases, development means the alleviation of the terrible poverty that
afflicts most of the world, but it requires additional resources as well as
increased productivity and a wider variety of goods and services. By
necessity, then, development requires new technologies. Development
also involves social modernization and political transformations, which
both require, and are affected by, new technologies.
Recognition of technology's essential role in development does not
imply a technological determinism. Not only can alternative products
and methods be chosen, but the wider effects of these choices depend
strongly on the political and economic environments in which they are
implemented. In particular, technological change is not necessarily
beneficial for all development goals. Depending on the circumstances
of their introduction and use, technologies that increase resource
productivity may, for example, also increase income inequality or
social stratification or urban crowding. Although there are many intui-
tively plausible relationships among technological decisions and eco-
nomic and social development, formulating policies for technology
requires a deeper understanding, than intuition alone provides, and the
more rigorous definition of relations.
"The poor have always been with us," but only since World War II
has there been general recognition and international concern for the
massive problems associated with the poverty of the developing na-
tions. This new attention to development problems has been closely
related to the achievement of independence by colonial areas and the
emergence of governments that have made economic progress a pri-
mary objective. The 1950's and the early 1960's appear, in retrospect,
to have been a period of euphoria with respect to development. There
was general confidence that the problems of poverty could be over-
come and, in particular, that rapid development would follow indepen-
dence and the implementation of conscious national development
policies. Those national policies had their counterparts in the eco-
nomic assistance programs of the advanced countries, which, by their
short-term nature, also implicitly predicted such rapid progress that
the need for assistance would be relatively brief.
Progress was indeed rapid in many of the developing countries
during the first 15 or 20 years after the end of World War II. In many of
these countries, great new construction projects were undertaken; new
industries were introduced; and, undoubtedly, growth rates of aggre-
gate and per capital income were far higher than they had previously

Summary 7

been, as well as high relative to those of many advanced countries.
There was also important progress on other social fronts: educational
and medical systems were expanded in many areas, new types of social
organizations were introduced, and political participation appeared to
increase in many countries.'
By the mid-sixties and even earlier in some cases, however, a
growing pessimism-perhaps realism-with respect to development
prospects emerged in the industrialized countries and in the developing
countries, whose growth objectives became increasingly modest. In
addition, the governments in some of the developing countries became
spectacularly unstable. Though such political instability may not have
been characteristic of developing countries alone, it was widely be-
lieved in the West that the instability reflected a failure of development
efforts. Both the industrialized and the developing countries began to
doubt the major themes that had guided investment and foreign trade.
A succession of crop failures due to drought in parts of Asia and Africa
came to be regarded as evidence, in part, of a mistaken preference in
investment policy for industry over agriculture. Several unforeseen
and important side effects of major construction projects substantially
diminished the hoped-for benefits. The optimism that accompanied the
Green Revolution decreased when the expectations for a solution to
the world's food problems turned out to be unwarranted. Inflation
intensified in some of the countries in which it was endemic and
appeared in other areas that had been previously immune. Other
development problems persisted, and progress appeared limited on
nearly all fronts. This pessimistic view of progress in the developing
countries has come to be widely held in the seventies, with only a few
countries somewhat controversially identified as "success stories."2
Partly because of that disillusionment, the economic assistance pro-
grams of most industrialized nations, including the United States, have
declined in real if not money terms and have received less and less
political support. Table 1 tabulates real per capital growth rates in the
industrialized and less-developed countries since the 1960's. As those
numbers indicate, the permission has not been uniformly justifiable.
In the early, optimistic period of development analysis and policy as
well as in the more recent pessimistic days, technology has been
attributed a critical role both in resolving and creating development
problems. The potential importance of technological decisions for
employment and unemployment and the possible conflicts between
employment, output, and other social goals of development were
recognized, and strong views were formulated early in the economic
analysis of development problems. The conflicting strategies of


TABLE 1 Real Rate of Growth of Gross Domestic Product Per
Capita (Percent)

Country 1961-65 1966-70 (Average)

Developing 2.9 3.3 3.8
Africa 2.3 2.3 2.4
Southern Europe 5.6 4.8 5.1
East Asia 2.9 4.7 4.9
Middle East 5.1 4.5 8.3
South Asia 1.3 2.0 -0.6
Western Hemisphere 2.4 2.9 4.1
Industrialized 3.9 3.6 2.8

"Preliminary data.
SOURCE: World Bank Annual Report 1975.

Mahatma Gandhi and his political disciple, Jawaharlal Nehru,
exemplify two of the ideologies regarding relations between technology
and development. Gandhi advocated reliance on small-scale agricul-
ture and village industry and the technologies associated with them.
Nehru, on the other hand, was responsible for the first three Indian
Five-Year Plans that vastly expanded the heavy-industry sector using
modern technologies with all their accoutrements. In the first debates
over Indian development policy, it was recognized that the sectoral
decisions also implied technical choices with far-reaching conse-
quences for employment.3
Most recently, there has been a resurgence of concern with the
policy implementations of technological decisions under the headline
of discovering and implementing "appropriate technologies" for eco-
nomic development. Avoidance of widespread unemployment has
moved into the forefront of economic issues in many of the developing
countries. This problem is related to the rapid population growth rates
that accelerated 20 to 30 years ago and that are now beginning to
deliver massive numbers of new entrants to the labor forces of many
developing countries. If the patterns of labor absorption of the recent
past continue into the future, many of the new potential workers will
not find jobs. That, in turn, would have important consequences for the
distribution of income and wealth unless consciously offset by gov-
ernment tax and expenditure programs.
Such diverse problems as balance of payments difficulties and ur-
banization can also be related in part to technological decisions. These

Summary 9

decisions affect the relative costs and the quantities of imports and
exports and, therefore, the characteristics of each country's balance of
payments. In addition, patent fees and royalties for imported
technologies are a direct expenditure for foreign countries. Similarly,
the exploding problems of urbanization in developing countries are
associated with the technological choices that affect relative sectoral
development. Technological decisions in part also determine the rela-
tive expansion of employment opportunities in rural and urban areas,
which, in turn, affects the patterns and intensities of internal migration.

Development Processes and Technological Decisions
The current output of goods and services depends on the efficiencies
with which labor, capital, and other available resources are used.
These are determined in part by the technological choices made from
the available alternatives. Technological choices. in turn, depend on
the character of economic motivations; on immediate influences such
as market prices of outputs and inputs and government taxes and
subsidies; and on political influences constraining engineering and
economic decisions.
The increases in output necessary to alleviate poverty require more
and more resources. Technological change in the production of old
products as well as the introduction of new products will also account
for a substantial part of the economic growth. The potential contribu-
tion of technical change is related to investment rates, however, since
new capital formation in many cases is the "carrier" of technological
improvements. Investment must be based on either domestic private or
public saving or foreign saving, but the substitutability of domestic and
foreign saving will depend, in part, on the nature of the technologies
employed. Although open and disguised unemployment are increasing
problems in many countries, population growth in many other countries
with rich resources and low population density will contribute to over-
all growth. Whether labor-force growth intensifies the unemployment
problems in developing countries depends in part on the employment
characteristics of the technologies employed. The education and skill
requirements of the labor force will also be determined by the tech-
nologies adopted.
Although there is controversy over the most desirable patterns,
changes in the sectoral composition of production are essential for
development. Both agriculture and industry must grow, and, if de-
velopment is successful, the latter sectors will grow faster at some
point to match the changes in patterns of demand that accompany

income growth. The precise patterns of sectoral growth and the effects
on general development goals will again depend partly on the
technologies utilized.
Changes in economic organization will both cause and affect
technological decisions as the role of household and small-scale enter-
prise in the economy changes and as public and private corporate
activity expands. Technological change is one of the major agents of
social modernization affecting society intimately at all levels. Directly
and indirectly, technological decisions will affect the processes of
political development and be affected by them. It is to be expected that
political considerations will often dominate narrowly defined economic
goals in the choices of projects, products, and technical methods.

Criteria of Technological Appropriateness
Since the use of any particular technology is not an end in itself, the
criteria of appropriateness for the choice of technology must be found
in the goals of development. These goals are concerned not only with
the volumes of output and income generated by an economy but also
with the way they are produced and distributed among the population;
they include, as well, particular patterns of national political change
and national independence.
The criterion for choice of technology most commonly applied,
either implicitly or explicitly, is net output maximization or cost mini-
mization. Application of this criterion will, in general, achieve both
physical and economic efficiency in the use of resources and, thus, will
do as well as possible in increasing the size of the total output pie. Yet,
this criterion would be satisfied by the operation of private markets
only if such markets function perfectly, but there are, in fact, many
sources and types of "imperfections."
An alternative criterion, maximizing the availability of consumption
goods in the present and the future, may have the same implications as
maximizing growth, depending on the index of growth used.
Both of the above criteria may be different from maximizing em-
ployment. The latter goal, though often thought of in terms of the
number of jobs provided, is, however, an ambiguous expression of
social policy. Employment serves a number of functions: output gener-
ation, income distribution, social recognition and personal satisfaction.
and "taxing" of the potential leisure of the community. Thus. if
employment is to be an operational criterion of technical choice, its
content must be specified in detail.
While income distribution is an issue of increasing social concern,
the relations between technical decisions and the achievement of this

Summary 11
particular goal are seldom clear. Regional development is a specific
aspect of the income-distribution goal. Again, technologies seldom
have specific regional implications, although decisions about regional
distribution will, in turn, emphasize particular sectors and particular
types of technologies.
Similar observations apply to balance of payments relief, which is
often set as still another alternative criterion of technological choice. It
has been argued that export-oriented production will contribute more
to development than import substitution. The claim is based partly on
the judgments that international competition faced by export produc-
tion will promote efficiency and that there is the potential of achieving
economies of large-scale production by selling to world markets.
However, governments appear as able to distort export patterns via
subsidies and other measures as they are to create and preserve
distortions in import-substituting industries.
Political development has also been proposed as a criterion of
technological choice. The creation of a national political system can be
both an instrument and a result of an interrelated national economy in
which large-scale projects undertaken with government initiative and
support have a major role. By comparison, self-sufficient village-level
production is more consistent with dispersed political power and
decision making and with lower capability of the political system to
evolve and respond to system demands. However, understanding of
the political sources and consequences of technical decisions is still at
an early stage.
Improvement of the quality of life has recently been emphasized, in
general terms, as a major consideration for technological choice.
Proponents of this criterion prefer qualities associated with village or
rural life, small-scale activity, self-sufficiency, minimum ecological
effects, and equality in income distribution. The desirability of these
particular qualities and the extent to which other objectives should be
sacrificed to achieve them is clearly a matter of preference. Moreover,
little attention has as yet been paid to the trade-offs among these quality
of life goals themselves and other development objectives.
The criteria proposed for technological decisions are often competi-
tive as well as complementary. It cannot be presumed that pursuing
any one will automatically satisfy all or any of the others. Therefore,
before an "appropriate" technology can be chosen, the criterion for
choice must be determined and, implicitly or explicitly, other criteria
must be rejected. Moreover, choices once made are not automatically
feasible. If a criterion other than cost minimization is used for private
or for public enterprise that must meet a market test, then the enter-
prise is not likely to be viable without continued government interven-

tion on its behalf. Thus, widespread use of other criteria will, in turn,
require a large-scale government program of taxes and subsidies or
regulation for the decisions to be effective.

Technological Alternatives and Information Transfer
Information about the range of available technological alternatives,
their precise characteristics, and their implications for the criteria of
appropriateness is essential for policy making. Yet such information is
not readily available, and it is costly and difficult to acquire. Overall
investigations based on existing statistics yield only equivocal results,
and case studies are seldom generalizable.
Although the weight of evidence suggests that most producing sec-
tors are not restricted to a single efficient combination of resources,
estimates of the range of choice available do not clearly demonstrate
that wide and easy substitution is possible. Case studies, mainly in light
manufacturing, do indicate that alternative techniques are being used,
particularly in activities peripheral to central production processing,
but these case studies often do not assure that the alternatives are all
efficient. Even less information is available to show how various tech-
nologies satisfy distributional or other criteria of appropriateness.
Many sources of technological information exist and are used. These
include some relatively inexpensive sources such as technical litera-
ture, formal and informal education and training of individuals, as well
as licensing of patented production processes and sales of expertise
and equipment. The intrinsic measurement problems and limited re-
search make it impossible to evaluate the relative importance and
potential substitutability of these sources. Some information on the cost
of using the various sources exists, but is so limited that generalizations
are not warranted. Adaptations of technology in the course of produc-
tion may be one of the most common and fruitful sources of technical
change, but again little hard information is available.

Determinants of Technological Decisions
Technological choices in developing countries are made by various
types of decision makers and under a variety of sources of influence.
They will not all necessarily have the same objectives or conform to the
same ideas of what is "appropriate." The owners and managers of both
national private enterprise and international enterprise will be con-
cerned with their own profits, but the latter will have different oppor-
tunities and constraints than the former. Both will be reacting to the
incentives of the resource and product markets in which they buy and

Summary 13
sell. Their technical decisions will be affected by the influences they
exert in their markets and the extent to which such markets are unified
and are affected by government tax, subsidy, and regulation policies. It
may well be optimal for a developing country as well as for a multina-
tional firm to import technologies from more developed countries when
both research and development as well as capital and operating costs
are taken into account.
Government corporations may or may not behave like private enter-
prises, depending on their organization and on the objectives and
controls imposed on them. In some cases, government enterprise has
been used to pursue particular employment and income goals, although
the pursuit has sometimes been piecemeal and inconsistent because
clear incentives were lacking.
The official national and international economic assistance agencies,
such as the U.S. Agency for International Development and the
International Bank for Reconstruction and Development, may have a
different conception of the development process than the developing
countries themselves. They have used their loan and grant programs to
advance strategies and impose conditions of performance that the
developing countries would not otherwise have adopted.
The methods of implementing decisions also vary among decision
makers. In private enterprises, decisions on technology may be made
directly but sometimes require approval of a government licensing
agency. Governments can resort to direct regulation and their fiscal
system to exert their influence. The official international sources of
funds rely on their expertise as well as their power to grant or with-
hold funds to persuade and guide.
While technological decisions in a developing economy should be
coherent and compatible with a particular set of goals, such coherence
and compatibility is never fully achieved, just as it is not achieved in
the developed countries. Rather, economic policy decisions in the
developing economies reflect sets of influences that, to some extent,
act at cross-purposes. A major task of development policy is to reduce
the frustration and inefficiency associated with inconsistent goals and
methods, but this task is never fully accomplished. It is especially
difficult to achieve when the choice concerns appropriate technologies,
because the knowledge, interests, and operating methods of the differ-
ent decision makers and sources of influence are often at variance.

Technological Decisions in Agriculture
The special features of agriculture in developing countries derive in
part from the diversity of their production conditions. Climate and soil

conditions can vary widely, even within relatively small countries, and the
farm products and the inputs they require also vary. In addition, there
is great diversity in the organization of farm enterprise. The family
enterprise, as distinct from capitalist organization, is of particular
importance in agriculture in the developing countries. In capitalist
enterprise, labor and other resources are purchased to maximize
profits, while in family farms, resources and technologies are chosen to
maximize the net returns to the family's capital and labor as a whole. A
number of different institutions prevail in both the formal and implicit
contracts that regulate tenancy, labor obligations, and use of other
resources as well as in decision-making processes. These are not all
equivalent to the practices of conventional business enterprises and
require individual analysis.
The search for technological improvements in agriculture is typically
centralized in research stations and is usually publicly sponsored
because the scale and time horizon of the research is often far beyond
what even large landowners can afford. This centralization of
technological research requires a system for disseminating results.
That system is formalized in the agricultural extension services, which
are also feedback mechanisms that spur research to meet farm prob-
Improved seed varieties, fertilizers, pesticides, mechanical equip-
ment, and improved farming practices have resulted in major "revolu-
tions" in agriculture in the developing countries. When individual
farmers adopt these innovations, their own interests dictate the crite-
rion applied, just as is the case of individual enterprise in other sectors.
The criterion is usually output maximization or cost minimization.
However, government policy can exert powerful influences on deci-
sions of individual farmers through taxes and subsidies as well as
through direct controls and the provision of new technologies. Though
these policies can be used consciously to direct agricultural develop-
ment, their effects are often unintended by-products as farmers pursue
their own interests.

Technical Decisions in Small Enterprise and the Service Sector
There is impressive evidence that the potential for choice of tech-
nologies to permit more intensive use of labor is greater within small-
scale enterprise than for larger enterprise in the same sector. However,
existing studies have not yet established that this greater employment
intensity is consistent with the other criteria of technological appropri-
ateness, in particular the criterion of economic efficiency or cost

Summary 15

minimization. Small enterprise operates in an environment with many
noncompetitive elements although their significance is difficult to
evaluate. Moreover, because it typically uses a relatively high propor-
tion of family labor and other resources not provided through markets,
small enterprise does not follow the conventional rules of profit
maximization. Even the continuing existence of small firms cannot be
taken as conclusive evidence of their ability to meet the test of
efficiency and survival in a competitive environment.
Some types of traditional small-scale family enterprise are vertically
integrated and completely transform raw materials into final products.
Other modern small enterprises have specialized to lower costs and
coordinate with larger enterprise. The successful achievement of this
goal has been facilitated in some cases by government-assisted trade
While there are questions about the economic efficiency of small-
scale enterprise, such enterprise may be successful when judged by the
other standards of appropriateness. Not all "inefficient" types of
economic activity are socially acceptable, but small-scale family enter-
prise is typically well regarded and is frequently given special protec-
tion. On the other hand, family enterprise has also been the locus of
what are now considered social abuses, for example, the intensive use
of children and women in production, or work in unhealthy and danger-
ous conditions, often resulting from the use of housing and other casual
structures for production.
The service sectors cover a particularly varied group of activities.
Some need highly trained professionals; other sectors depend mainly
on persons with little or no training. In the construction sector, which
has a particularly critical role in the development process, a number of
alternative technically feasible technologies can often be used. Re-
search on highway construction techniques indicates that the most
labor-intensive methods and intermediate technologies are not as eco-
nomically efficient as the capital-intensive methods, but it may still be
possible to develop alternative highway designs for which labor-
intensive methods do meet the efficiency criterion. Alternative con-
struction technologies may also have different implications for the
satisfaction of the criteria of appropriateness other than cost minimiza-
Within the service sector, the health care delivery and educational
systems appear to have important technological alternatives. These
alternatives also provide different "qualities" of services, and these
quality differences make it difficult to judge the acceptability of the
technological alternatives by the various criteria of appropriateness.

This is also true of the marketing and storage sectors, which are
considered important in the achievement of greater benefits from
recent increases in agricultural productivity. The potential contribu-
tions of alternative technologies in these sectors remain to be carefully

Policies for Improving Technical Decisions
Because the amount of information and expertise on which to base
policy is limited, a major goal at this point must be to improve the
knowledge base.
While many influences affect technological decisions, it is not clear
how they operate and what their relative weights are. Thus, no single
emphasis or policy can be recommended that will ensure that more
appropriate technologies can be discovered for developing countries. A
single-minded approach is likely to lead to decisions that are inappro-
priate by some important development criteria. To be appropriate,
technological decisions must be tailored to the individual country, the
particular sector, and the particular criterion being pursued. Village-
level or intermediate technologies are seldom defined precisely and
cannot be regarded as solutions for the problems of development.
There is little reason to believe that the intensive search for such
"intermediate" technologies will facilitate development and that a
substantial diversion of resources to this search is warranted.
Three types of policies can be distinguished: (1) policies that improve
the incentives operating in the choice of technology; (2) those that will
expand knowledge of technological alternatives particularly suited to
developing countries; and (3) institutional changes to improve and
lower the cost of disseminating technical information.
To help make the choices of technology more consistent with de-
velopment goals, the economic and political incentives operating on
policy decisions should be examined carefully. Market prices may not in
themselves ensure the adoption of appropriate technologies, but they
certainly should not favor inappropriate technologies. Projects should
be evaluated with correct shadow prices as well. Economic research
should be sponsored to improve understanding of the technological
choice process; this research should include the study of the costs of
information and its relation to the exercise of influence over technolog-
ical decisions and the study of the efficiency of small-scale enterprise.
To expand knowledge of technological alternatives, task forces of
engineers and economists should be formed to generate priority lists of
production methods and problems. These lists should indicate the

Summary 17
sectors in which it was both of particular importance to extend the
range of efficient techniques (especially at relatively low levels of
output) and in which research would have reasonable opportunities for
success. Research to actually explore new technologies should then be
supported on the basis of the priorities established.
The general presumption should be that technological investigations
should be done in close conjunction with the potential users of the
technologies. Technological information should also be accumulated
and disseminated with careful attention to the evaluation of its quality
and its economic as well as technical implications. Research should
also be sponsored to improve the understanding and effectiveness of
the public and private mechanisms customarily used to collect and
disseminate technological information.
Although institutional conditions are important in the determination
of technological choices, these factors are not well understood and
research in this area needs to be expanded. For example, the role of
trade associations and government standards in expanding and modern-
izing small-scale enterprise should be investigated. Since technologi-
cal decisions and political factors are interdependent in development,
These interdependencies deserve more study as well.
At present there is some degree of monopoly power in the differen-
tial possession of technological information and project-by-project
bargaining. Thus, existing licensing methods can be expected to lead to
inefficient uses of technological information. Consideration should be
given to finding means for improving the terms on which technological
licensing is done.
The body of this report begins in Chapter 2 with a brief survey of
development processes, with particular attention to technical influ-
ences. This survey is intended only to provide a background for the
later more detailed analysis of the relations between technological
decisions and economic and other social goals and processes. Alterna-
tive criteria for technological appropriateness are described and
analyzed in Chapter 3. Chapters 4 and 5 review the factors that
determine the general effectiveness of a developing country in making
appropriate choices of technology. These factors include both the
general, systematic influences that condition choice processes, the
decision methods implicit in the functioning of the economy, and the
conscious and purposeful use of techniques for choosing among de-
velopment projects. In Chapter 6, some of the unique features of the
development processes in the agricultural sectors are surveyed, and
their significance for the choice of appropriate technologies is examined.
Chapter 7 considers briefly development processes and technical


choice in small-scale enterprise and the service sectors. Finally, Chap-
ter 8 suggests technological, economic, and institutional policies to
help improve the coherence of development goals and performance.

1. There is an enormous literature on development history and policy. A useful survey is
provided in A. Maddison (1970) and in H. Chenery et al. (1974). The annual reports of
the Organization for Economic Cooperation and Development, Development Assis-
tance Committee and the International Bank for Reconstruction and Development
(World Bank) provide up-to-date information.
2. For a skeptical view of the achievements of the post-World War II development
efforts in improving the conditions of the poorest people in most of the developing
countries, see I. Adelman and C. Morris (1973). The observed patterns of develop-
ment are also surveyed in H. Chenery and M. Syrquin (1975).
3. J. Lewis (1962).

2 Interaction of
Technological Decisions


Development Processes

A survey of development processes and the role of technology is
undertaken in this chapter to provide a general background for the
later, more detailed analysis of technological decisions and related
policies.' Understanding the relationships in development, however,
involves special difficulties. First, there is great diversity among the
developing countries. Second, despite recent intensive studies, only a
limited amount of relevant information is yet available. Finally, and
fundamentally, there is only a limited theoretical understanding of the
processes of economic growth and social change involved in develop-
Most of the characteristics of the developing countries vary enor-
mously. The geography of the developing areas includes, for example,
the Ganges basin and the teeming cities of the Asian subcontinent, the
high Andean slopes and valleys, the semidesert of sub-Saharan Africa,
the richly endowed and relatively sparsely populated African coun-
tries, and the great river valleys of the Mideast. There is also wide
diversity in the cultural and political patterns of the societies and in
their political systems. There are areas in which tribal organization is
still important, areas with a long history of political independence, and
countries that have emerged from colonialism only recently. There are
economies in which there is a high degree of public intervention and
direction, economies that are primarily organized by private enter-

prise, and many different mixtures of public and private enterprise and
control. This diversity must be fully respected in any analysis and is a
major barrier to generalization.
While it is easy to recognize the heterogeneity of the developing
countries, it is difficult to document in detail the characteristics of any
one of them. Their statistical services are often relatively new;
scholarly investigations are burgeoning in some places but just begin-
ning in others, and they still do not add up to a comprehensive view of
any of the countries.
It is especially difficult to obtain those quantitative measurements of
the characteristics of technologies that are necessary to determine the
economic and social implications of specific methods. While the ulti-
mate goal is to be able to generalize about the role of technology in the
development process, these generalizations must be based on the accumu-
lated knowledge of particular technologies. For this purpose, qualitative
and verbal descriptions are inadequate, although frequently they are the
only basis presented for policy propositions. Quantitative technical
information is lacking for several reasons. Technical decisions are
made at the plant and process levels, and the collection of information
at that level is a time-consuming and costly process. In addition,
technical data in the quantitative form desired are frequently at the
heart of the competitive advantage of particular plants and are, there-
fore, regarded as privileged information.
Although theoretical understanding of development processes has
undoubtedly progressed, it is necessary to be modest about what has
been achieved. Development is a flourishing field of specialization in
political science and economics, but advances are achieved through
slow accretion rather than spectacular breakthroughs.
Of all the influences on development, the economic influences of
technological decisions can be most thoroughly traced, and they will be
emphasized in this survey. The character of other political and social
transformations in the development process and their relation to
technological decisions will then be discussed briefly.

Efficiency and Technological Change as Sources of Growth
One of the most obvious features of the economic life in the less-
developed areas is the strikingly "old-fashioned" technology often
employed. While "modern" facilities are being rapidly created and are
expanding beyond industrial enclaves in some of the more successful
cases, traditional techniques still prevail in many sectors. No special
census or accounting of inputs is needed to see that a bullock cart is

Interaction of Technological Decisions 21
different from a truck and that handsaws are different from power
tools. Neither is special insight needed to understand that a person
driving a truck or using a power saw can accomplish more than
someone guiding a bullock cart or using a handsaw. And there is little
doubt that higher productivity of individual workers is one of the
essential requisites of development.
Yet it does not follow that providing a worker with mechanical
equipment and power tools to improve individual productivity will
benefit the economy as a whole. Typically, there are not enough trucks
and power saws for all the drivers and construction workers who must
transport and build. In such circumstances it is usually unwise for the
entire economy to provide only a few workers with modern equipment
and leave the rest with primitive tools. What is right for the economy as
a whole depends on the size and quality of the labor force, the total
availability of equipment and other resources, and the future and
present goals of the economy.
This reasoning suggests that technical feasibility and labor and land
productivity by themselves are not adequate criteria for choosing tech-
nologies. Rather, technological choices should be made in terms of
general economic and social criteria. The economic evaluation of
particular technologies is known as cost-benefit analysis. To perform
such analysis, qualitative technical characteristics must be translated
into estimates of how much of each type of resource is required by the
technique. Then, by using wages and prices to value the costs of hiring
and buying resources, the total costs of alternative production methods
can be compared as well as the specific costs of using each resource.
While such detailed comparisons will not be attempted in this report,
the method will be implicit in all the economic analyses of the tech-
nological alternatives.
Describing technologies in terms of the amount of each resource they
use makes it possible to distinguish two kinds of inefficiency. The
common observation of technical backwardness in the developing
areas has often been considered an indication of physical inefficiency in
the use of resources. Strictly speaking, physical inefficiency exists
whenever more of at least one resource is used than is absolutely
necessary to achieve a particular amount of output. Assessment of
economic efficiency, on the other hand, requires comparison of the
total costs of using different sets of resources to achieve a particular
level of output. A particular product may be yielded by a number of
physically efficient techniques. But, of these alternatives, only the one
that minimizes social costs will be economically efficient.
The proper choice of prices for evaluating economic efficiency is

itself a complex issue. The "proper" prices should reflect the real
scarcities of the resources in terms of the goals of the economy. These
prices, called "shadow prices," would be generated by an economy
that achieved its social goals with perfect efficiency. In fact, such
economies do not exist. There are many sources of deviation from
"perfection," some inherent in the nature of technology itself. For
example, when "natural monopolies" result from economies of large-
scale operations, market prices will not necessarily reflect the real
productivity of the resources used. The essential operations of gov-
ernment in its taxing and spending may also interfere with efficient
production. In addition, government powers are often used to support
special-interest groups at the expense of the economy as a whole, with
distorted prices the result. Because of the complexity of the problems
of pricing resource inputs correctly, assessing economic efficiency is
generally more difficult than assessing physical efficiency.2
Even in advanced countries, the "best-practice" technique typically
differs substantially from the "average-practice" technique.3 Yet the
production inefficiencies of the developing countries are considered
more profound and extensive. They have been associated with limited
access to technical information and with a relatively poorly educated
and trained labor force. Conventionally, such inefficiencies are attrib-
uted to the influence of traditional habits of thought and behavior on
technical decisions as contrasted to the rational problem-solving ap-
proaches to management considered characteristic of industrialized
The view that traditional production is inefficient has been chal-
lenged in recent years, especially with respect to peasant agriculture. It
has been argued that peasant agriculture actually represents an optimal
use of available resources. The means by which this optimality has
been achieved have not been set out in detail, but the argument appears
to be that a kind of Darwinian process of survival of the most efficient
technologies has eliminated both physically and economically ineffi-
cient methods.4 While this new view is plausible, the empirical evi-
dence for peasant efficiency is equivocal at best, and the analysis and
evidence for the existence of an effective selection process for efficient
techniques is almost nonexistent.
A view still widely accepted and supported by some empirical
studies is that important inefficiencies are commonly found in nonag-
ricultural sectors of developing countries, both in the traditional "un-
organized" sectors and in the modern "organized" plants. The fre-
quent references to the necessity for greater literacy among workers
and better engineering and organizational techniques among managers,

Interaction of Technological Decisions 23
taken in context, imply that there are neglected opportunities for
improving output without substantial additions of physical resources.
Yet the evidence is, perhaps by necessity, mainly anecdotal; it consists
of isolated observations or, at best, a case study that is thorough but
limited to a particular country at a particular time."
Technological change means the replacement of economically effi-
cient methods by still better methods. The replacement of steam by
diesel locomotives on most of the world's railways and the replacement
of open hearth by basic oxygen furnaces are dramatic examples. These
innovations made it possible to produce more transport services and
more steel with less equipment and manpower. As has been said, the
effects of such technical change make it seem as if the laws of
thermodynamics were repealed, for more output is achieved with given
Studies beginning in the 1950's that tried to identify the sources of
economic growth in Western Europe and the United States have
suggested that technical change was responsible for most of the growth
during this period.6 These conclusions, as well as the obvious impact of
new technologies in many areas of modern life, appeared to validate
the post-World War II emphasis on research and development as the
source of technological change and economic growth. The conclusions,
which were also extended to Western Europe, were substantially
revised in subsequent studies. Sources of growth other than simple
capital accumulation and increased use of labor were diligently tracked
down, and thus the role left for technical change was reduced.7
Yet technological change is, indeed, a remarkable phenomenon. It
generates new products that relieve pain, cure sickness, and improve
communication, and it increases output of all kinds. It creates problems
of adjustment as well, and it has been argued that the problems it
creates are more onerous than the problems it alleviates.8 A complaint
gaining currency is that the new technologies imported by the develop-
ing countries from the industrialized countries are not "appropriate" to
their resource endowments and conditions. As a result, imported
technologies do not help absorb their rapidly growing labor forces and
make demands for materials, organization, and skilled personnel that
cannot easily be met. It is implied that appropriate technologies can
provide more of the benefits of technical change with fewer deleterious
side effects.9 The essential questions are the availability of specific
alternative production methods and their economic and social costs
and benefits. These issues cannot be settled by assertion. They will be
the continuing focus of analysis in this report.
It has also been argued that the agents of technical change-the

innovators and entrepreneurs-are the central figures of the develop-
ment process. It is they who implement the new methods and introduce
the products that provide the major new investment opportunities, and
these in turn result in the improved worker productivity and the
production of new goods that are the essence of economic growth. The
success of these innovators encourages others, more timid, less qual-
ified, or less opportunely placed, to imitate the specific innovations and
entrepreneurial patterns of economic behavior.10 By this reasoning, the
process as well as the content of technological change is significant for

Capital Accumulation and Saving
Outside their modern production enclaves, developing countries have
limited tools, equipment, and other capital facilities. The standard
examples that come to mind are farmers with simple stick plows,
construction workers using picks, hoes, and baskets to move dirt and
rock, and artisans using mostly hand tools and little, if any, powered
equipment. While these examples do not accurately characterize all the
production activities of any developing country, they do reflect in
extreme form the relative capital scarcity common to all the developing
When capital is so scarce, it is not surprising that labor productivity
and income are low and that poverty is widespread. The obvious
remedy is more capital, and that is certainly a necessary condition for
development. But through emphasis and/or neglect of other conditions,
capital investment has often been treated as a sufficient condition."
That is certainly wrong, since a great deal depends on the effectiveness
with which capital is used and that, in turn, depends on many other
features of an economy.
The theories of growth developed in the 1950's, which related
increases in income and output to increases in capital created by new
investment, were attractive in their simplicity and their apparent power
both to explain the existing poverty and to plan for future development.
Though the investment-growth theories often neglected the role of
labor, that did not seem to be a fundamental omission when they were
first applied to the "labor surplus" economies of south Asia. Even now
a simple relation between increases in capital and associated increases
in output is often the basis of much development planning. When used
skillfully and supplemented by consideration of manpower and other
requirements for growth, the capital-output relation can be a powerful
analytical and policy tool.

Interaction of Technological Decisions 25
To increase the capital stock by net investment there must be saving.
When resources are fully employed, they must be saved from current
consumption to make them available for capital formation. Saving can
be generated from several sources. Saving is perhaps most often
thought of in private and personal terms: individuals, by not spending
all their income on consumption goods and, thus, by not exercising all
their claims on current output, make some of it available for new
capital formation. Private business saving is at least as important as
personal saving in many of the developing countries. Businesses, by
not distributing to their owners all their net profits, also effectively
restrict consumption and, thus, free resources for investment in new
capital. Governments and government enterprises also generate saving
by restricting their expenditures on currently produced and consumed
services to less than their tax income or revenues. Typically, any
surpluses of this type are moved directly into capital formation by
government, but they could be made available, directly or indirectly,
for private capital formation. Finally, foreign resources used for capital
formation can also be considered "saving"-foreign saving-and the
basis of new investment.
One advance in the economic understanding of development pro-
cesses has been the recognition of the limits to the potential substituta-
bility of savings from domestic and from foreign sources. Specific
types of facilities and equipment are necessary to create each specific
type of investment goods. If those facilities exist domestically, then
domestic saving can help make them available for new capital forma-
tion. However, if the necessary production facilities are not available
domestically, no amount of domestic saving will directly free resources
for new investment. For example, a developing country that wants to
invest in a hydroelectric station of the largest size must build dam and
powerhouse structures of concrete and install large turbogenerators. If
there are domestic cement plants, the cement for the concrete will be
released from domestic supply-if personal saving limits private hous-
ing demand, for example. But only a few countries in the world have
the specialized metal-making and metal-working equipment required to
produce the largest turbogenerators. So those must be imported.
However, technically necessary imports of investment goods can be
made available indirectly via domestic saving if such saving makes
possible export of goods that would otherwise not be exported. Thus,
domestic saving, which reduces domestic purchases of textiles and
makes them available for export, in effect earns foreign exchange to
pay for imports of turbogenerators as well as other types of goods. Yet
there are limits to the rate at which exports can be expanded, and these

in turn create limits to the indirect substitution of domestic for foreign
saving for most developing countries.12
According to one view of economic growth processes, there are
limiting factors to what could be achieved by capital accumulation.
Diminishing returns to new investment, like diminishing returns to
labor in the Malthusian view, would steadily reduce the net contribu-
tion of new capital formation. A more recent view, embodied in most of
the development planning models, holds that the right kind of invest-
ment can break nearly all production bottlenecks. In this view, almost
all "fixed factors," other than geographic conditions, can be expanded
by investment without diminishing the returns to the investment. Even
the availability of arable land can be increased by irrigation and other
land-improvement schemes.13 However, there will never be any defini-
tive tests of the alternative viewpoints, partly because of another
essential aspect of capital formation: its function as a carrier of new
Although difficult to measure precisely, an important part of most
new investment carries technical change with it. Standard types of
investment goods can be bought "off the shelf," but much new
investment is "tailor made," and the tailoring process typically adds
new technical arrangements, devices, and other improvements even
without major innovations. Thus, capital formation and technological
change are not separate growth processes, though it is often convenient
to discuss each separately.

Growth and Improvement in Quality of the Labor Force
One of the important differences among the developing countries is the
relative population densities and the ratios of available labor to capital
and other productive resources. There are "overpopulated" and "un-
derpopulated" developing countries, where these terms reflect general
impressions of differences in labor/resource ratios. These differences
make generalizations difficult, but some features of population and
labor-force issues are common to most developing countries:

Rapid population growth rates;
Substantial and growing open unemployment;
Large-scale rural-urban migration; and
Education and skill patterns in which significant shortages at some
levels may coexist with surpluses at other levels.

With only few exceptions, developing countries have high rates of

Interaction of Technological Decisions 27
population growth. These rates are nearly always substantially over 2
percent per year and often 3 percent per year or more, whereas the
growth rates in most advanced countries are 1 percent or less. The
relation of the labor force and its growth to total population and its
growth depends on the age structure of the population and participation
in the labor force. A relatively high proportion of a rapidly growing
population is too young to be in the labor force. In addition, for reasons
of culture and religion but also because of preoccupation with child
rearing, the participation of women in the labor force, at least in urban
economic activity, is often relatively low in the developing countries.
Nonetheless, because the relatively high overall population growth
rates have been sustained for some time in most of the developing
countries, the growth rate in their labor forces is also now relatively
The projected rapid increase in the labor forces of the developing
countries compared with current labor-absorption rates has made
unemployment one of the dominant issues in the developing countries.
The current unemployment figures, partly because of their statistical
unreliability, do not now indicate a widespread and overwhelming
problem. However, urban unemployment in many developing coun-
tries is growing rapidly and is already considered a major issue. The
growth statistics and expectations, combined with the information now
available on the growth of employment that has accompanied de-
velopment over the past 30 years or so, produce a gloomy forecast. A
calculation by Turnham is typical:
By simple arithmetic, a manufacturing sector employing 20 percent of the labour force
would need to increase employment by 15 percent per annum if only to absorb the
increase in a total labour force growing at 3 percent per annum. In addition, because of
productivity increase, about 3 percent growth per annum in output seems needed to
maintain a constant labour force in manufacturing.14
The potential for expanding employment depends on capital accumu-
lation rates, the labor force/resource ratios prevailing in these sectors,
and the degree to which the new technologies permit additional labor to
be used with additional resources. For example, some developing
countries, though capital-poor, are rich in mineral resources and in
underutilized agricultural land. These resource-rich developing coun-
tries have "frontiers" to exploit and, as claimed for the United States
during its most rapid population growth in the nineteenth century,
these frontiers may provide a "safety valve" for the absorption of
labor. However, even in the relatively resource-rich and "underpopu-
lated" developing countries, difficult unemployment problems may
emerge as they adjust to rapid population and labor force growth rates

and to the new patterns of sectoral growth. In developing countries not
richly endowed with agricultural land and/or important mineral de-
posits, urban manufacturing and the service sectors must absorb most
of the increases. That, in turn, requires new investment to provide the
equipment necessary for the productive use of labor. So labor absorp-
tion depends on investment rates as well as the degree to which the
technologies employed permit intensive labor use.
Although overall growth in population and the labor force is one of
the major factors aggravating unemployment, rural-urban migration is
also a contributing factor. These migration processes are not com-
pletely understood, but rural-urban income differentials apparently
play an important role. These differentials may arise from real produc-
tivity differences in rural and urban areas that, in turn, depend on the
use of modern capital-intensive technologies in urban production. The
income differentials may be created to induce higher productivity by
reducing labor turnover and/or by improving nutrition and providing
other social amenities associated with higher income. But the rural-
urban income differentials may also be the result of minimum-wage
laws or union organizing, which is effective in urban areas.1
Typically, there are significant shortages of skilled labor in develop-
ing countries even when unemployment is substantial. While develop-
ment based on natural-resource exploitation is the easiest in many
respects, even that requires new processes. And, if development is to
spread and become permanent, manufacturing and construction sec-
tors must be expanded and new processes must be adopted. In the
agricultural sector, development typically requires new seeds, fertiliz-
ers, pesticides, and, often, changes in irrigation patterns. All of these
changes require new and additional skills in the labor force.
The creation of labor skills through education and training is now
conventionally considered to be a kind of capital formation-"human
capital" formation. Time and resources are required for education and
training, just as they are in the formation of physical capital. And
education and training can increase the return to labor, just as invest-
ment increases the potential productivity of resources. To the extent
that on-the-job or in-the-field training also creates skills, it is also human
capital formation. However, while there are insightful analogies be-
tween physical and human capital formation, there are important
differences as well, in the inalienability of human labor and in the
individual and social character of the decision processes that determine
its use. These differences require caution in the use of the human
capital-physical capital analogies.16

Interaction of Technological Decisions 29
The developing nations vary widely in their stock of human capital
and in their facilities for creating additional amounts of such capital. In
some countries, formal education is an old tradition, so that it is a
familiar, if not universal, social institution. In other countries, the
experience with formal education systems is relatively new. In some
countries, artisan and household enterprises and small factory estab-
lishments serve as important training grounds for workers. In other
countries, this type of undertaking is relatively unimportant. Also,
depending on their culture, language, and relative proximity, some
developing nations can more easily and extensively use skilled workers
from other countries. Developing countries that can produce highly
qualified persons, but cannot compete in the real income offered to
these people, can be faced with "brain drain." This "export" of talent
may or may not be beneficial, depending on the returns that accrue.
That new technologies often require new labor skills is clear, but
there is so little evidence on the relations between capital intensity and
labor-skill requirements that no generalizations seem warranted. In
some cases, for example, modern and capital-intensive equipment
simplifies the labor operations by specializing the tasks, or it carries
automatic control devices that reduce the need for human skill in
regulating the equipment. Adoption of new technologies will almost
certainly change the composition of workers employed. In some situa-
tions, only a few highly skilled workers may be necessary, and the rest
of the labor force can be relatively untrained. Thus, the education and
training of labor required by new technologies is not readily evaluated,
and few studies are available.

Changes in Sectoral Composition
Development in every country certainly means some degree of indus-
trialization. It also means expansion of the power-generation sector,
the transportation and communication sectors, the construction sector,
and all the marketing and financial service sectors. It requires growth
in public services of all types: education, public health, administration,
and so on. Some patterns of sectoral change are more conducive to
rapid growth than others, especially when recourse to foreign trade is
limited. In general, rapid economic growth is associated with concen-
tration of a relatively large proportion of investable resources in the
capital-goods sectors at the early stages of the growth process. Yet this
generalization is subject to so many qualifications that it does not
provide a useful general guide to policy. Each country requires its own

separate analysis to establish its own optimal growth path. For exam-
ple, small countries cannot establish efficient capital-goods sectors as
readily as large countries, because economies of scale are important in
industries that require relatively large levels of output to be efficient.
The emphasis in development policy on the different economic sectors
implies an emphasis on technologies having different implications for
employment opportunities as well as for overall economic growth and
income distribution. It is understandable that the relative growth of the
various sectors has been one of the most controversial issues in devel-
opment policy.
In addition, sectoral interdependence arises because flows of inputs
and outputs among sectors are necessary if each sector is to deliver the
final goods and services demanded of it. To invest in fertilizer plants,
cement and steel have to be delivered to construction projects. To
produce cotton textiles, fertilizer must be delivered to farms and used
with labor and other resources to grow the cotton, which must then be
delivered to textile factories, and so on. It can be intuitively ap-
preciated that the interdependence of the sectors requires some kind of
balance to prevent major bottlenecks in the economy. Factories con-
structed to use centrally generated electric power will stand idle if the
power-generating stations are not ready when the factories are com-
pleted. That idleness means that the investment will pay off more
slowly and less well than it should have, whereas coordination of
factory and power-station construction would avoid such idleness and
Thus, sectoral interdependence makes it impossible to judge a
technology solely in terms of its direct effects on output, employment,
and so on. Because nearly every productive activity both sends and
receives goods and services from other productive activities, its indi-
rect consequences must also be taken into account in a comprehensive
One of the major development-policy controversies regarding sec-
toral balance has concerned the relative emphasis on production of
goods intended to substitute for imports or of goods intended mainly
for export. Import substitution, it has been claimed, would reduce the
developing countries' need to export primary products at increasingly
unfavorable terms of trade to purchase the manufactured goods neces-
sary for their development. Against this policy, it is argued that
because domestic markets for many manufactured goods in developing
countries are relatively small, import substitution has actually led to
the establishment of uneconomically small-scale plants.17 In addition,

Interaction of Technological Decisions 31
it is claimed that protection has permitted developing countries to use
technologies that do not take full advantage of their particular labor and
other resource supplies.18
On the other hand, it is claimed that export promotion would exploit
the comparative advantage of the developing countries in their special
natural resources and, often, relatively inexpensive labor. Exposing
the economies of the developing countries to international competition
would both permit them to buy needed goods at relatively low prices
and stimulate internal efficiency in the use of resources. Also, foreign
trade is a many-purpose balancing sector. Because foreign trade can
provide many kinds of inputs not domestically available and take many
kinds of outputs not domestically absorbable, it can make possible
large-scale and sometimes less costly production. Foreign trade can
also reduce, though not eliminate, the need for input-output balances
among sectors within a country.19
But the sectoral-balance controversy goes beyond the import
substitution-export promotion issues. Special emphasis on industrial
expansion-and within industry on the expansion of heavy manufactur-
ing industry-as the key to development progress was an early part
of development ideology. Concentration on investment-goods produc-
tion necessarily implies relatively high savings rates and lower con-
sumption rates. As a result, however, the potential levels of consump-
tion in the future would be higher. Thus, the sectoral emphasis on
investment goods or consumption goods also involves a decision about
the distribution of the benefits of development over time.
There also appear to be characteristic differences among sectors in
the opportunities for the efficient and intensive use of labor.20 If
electric power is to be generated in a central station, it is difficult to
find efficient, labor-intensive methods. Likewise, there are no known
ways of producing steel efficiently except by using a lot of plant and
equipment. In both cases, there may be more opportunities for inten-
sive use of labor in the peripheral processes such as fuel and other
material transport than there are in the core processes.21 However,
that these have a major effect on employment opportunities is yet to be
shown. On the other hand, other manufacturing sectors (such as in
some consumer goods, and certainly in agriculture, the services, and
construction sectors) appear to permit more opportunities to absorb
labor efficiently. These possibilities will be examined carefully below.
Agriculture, which was relatively neglected in the early stages of the
development policy of a number of countries, has more recently come
to be widely regarded as a premier sector. In agriculture, the technolog-

ical choice issues have focused on the emphasis that should be given to
methods that rely heavily on mechanical equipment and chemical
fertilizers and pesticides. These questions have been related to agricul-
tural organization, the concentration of land ownership, and the poten-
tial displacement of labor as a result of new technologies.

Changes in Economic Organization
Changes in economic organization are often considered one of the
major aspects and conditions of development. Some of the envisaged
changes have to do with the relative role of private markets and
government ownership and direction in the economy. But development
also involves changes in such fundamental social patterns as the role of
the individual household enterprise as compared with wage-paying
capitalist or cooperative types of enterprise. While it is difficult to
evaluate the significance of such changes, they cannot be ignored as an
essential part of the development process.
Rightly or wrongly, the lack of development in many countries has
been considered a failure of private capitalism. This reasoning is due
partly to the association of colonialism with the capitalism of the
colonizing country. In countries without a colonial history, the stagna-
tion of the past has also occurred when the dominant economic
organization was a form of private capitalism. While historical relations
are often far from conclusive, many of the developing countries have
reacted against private capitalism as part of a widespread ideological
trend. However, the government's role in the economies of the de-
veloping countries has been related not only to their economic ideology
but also to their political development. Government intervention in
economic decision making and control has enabled them to penetrate
the social systems and institutionalize authority when action with other
instruments and in other areas has been difficult to sustain.
The replacement of the private organization of production with
public or quasi-public organizations has to some degree been a feature
of the development process in many countries. Nationalization of
enterprise has been due partly to economic ideology, but partly also to
the political goals to which it contributes. Typically, the major "util-
ity" sectors of the economy, such as transportation and power, have
been nationalized and, in some countries, important parts of manufac-
turing have been nationalized as well. In some cases, the nationalized
sectors have operated directly under government ministries, but, in
other cases, new government corporations have been created.
Government operation of production enterprise, directly or indi-

Interaction of Technological Decisions 33
rectly, has encountered many problems. Staffing patterns and organi-
zations different from the civil service have been difficult to establish
successfully. The multiple objectives typically set for government or
quasi-public firms have created special difficulties in the developing
countries. A private firm is expected to maximize its profits within the
technological and demand constraints of its sector and the taxes and
controls of the government. However, public enterprises are often
subject to special constraints on the pricing of their product, the level
of their wages, and their employment conditions. The degree to which
public enterprises can adjust to their goals and constraints depends in
part on the nature of the technologies they employ.
Many of the developing countries have turned to government plan-
ning and government controls for the direction of investment alloca-
tion. Review and control processes in the economic ministries may
both direct government enterprise and regulate private enterprise. The
level of detail at which the centralized control operates varies con-
siderably among countries. Often only the sectoral composition of new
investment is controlled, and the choice of technological methods is
left to individual enterprise. But in some cases the latter is also
centrally regulated.
Other countries rely heavily on various market mechanisms to
determine the sectoral allocation of investment. Under these condi-
tions, the financial markets have a central role in development deci-
sions; they affect investment allocations and the choice of technology,
and they determine which individuals or groups will undertake particu-
lar investments.
In addition to overall policies for economic organization, there may
be special sectoral policies, as in the formation of farm cooperatives,
for example. Such cooperatives may be encouraged, when individual
farms are relatively small, to facilitate the use of technologies that are
most efficient if used on a large scale. Or cooperative organizations
may be used to achieve other advantages of large scale in purchasing
and marketing.
Finally, systematic changes occur in the most prevalent types of
economic organizations in the course of development, even when there
are no explicit government policies toward enterprise organization. In
the developing countries, much production is typically organized in
small-scale, often family, enterprise, not only in the agricultural and
service sectors but also in the early stages of manufacturing. This type
of enterprise is essentially "noncapitalist" in its use of labor and may
be relatively labor intensive since family labor need not be paid
standard wages. Thus, the technologies chosen and used in family

enterprise can, in general, be expected to be different from those
adopted by enterprises that hire and pay for labor through more explicit
wage contracts.22
Although family enterprise persists over the course of development,
it tends to become less important in one sector after another. Its
changing importance is only one example of the spread of market-
determined resource decisions in the course of development. For
example, in some developing countries agricultural land is not bought
and sold or rented in markets where it is valued only for its productiv-
ity. But these conditions change over time. This, too, has important
implications for technologies, since absence of market-determined
resource valuations will lead to different land-labor-capital combina-
tions than those determined by markets.

Social and Political Development
Economic development and the accompanying political changes are an
aspect of the modernization of the developing nations. The term
"modernization" is not intended as a pejorative characterization of the
"traditional" societies from which the developing countries are
evolving. Rather, modernization is intended to refer to the whole
complex of changes, economic and otherwise, which occur in the
course of development. These include the differentiation of institu-
tional structures to perform societal functions and changes from per-
sonalistic to more formalized structures and procedures of social
organization and decision making. The economic changes are.
perhaps, the most obvious and are most obviously associated with
technological decisions, but the effects of modernization are pervasive
and are intimately associated with the production technologies used.
Some of the clearest and most dramatic examples of the relations
between social structure and technology can be found in economies
that still rely on hunting to supplement small-scale agriculture. In these
societies, hunting is typically a male activity, and status as a male and
among males depends on success in this activity. Substituting other
food for game is a technological innovation that destroys a status-
creating activity and can impose severe social tensions on such a
society.23 While the social effects of technological changes in other
societies are often not so obvious, they may be just as profound. Such
change, however, is not necessarily destructive in all respects. For
example, it is commonly argued that in the early phases of Japanese
industrialization the conversion of the samurai warrior class to indus-
trial entrepreneurs and managers was both cause and effect of the
technological changes transforming that economy.25

Interaction of Technological Decisions 35

The changes in the distribution of work and production respon-
sibilities among family members that accompany modernization in
turn change the family structure itself. One of the major concerns of
the families of traditional societies is organization for production. This
responsibility is transformed with new ways of production so that
family structure itself undergoes major alterations.
Modernization also changes the way people associate themselves in
larger communities and particularly their politicization. The political
development accompanying modernization is at least indirectly related
to the technological changes that are a fundamental characteristic of
modernization. Specific sectoral emphases and technological innova-
tion may promote types of political developments that strengthen the
capacity of governments to interact with their citizens. For example,
concentration on large-scale industry is more consonant with centrali-
zation of political authority than decentralized development based on
rural industry. In turn, political development can contribute to
economic development by strengthening the capacity of governments
to identify and deal effectively with economic problems. Even in the
most private enterprise and capitalistic economies, government has
important economic functions to perform that will determine, to a
considerable extent, the effectiveness of the private sector. When
government participates substantially in the economy through owner-
ship, control of enterprise, and/or planning and overall direction, the
government's ability to carry out its responsibilities is even more
critical to development.25

1. For a somewhat deeper but still relatively accessible survey of development pro-
cesses see J. Bhagwati, The Economics of Underdeveloped Countries (1970).
2. The nature of the distortions that require the estimation of shadow prices are
discussed in P. Dasgupta et al. (1972) and I. Little and J. Mirrlees (1969).
3. H. Leibenstein (1966).
4. T. Schultz (1964), Chap. 3.
5. Louis Wells, "Economic Man and Engineering Man," in C. Timmer et al. (1975).
6. R. Solow (1957).
7. E. Denison (1962), (1967) and Z. Griliches and D. Jorgenson (1967).
8. For a critical statement on the problems created by technological change-and
economic growth-see, for instance, Mishan (1970).
9. E. Schumacher (1973). Part II, Chap. 5.
10. E. Hagen (1962)
11. There are few if any analysts who have explicitly taken this position. But it has
emerged from the central place given to investment policy in most development
analyses and, more to the point, in policy actions in developing countries.
12. These insights into the differential requirements for domestic and foreign saving

required for capital formation have entered the development literature primarily
under the headline of "two-gap" analysis. One gap is the possible difference
between the domestic saving and the investment desired for the development
program. The other gap is the difference between foreign-exchange resources and
import requirements. Of course, actual savings from all sources must finally match
the actual investment actually carried out, and the two gaps must both disappear.
Ex post, see H. Chenery and M. Bruno (1962).
13. For a comprehensive view of development planning models and their embodied
assumptions see C. Blitzer et al. (1975).
14. D. Turnham (1971), p. 10.
15. J. Stiglitz (1972) and J. Harris and M. Todaro (1970).
16. See, for example, R. Eckaus (1962).
17. See I. Little et al. (1970).
18. See, for example, G. Ranis (1974).
19. For the arguments for an "export-led" strategy of development see, for example,
I. Little et al. (1970).
20. These characteristic differences are at the heart of the modeling of production in
dualistic economies in A. Kelly et al. (1972).
21. See G. Ranis (1974).
22. For examples of arguments along these lines see A. Sen (1975), Part III.
23. For examples of such changes see E. Service (1962).
24. Y. Horie (1965).
25. N. Choucri (1976).

3 Alternative Criteria of
Appropriateness of

Technological Decisions

Since technological decisions affect the patterns of all development
processes, they may to some extent be consciously used as instruments
of policy to affect the course of development. Recognition of technolo-
gy's essential role and its potential as a policy instrument has been
implicit in the recently burgeoning interest in implementing "appropri-
ate" technologies. This chapter will consider the specific development
goals to which technological choices may contribute and the degree to
which these goals are complementary or competing.
The only standard for deciding the "appropriateness" of technologi-
cal decisions is reference to the general goals of development. The
qualities of technology are not more or less desirable in themselves but
only for their output potential, their corresponding input requirements,
and their effects on social and political organization. In particular,
small-scale or labor-intensive technologies are not necessarily "appro-
priate" because they are small scale or labor intensive. Whether they
are appropriate depends on their ability to contribute to development
objectives. Only if inherent features of technology dictate other pat-
terns of life and development can technology be treated as an end
rather than as a means. While this possibility must be considered, the
resolution of the issues involved should not be prejudged by adopting
this view. The alternative viewpoint is that, although technological
choices have broad economic and social implications, a technology

seldom has inevitable consequences that dictate a particular pattern for
the economy and society. Many economic and political instruments
can, in principle, be used to guide and modify the impact of technical
decisions. However, this viewpoint also requires critical examination,
since there may be many circumstances in which there are limits on the
extent to which the undesired effects of a particular technological deci-
sion can be ameliorated.
The major goals enunciated for development will be examined in this
chapter and interpreted as criteria for appropriateness of technological
decisions. Essentially, these goals and criteria are related not only to
the quantities of income and output generated over time but also to how
they are produced and distributed. As will become clear, there is some
redundancy in these goals, and even opposition. The analysis will not
be organized to avoid these features but rather to allow them to emerge
in the course of the discussion.

Maximization of Net National Output and Income
Perhaps the most conventional formulation of the national economic
objective of "doing as well as possible" is the maximization of current
national output and income. Yet the goal should not be understood in a
static sense to mean doing as well as possible only in the current year.
Rather, it nearly always implies maximizing the discounted sum of
future outputs over some time horizon or, alternatively, maximizing
output in some future target year.' This objective is attractive as a
primary goal of development because, in general, it provides the largest
sized "pie" to be divided among various uses and allocated to various
groups. There can be little doubt of the support for this general goal:
the desire for more of the material things of life-food, clothing.
shelter, and some "luxuries" beyond that-is evident in the intensity
with which economic objectives are pursued. And suppose the goal
were not being achieved. Then it would be potentially possible to
increase the amounts of goods and services available to everyone, or at
least to some groups and uses, without decreasing the availability of
goods and services in any other direction. Thus, if the goal were
abjured, the potential would not be pursued, in spite of obvious needs.
The rules for technical choice to maximize the national output and
income at any time and over time require physical and economic
efficiency in each productive establishment and in the economy as a
whole. The specific content of these rules has been worked over
intensively by economists. They are embodied in the methods of
cost-benefit analysis that can be applied to new projects and technical

Alternative Criteria 39

decisions and in the normative overall planning techniques that have
been developed. In general, neither efficiency within any enterprise
nor overall efficiency is achieved by maximizing the productivity of
any single resource in any particular use. Rather, efficient use of
resources is achieved by equalizing the incremental returns to equal
expenditures on each type of resource. If such returns are equalized,
then no reallocation of resources could possibly increase output.
Concentration on national output and income maximization as a
development goal has significant limitations, however. It avoids the
essential questions of what should be the composition and distribution
of national output, and it implicitly assumes that the national output
can be used directly or indirectly for any social purpose and distributed
according to any social rule. Yet these assumptions are not warranted.
For example, investment goods are seldom useful for consumption
purposes, products for export may well be different from those con-
sumed domestically, and political constraints may limit the scope of
income redistribution that is feasible. Thus, development goals increas-
ingly are cast in terms of more fundamental national objectives such as
achieving particular levels of consumption, patterns of income distribu-
tion, self-sufficiency, and so on.
However, even when the overall goal is not to maximize output and
income, and when the use of economic policy instruments is con-
strained, and when market or other allocation mechanisms are not fully
effective, efficient use of resources within each production establish-
ment may still be desirable.2 So caution in the sacrifice of such
efficiency is always necessary.

Maximization of Availability of Consumption Goods
Since most would agree that consumption in some form is the final goal
of economic activity, and since consumption goods are not perfectly
substitutable with other goods, it is reasonable to set maximizing
consumption as the goal for the developing countries. This goal is in
fact enshrined in many development plans and planning models as well
as in one major approach to cost-benefit analysis.3
However, the goal should not be interpreted in the narrow sense of
creating consumer enjoyment in the short run. Instead, the maximiza-
tion is intended to occur over some foreseeable future period. Thus it is
necessary to consider how present decisions about technology will
affect the production of consumption in the future as well as in the
present. Those effects in turn depend on investment and savings rates.
So the choice of appropriate technology (that is, one that would

maximize the attainment of the consumption availability goal), must
consider any possible differential effects on saving. The differential
effects may arise because alternative technologies imply different
proportions of wages, capital income, and tax payments in total in-
come. Since a future loaf of bread cannot satisfy current hunger, the
choices must also weigh the relative importance of present and future
consumption. The weighing of present and future must also reflect
foreign-trade constraints, which affect the viability of any particular
time stream of consumption.
Again, however, this goal is stated in terms of a national aggregate. It
implicitly assumes no need to worry either about the composition or
the distribution of consumption, perhaps because taxes and subsidies
can be used for particular distributional objectives. When this is not the
case, more specific development goals must be set.

Maximization of Rate of Economic Growth
One of the most frequently articulated goals of the developing coun-
tries is to increase their rate of growth or, in a more extreme version, to
grow as rapidly as possible-that is, to maximize their rate of growth.
This objective seems plausible until it is realized that growth requires
current sacrifices, and maximizing growth means maximizing those
sacrifices. The critical sacrifice is current consumption because, as
pointed out in Chapter 2, consumption must be restrained to free
resources for investment to add to production capacity to create more
investment, etc.
A more acceptable version of the objective is to maximize growth
subject to the maintenance of a minimal and, perhaps, rising level of
income or consumption. In this form, the goal becomes a version of
that articulated just previously: maximizing a combination of present
and future consumption, since the object of growth is to increase
consumption possibilities in the future.
In effect, this criterion of appropriateness of technological decisions
may only imply a different weighting of present and future output or
consumption rather than an objective completely different from the
previous two. But the significance of the different weights for the
composition of output and the choice of technology can be profound.
An emphasis on growth will place stress on technologies that help
generate a high rate of saving for investment. An emphasis on growth
also tends to be associated with relatively high interest rates, which, in
turn, works against capital-intensive technologies and for methods that
pay off rapidly.

Alternative Criteria 41

Reduction of Unemployment
As noted, the problems generating the greatest concern about the
appropriateness of technological choices have been employment and
unemployment. Despite the substantial amount of investment that has
occurred in most of the developing countries, their unemployment
problem has not been relieved. Indeed, partly because their population
growth rates and their labor force growth rates are high compared with
their growth rate of employment, unemployment is expected to in-
crease. It has been easy, though not necessarily correct, to attribute
the low growth rate of employment to the choice of "inappropriate"
technologies, either locally generated or imported from abroad. The
next step is usually a call for more labor-using technologies. However,
before taking this step it is desirable to consider the social functions of
employment in more depth.
First, employment in conjunction with other types of productive
resources generates output. To perform this function as well as possi-
ble, the appropriate choice of technology should be that which
maximizes output, taking into account the type of output desired and
the preference for present or future increases in consumption.
Second, employment is a means of distributing income. In effect,
labor is made a condition of the receipt of income. The amount of
income is associated with work time and usually with the quality of the
labor as well, where quality reflects the work's contribution to the
output's value. While the income-distribution function of employment
is not always separated from its output function, both in principle and
in actuality the distinction can be and is actively made, as, for example,
in welfare assistance and old-age-support programs.
Social recognition and personal satisfaction are other features of
employment: the recognition aspect.4 Social status and self-evaluation
are a function of the kind of work people perform, not just in relatively
traditional societies but in modern systems as well. The individual
trauma associated with involuntary unemployment in the advanced
Western economies is an example of the relation of employment and
status. Typically, the socially approved roles for individuals permit
only small and specially identified classes to be more or less perma-
nently absent from formal or informal employment without loss of
status and self-esteem. In addition, the redistribution of income
through government programs may not be a realistic option, given the
limited capacities of the government to mobilize resources and distrib-
ute them to potential recipients. It is partly for these reasons that
distribution of income through employment, even with economically

inefficient technologies, may be socially acceptable, while distribution
through dole or welfare schemes is less so. In addition, if there is
obvious inefficiency in terms of the output or growth criteria, any
burden or blame can be shifted to less easily identifiable bureaucrats
and managers.
Finally, the reverse of the income distribution function of employ-
ment is its "tax" function. Employment requires the sacrifice of leisure
and, usually, the personal direction of each individual's own activities.
If the income received from employment were totally within each
worker's disposal, the burden would be completely self-determined.
But taxes and rules for income subsidies make employment to some
degree a socially determined tax.
Because employment performs several different functions in every
society, there is no single employment or unemployment criterion for
the appropriateness of technological choices. In fact, there may well be
conflicts among employment criteria. It has been argued that when
technologies considered appropriate by income-distribution standards
lead to inefficient choices by output maximization or growth standards,
the latter criteria of appropriateness should prevail. This would result
in the largest output to distribute. Conceivably, then, any desired
pattern of income distribution might be achieved by one or another
redistribution scheme. In principle, everyone would then be better off
than if output, in the short or long run, were sacrificed for a particular
distributional goal. Yet income redistribution in the patterns desired
may not only be beyond the capacities of government agencies in
developing countries but may not be as socially acceptable as a
government subsidy. By comparison, a particular income-distribution
pattern achieved by the conscious adoption of inefficient technologies
may be more readily tolerated.
The sacrifices of output in the choice of technology to achieve
distributional goals will often be difficult to identify. When distribu-
tional goals are achieved through technologies that use more labor than
is strictly necessary, the distinction between persons "on the dole"
or "on welfare" and those "righteously working" disappears. Tech-
nologies that increase employment may also go far toward meet-
ing the standards of appropriateness by the criteria of social recogni-
tion and personal satisfaction as well as income-distribution criteria.
Employment-based criticisms of the technological choices made in the
developing countries tend to emphasize the distributional and self-
esteem functions performed by employment rather than the production
and tax aspects. What is not clear in many of the criticisms, however,
is the extent to which these latter functions of employment should be

Alternative Criteria 43
sacrificed for the former goals, nor is there even a recognition that a
trade-off may be unavoidable.5

Redistribution of Income and Wealth
There has been increasing concern in recent years about the distribu-
tional aspects of development and a growing interest in policy instru-
ments. that can affect these aspects. It is natural, therefore, that
technological decisions should be given attention in this search. Yet
any policy movement in this direction is strongly constrained by lack of
general understanding of the factors determining the personal distribu-
tion of income and, in particular, the effects of technological decisions.
For example, there is some evidence that the distribution of income in
the developing countries is more unequal than in advanced countries
but no firm comprehension of why this might be so.6 However, it has
been noted that appropriate technological decisions can affect the
distribution of income by affecting the volume of employment pro-
vided. These decisions may have potentially similar effects in the
quality of employment created, depending on whether the employment
is mainly of skilled or unskilled workers.
Distributional issues also arise in relation to the precise meaning
given to the consumption criteria for technological choice, since poten-
tial additions to the consumption of high-income groups generally are
not given the same social valuation as additions to the consumption of
low-income groups. The benefits of alternative technological choices
would, therefore, be weighted differently, depending on whether they
give employment and thus generate income mainly to low- or to
high-income groups.
The goals of maximizing aggregate or average per capital consump-
tion over some foreseeable period on the one hand, and that of
spreading consumption to specially benefit low-income groups on the
other hand, may conflict because of different savings levels associated
with technologies chosen by the different criteria.

Regional Development
Regional development goals are simply economic, political, and other
social goals specified for a particular locality. If there is any difference
between such regionally specified goals and those set forth at the
national level, it is that the regional goals may be described in terms
relative to the average of the nation as a whole or some particular
"advanced" region. For example, the growth goals for the less-

developed south Italian Mezzogiorno have been specified relative to
the highly industrialized Italian north.
The reasons for assigning special priority to particular regions are
fundamentally political, but they may emanate from an appreciation of
particularly profound differences in economic conditions and the de-
gree of participation in overall national development, as appears to be
the case with the Brazilian northeast region. However, regional de-
velopment goals may also be frankly political and involved with
conflicting international territorial claims, as in the case of the extreme
north of Chile vis-a-vis Bolivia. Or the politics of regional development
may be completely internal and related to differential support to
particular factions, parties, or ideologies.
Differential rates of regional development are responses to differen-
tial availability and use of resources, including regional differences in
the level of investment. Production methods are likely to be less
important than investment in particular types of production and loca-
tion decisions. However, because regional potentials differ, a decision
to emphasize development in a particular area will result in emphasis
on particular sectors and processes. For example, the decision to
encourage the development of an agricultural region characterized by
inadequate water supply will give special weight to all the investment
and technical decisions related to water use. If a water supply technol-
ogy requiring relatively large-scale investment by individual farmers is
adopted, then the large farmers will be favored over small farmers,
unless cooperative organizations and special financial programs for
small farmers are sponsored. Similar considerations also apply when
technological decisions are not guided primarily by regional objectives,
but when they are related to the explicit and often large-scale problems
of regional development the implications may become more obvious.
Regional development is also affected strongly by regional availabil-
ity of natural resources. The great variability in these will typically
dictate local adaptations of technology: in agriculture, mining, forestry,
and so on. In resource-based production, the local resources greatly
affect the choice of technologies appropriate by the economic criterion
of cost minimization or output maximization. However, in resource-
based production, there is no greater understanding of technologies
appropriate by criteria other than output maximization than there is in
other types of production.

Balance of Payments Relief
There are two quite opposite theoretical and policy arguments in which
technological decisions are associated with balance of payments pol-

Alternative Criteria 45

icy. One argument claims that concentration on export promotion
policies, i.e., investment in industries that expand export sales and
earnings, will cope best with the balance of payments problems that
arise in the course of development. The opposite policy would place
more emphasis on industries whose output substitutes for imports.
While the arguments are mainly about sectoral investment allocations,
it has also been argued that the different sectoral investment policies
employ different technologies as well.7
The argument for export promotion, briefly, is that for export
industries to thrive in the competitive world environment they must
reflect the real comparative advantages of the developing country.
This, in turn, can be achieved only if the technologies used in the
export industries make optimal use of available resources to minimize
costs. Thus, the criterion for the identification of appropriateness in
this connection is really the one first mentioned above: output maximi-
zation or cost minimization with available resources for any particular
level of output produced.
The argument for investment in import-substituting industries has
several themes. One is that the terms of trade are steadily worsening
for the typical primary commodity exports of developing countries and
that the way to avoid the consequences of this trend is to create
domestic industries that will produce the imported goods that had been
paid for with commodity exports. However, these industries were also
intended to use resources efficiently in the sense that output was
maximized or costs minimized for any particular output. Another
theme in the import-substitution argument is that the policy will make
developing countries more independent and self-sufficient economically,
and that end is desirable in itself.
Thus, both the export-promotion and import-substitution policies are
focused on output maximization but differ in the means chosen, except
that the latter may also be justified on nationalistic and autarkic
However, it has also been argued that actual pursuit of the export-
promotion policy has led to the choice of efficient and appropriate
technologies, while this has not been the case when import-substitution
programs are followed.8 Although import-substituting industries were
not intended to be inefficient, that often appears to have been the
outcome because the domestic markets are small in most of the
developing countries. Moreover, it has been argued that government
interference with, and distortions of, production and pricing are more
pervasive when import-substitution policies are followed, because
domestic production can be "protected" while exports are exposed to
international competition. Because of this competition, pressure to find

technologies that minimize costs will be greater in production for
export than production for import substitution.
It is well-known that government can use subsidies and other favored
treatment to influence decisions in export sectors just as much as in
import-substituting sectors. The claim that governments have foregone
the opportunities to exercise distorting influences on exports remains
unsubstantiated. However, the issues are controversial ones, and the
arguments over these two policies emphasize the need for cautious
judgments. Technological decisions and balance of payments policy
are also related by the foreign-exchange cost of importing technologies
from abroad. Some studies have indicated that license fees and patent
royalties involved in importing foreign technologies can become a
significant part of foreign-exchange requirements.9 So domestically
developed technologies or foreign technologies acquired at relatively
low cost would be the more appropriate technologies by this test alone.
To increase foreign-exchange earnings, some developing countries
that export raw materials insist on increased domestic processing of
these exports and have raised this policy to the status of a development
goal. Yet raw-materials processing must be just a means for the
achievement of more basic objectives and should be evaluated in those
terms. It may provide a means for increasing industrial investment to
increase employment, net domestic output, overall growth, or export
earnings. However, there may well be trade-offs for foreign firms
between investment in processing and in extraction facilities. Thus,
additional foreign investment may not become available as a result of
this type of policy. Domestically sponsored investment in raw-
materials processing should also face the competition of alternative
projects that may be more effective in achieving overall goals.

Promotion of Political Development and
National Political Goals
Political development has been identified in terms of the characteristic
difficulties, crises, or problems that developing societies encounter and
must resolve to become viable. These are the establishment of national
identity and recognition of the political legitimacy of the governmental
system, the increased participation of the citizenry in the political
process, the extension of government influence more broadly into
various aspects of life, and the distribution of both material and
associational values. These problems must be surmounted by every
political system if it is to survive, and the ways in which these problems
are resolved both shape and are shaped by the capabilities of govern-

Alternative Criteria 47
ments.'0 The rapid changes in the means used to deal with these
problems give a particularly urgent quality to political development.
That political problems must interact extensively with technological
decisions is apparent from the inclusive nature of the issues. Therefore,
it might be possible to define appropriate technologies according to
how much they help in the management of these problems. However,
the characteristics of technology have not been systematically as-
sociated with political development to any great extent. At this point it
is difficult to do more than use the political development problems for
rationalizing technological decisions that are otherwise difficult to
justify. Thus steel plants, even when not warranted on a narrowly
defined economic criterion, may be an appropriate technology on the
ground that they contribute to the creation of national identity. Such
"national monuments" can serve an important symbolic function and,
as such, may be a major political tool in assisting national integration.
Public-communications methods and transportation systems not
necessary to meet the economic requirements of development may be
justified in terms of the capability they create for the government to
spread its influence throughout the country. As noted above, the
technologies associated with relatively large-scale projects may be
both reflections and instruments of an intention to facilitate a "mod-
ern" centralized political development rather than a decentralized
village and rural society.
Another criticism of some technologies is that they are foreign and
thus maintain or increase national dependency upon external
capabilities.12 Although dependency has received a good deal of atten-
tion, especially in Latin America, it remains an elusive characteristic.
It is often defined tautologically or in terms of such general constraints
on national autonomy that its significance cannot be tested. Use of
imported technology does mean dependence on foreign technology in a
simple sense. But the social and political consequences are not neces-
sarily negative in the sense attributed to them by the dependency
arguments. Nonetheless, it is not surprising and must be recognized
that, with technology as with foreign trade, partial if not complete
autarky can be an important political goal.

Improvement in the "Quality of Life"
Although all the above economic and political criteria for appropriate
technologies which are standards based on technology's contribu-
tion to specific aspects of economic and political development, another
approach claims to base its standards on a different valuation of the

major features of modern societies. This alternative view places a high
value on small-scale economic activities, on rural or village life, on
local self-sufficiency, on the maintenance of a "natural" ecology, and
on employment and equalization of the distribution of income. These
themes are, perhaps, most clearly identified with Mahatma Gandhi.
However, most of the elements have a wide range of advocates,
historical as well as contemporary. Among the most well-known of the
latter is E. F. Schumacher, who has popularized the notion of an
intermediate technology:
It is vastly superior to the primitive technology of bygone ages but at the same time much
simpler, cheaper, and freer than the super technology of the rich. . .3
The intermediate technology would also fit much more smoothly into the relatively
unsophisticated environment in which it is to be utilized. The equipment would be fairly
simple and therefore understandable, suitable for maintenance and repair on the spot ...
far less dependent on raw materials of great purity or exact specifications and much more
adaptable to market fluctuations than highly sophisticated equipment. Men are more
easily trained; supervision, control and organization are simpler and there is far less
vulnerability to unforeseen difficulties.14
These specifications are less criteria for choice among existing
technologies than aspirations for a new set of technologies. The
hoped-for intermediate technologies would not sacrifice anything in
terms of output and income and yet would employ more labor, use less
of other resources, and so on. It is far from clear that such technologies
can ever be created for a broad range of production activities. This
approach to technological choice does emphasize the many disadvan-
tages of small farmers and artisans relative to larger producers who
have greater economic and political power. The search for intermediate
technologies has been justified in part by the desire to reduce these
disadvantages. The technological remedy proposed would essentially
permit a withdrawal or at least provide protection from market interac-
tions by permitting a greater degree of self-sufficiency. However, the
existence of market imperfections and political disadvantages does not
necessarily imply that technologies that compensate for the imperfec-
tions should be sought and adopted. For example, a criterion proposed
for village-level technologies is that the cost of the required capital
equipment should be low enough for a single farmer or artisan to afford
it. This criterion is justified by the customary shortage or unavailability
of loans necessary for larger-scale technologies. In effect, a financial
market imperfection-the unavailability of loans to small farmers at
market interest rates-is used as the rationale for a particular type of
technology. But that implies adjusting to, and preserving the effects of,
the financial market imperfection. It would be better economic and

Alternative Criteria 49
social policy to attempt to eliminate the market imperfections them-
selves instead of trying to create techniques that preserve the effects of
such imperfections.
Although the Intermediate Technology Group and other organiza-
tions claim some success in finding technologies that satisfy their
requirements, their examples appear to be clustered in a few agricultural
activities and small manufacturing activities outside the major produc-
tion processes of both sectors. In the absence of cost-benefit analyses
for new intermediate technologies, it must be concluded that for the
major processes in mineral transformation, power generation, and the
metal and electrical industries-i.e., much of the stuff of modern
life-village-level intermediate technologies remain unfulfilled goals.
Perhaps a Gandhi or Schumacher would respond, first, that insufficient
effort has been put into the creation of intermediate technologies to
judge the potential. Yet it is not correct to argue that there have been
no incentives to create such technologies. Moreover, it should not be
expected that modern science is necessarily capable of doing anything,
even developing an intermediate technology for producing stainless
steel. A second reaction would be that man should learn to live without
so much stainless steel and the other commodities of modern life that
are so demanding of resources and that, to a considerable degree, impel
production into large-scale, low-labor intensity, agglomerated indus-
trial complexes. But this is a matter of choice on which the overwhelm-
ing evidence is a preference for modernity.
Thus, the Gandhian-Schumacher prescription for appropriate
technology is, at this point, mainly a desire for a life-style focused
around villages and a relatively simple, rural existence compared to the
mass production-consumption, urbanized life-style more typical of the
relatively developed countries. There is no criterion for economic
choice involved because the intermediate technologists promise more
of everything except what they consider undesirable in life. The
arguments are not in themselves conclusive, and, if research does
generate intermediate technology alternatives, choice will finally be
one of individual taste. About matters of taste, there is, finally, no
These two critical aspects of the Gandhian-Schumacher prescrip-
tions for appropriate technology-(1) that it is at present a research
agenda rather than a practical set of alternatives, and (2) that if the
research were successful, it would be a prescription for a different
life-style-are typically obscured in debate about the characteristics
and implications of intermediate technologies that do not now exist.
Nonetheless, there can be little objection to the consideration of adding

the intermediate technologies to the technical-research agenda for
developing countries. If intermediate technologies having the charac-
teristics described above can be found, they would improve the eco-
nomic conditions of life for the majority of persons in most of the
less-developed countries who are still living in rural villages. However,
it is not certain that everything called an intermediate technology
would have that effect. The economic outcome depends on the relative
effectiveness with which such technologies would use resources com-
pared to other techniques. Beyond that, their acceptance and use
depend on other relevant criteria and the preferences for the life-styles
they make possible. But there is really not much point in debating at
length the merits of intermediate technologies that do not now exist and
whose contrivance remains to be demonstrated.

Complementarity and Competitiveness of Alternative Criteria
for Appropriate Technologies
To decide what is an appropriate technology, it is necessary to have a
criterion. Since particular technologies are not ends in themselves but
means for achieving economic, political, and other social goals, the
criteria for appropriateness must be found in the goals of development.
Yet there are many different development goals, and of these some are
alternative and competitive rather than complementary. For example,
maximization of output and income in any particular year may not
produce the fastest rate of overall growth over a longer period. Increas-
ing employment and reducing unemployment may or may not be compat-
ible with either output and income maximization or maximizing the
overall growth rate. Providing employment may lessen or increase
inequality in the distribution of wealth and income, depending on the
sectors in which employment is created. Employment creation in the
short run may well compete with employment creation over a longer
period if the former results in the sacrifice of investment that would
create employment opportunities in the future.
Extending the regional distribution of development may also be
consistent with all the other goals but, on the other hand, may be a
competitive objective, depending on the reasons for the regional in-
equalities and the means used to reduce them. For example, if
technologies chosen on other grounds create balance of payments
problems, then certainly sacrifice in some other dimension will be
necessary to help redress the balance of payments.
With respect to the relations between political development and
other goals achievable in part by appropriate technological decisions,

Alternative Criteria 51
few generalizations appear warranted. Political criteria that emphasize
the political symbolism and penetration or social mobilization values of
projects may conflict with economic criteria. Certain types of eco-
nomic organization and technologies may contribute more specifically
to centralized political development goals than others. But, for the
most part, the relations remain obscure, even though the potential for
the relationships is clear.
Finally, intermediate technologies stressing small-scale, relatively
self-sufficient production and village, if not rural, life are competitive
with social and economic goals associated with urbanization. How-
ever, the reality of the choice cannot now be assessed. There are
undoubtedly economically efficient technologies in some sectors that
are intermediate in the sense that they employ capital and labor in
proportions different from those used by modern large-scale industry.
However, it is by no means clear that even these satisfy the Gandhian
goals. Of those specific technologies proposed for these latter purposes,
few, if any, have been carefully assessed with respect to either their
narrowly economic or more broadly social characteristics.
When goals and criteria are consistent and complementary,
technological decisions are relatively straightforward. But what can be
done when the goals of development and criteria for decisions about
appropriate technology are inconsistent and competing? That is a
complexity of reality that cannot be escaped and for which there is no
analytical resolution. For example, suppose it is conclusively demon-
strated that some techniques are more labor intensive, yield as high a
return on capital as alternative methods, but generate lower savings.
This would imply a lesser contribution to further investment and
growth. The different goals are competing and must, in effect, be
valued against each other. An essentially political decision is required
in determining the extent to which one goal will be pursued at the
sacrifice of others.
Moreover, to determine which criteria of appropriateness will be
observed or the relative weights each will have, it is not sufficient to
make an initial decision. When there is competition among the criteria,
the competition will continue through the lifetime of the project. In
particular, cost minimization is the only criterion that is surely consis-
tent with the survival of the private enterprise that employs it. That is
true as well of public enterprise required to meet a market test. The
technical decisions of a government-operated railroad system, for
example, can be made primarily to provide employment, rather than to
minimize costs. But suppose also that the government system has to
compete with a private road-trucking system. Since the government

railroads do not minimize costs, they would lose business to private
trucks, and only recurring government subsidies would compensate for
the employment-intensive but high-cost railroad technology.
Enterprises that must directly or indirectly meet market tests of
performance and do not use cost-minimizing technology will require
continuing government intervention to guarantee their survival. Thus,
before any criterion of appropriateness other than cost minimization is
actually applied, the government's ability and willingness to provide
the continuing support required by the technology should be appraised.
Choice is easy; survival is more difficult.

1. For an example of planning to achieve future targets see Third Five Year Plan,
Government of India Planning Commission.
2. See P. Diamond and J. Mirrlees (1971).
3. See, for example, R. Eckaus and K. Parikh (1968), Chap. 1, and P. Dasgupta
et al. (1972).
4. This terminology is due to the elucidating discussion of A. Sen (1975), Chap. 1.
5. E. Schumacher (1973), passim.
6. See Chenery et al. (1975), Chap. 1.
7. See G. Ranis (1974) and H. Hughes (1974).
8. For a review of the debate that favors export promotion see Little et al. (1970).
9. For a review of studies of this question see United Nations Conference on Trade
and Development (1975). pp. 57-59.
10. For a discussion of these characteristic political development crises, see L.
Binder (1971).
11. See N. Choucri (1976).
12. See, for example, 0. Sunkel (1973).
13. E. Schumacher (1973), p. 145.
14. Ibid., p. 176.

4 Technological Opportunities
and Transfer of

Technical Information

In attempting to make and implement optimal decisions with respect to
appropriate technologies, the acquisition and transfer of information
about technological alternatives is, in itself, one of the foremost
difficulties. While in many cases there are obviously a number of
alternative sources and channels, in other cases there is only one or a
few. And in most cases, information about the costs and returns of the
alternatives is seldom easily ascertainable. This chapter will survey the
state of information with respect to technological alternatives and the
methods of transferring technological knowledge, to provide back-
ground for the subsequent examination of the various processes of, and
influences on, decision making.

Range of Technological Alternatives from which Choices
Can Be Made
If only one production method for any particular type and quantity of
output were physically efficient, then the problems of technological
choice would be greatly simplified, though not eliminated. With no
alternatives, there would be no need to consider criteria for choice
unless physical efficiency were willingly sacrificed. Even so, by
considering diverse products, each requiring a single but different set
of inputs, it might still be possible to achieve a range of choice among

alternative resource-input combinations. Assessment of this last pos-
sibility requires information not only about technology but also about
demand patterns. Choosing an output mix to get an appropriate input
mix is pointless if the outputs cannot be matched to demands.
It has been hypothesized that in many of the important economic
sectors there is, in fact, only one, or at most a few, alternative methods
for carrying out the central production processes and that all methods
require inputs in much the same proportions.' Such a limited range of
alternatives, it has been suggested, does not permit the physically
efficient employment of all the labor that is available in some develop-
ing countries because of their limited amount of capital and other
resources. If this strong hypothesis about technology and labor and
other resource availabilities were correct, then some unemployed labor
simply could not be usefully put to work to produce additional output.
A variation of this single-technique hypothesis applies not to the entire
economy but to some particular sectors, usually utilities and manufac-
turing, or at least the heavy industry part of it. The single-technology
view appears most prevalent among engineers who think in terms of a
best-practice technique that is virtually independent of wages and
prices of other resources.
Even if only a single efficient combination of labor and other
resources could be used to produce any output, it would still be
possible to employ labor or other resources more intensively by not
using them in a physically efficient way. Where one man could do the
job, two or more could be employed. Labor and other resources can be
used in many inefficient ways, and in certain conditions inefficient use
of labor might be desirable if that were the only way to achieve some
other goal such as, for example, income redistribution. Such a con-
scious decision should, however, be distinguished from the many
instances in which inefficiency is not recognized and not desired.
The more conventional view of technology, within the economics
profession at least, is that many different combinations of inputs can be
used to produce more or less the same output. This production
hypothesis implies that one input can generally be substituted for
another over a wide range of input proportions to achieve any
particular level of output. If this view of technology is correct, then the
technology, though still having great economic influence, does not so
completely determine the quantities of resources used in production as
in the single-technique hypothesis. Moreover, if the conventional
hypothesis is correct, a different explanation of unemployment in
developing countries is implied. Aside from the problems of effective
demand, unemployment must be the result of rigidities in the wage-

Technological Opportunities 55
payment mechanisms, such as minimum-wage laws imposed by gov-
ernment, union wage agreements, or by some other departures from
the competitive markets. Otherwise, labor could always be substituted
productively for other resources.
Clearly, to understand and to make policy in developing countries, it
is important to know the range of substitution possibilities and the
relative trade-offs. As a result, substantial attention has been devoted
to empirical investigations of the issues. These investigations are
unusually difficult and often costly. Detailed information on the use of
resources exists only at the plant level, and even there it is often not
conveniently organized. Manufacturing census data are usually too
aggregated and incomplete to reveal the type of technological informa-
tion required, so that special case studies and surveys are necessary.
However, detailed production data is often difficult to obtain because
of proprietary interests. Therefore, it becomes necessary to resort to
indirect and less-penetrating methods of investigation. As a result, the
studies that have been undertaken are, unfortunately, inconclusive,
although information is slowly accumulating about particular sectors
and production processes.
One kind of empirical investigation of the opportunities for
technological choice has relied on statistical information from censuses
or surveys about the resources actually used in production. The hope
has been that inferences about technology could be drawn from
information about resources used. Yet the information often includes
only the amount of resources used, employment and wages paid to
labor, the profit share, and, perhaps, estimates of capital stock.
Typically, these sources also contain only highly aggregated informa-
tion on products that are not strictly identical. The data may cover
different periods of time during which significant technological im-
provements may have been implemented, although at uneven rates. As
a result, much ingenuity and many restrictive assumptions have been
necessary to produce any measures of the potentiality for substituting
inputs for each other, typically capital for labor or vice versa. Despite
the ingenuity and because of the inadequacies of the data, the results
are inconclusive. Some studies seem to show only a restricted range of
alternatives, while other studies indicate a wide range of easy substitu-
tion possibilities between capital and labor as potentially the most
important trade-off. Furthermore, none of the findings appears to be
"robust"; apparently minor changes in estimating procedures will lead
to substantial differences in the results.3
Another type of attempt to establish the range of available technolog-
ical choices has involved detailed case studies of particular types of

production processes. In some cases, the study has been made at a
highly disaggregated level and has proceeded by isolating subprocesses
within the overall production of a particular commodity. In other
cases, the study has not tried to penetrate below the level of what goes
into and comes out of the factory or farm. One especially assiduous
technique requires visiting and surveying individual plants. The case-
study methods are relatively slow and expensive, though the results are
often not only insightful for the the specific methods and products but
also suggestive with respect to a wider range of technologies.4 By their
very nature, however, the case studies produce piecemeal and incon-
clusive evidence. In some cases, the advantages of labor-intensive
techniques appear to have been demonstrated; in other studies, the
most labor-intensive techniques do not appear to be the most appropri-
ate for minimizing the cost of production. An interesting suggestion
from some investigations is that, even if there are only a few efficient
technological alternatives in the central manufacturing processes,
significant substitution possibilities exist in the "peripheral" materials
handling and transport functions.5
A major problem in interpreting the typical case study of particular
establishments and industries is the difficulty of determining from the
available evidence whether the observed alternative input ratios are all
efficient in the least-cost sense. As noted above, apparent differences
in input proportions for the same type and quantity of output can be
generated simply through inefficiency. Presumably, few establish-
ments set out to use resources inefficiently, but many of them might be
doing so, as suggested by the previously noted wide and prevalent
differences between actual and best-practice techniques.
While it is important to know whether the methods observed through
case studies reflect efficient resource use, such information can only be
obtained by comparisons beyond the scope of such studies. The
problem typically is avoided by the implicit or explicit assumption that
the production establishments studied are participants in perfectly
competitive product and resource markets. If this were the case, then
the observed labor and other intensity resources could be considered to
have "passed the market test" and, therefore, to be efficient. How-
ever, the assumption is seldom if ever tested explicitly, so the conclu-
sions of the case studies about efficient alternatives must often be
accepted only with reservation.
The economies of scale characterizing many industrial processes are
a boon because of the relatively low cost they make possible. But they
also create problems, because developing countries are typically less
able than industrialized countries to take advantage of economies of

Technological Opportunities 57
scale. The lower costs of output often associated with larger plants can
be achieved only if the plants are operating at near capacity. This
means that many developing countries with small markets due to their
relatively small populations and/or low per capital incomes cannot
achieve the lowest existing production costs if plants are built to serve
the domestic markets alone. If national self-sufficiency is set as a policy
goal nonetheless, economies of scale mean that developing countries
with relatively small plants will bear higher real costs of production.
International trade provides a potential method of using the capacity
of large plants through exploitation of foreign markets. However, costs
and time delays in penetrating foreign markets make it difficult for the
newly industrializing countries to displace the production of advanced
If international financial markets were more perfect, the developing
countries would not be at a disadvantage just because of the large
amounts of financing needed to create firms that can achieve the
significant economies of scale. However, financial markets are not
perfect, and large loans to finance large plants in small countries may
be considered particularly risky. For these reasons, large size and the
consequent large amounts of funds required can, in itself, be a barrier to
low-cost production by developing countries.
The evidence is not conclusive for one feature often attributed to
production processes characterized by economies of scale. That feature
is the argument that such processes are also relatively capital inten-
sive rather than labor intensive and, on this ground, unsuited for de-
veloping countries. What typically captures the eye and generates
such observations about large plants is the scale of plant and equip-
ment, not the relative amounts of factors used in efficient production,
which is the relevant issue. It may be suggested that output levels
rivaling those of large-scale plants could be achieved by replicating
more obviously labor-intensive methods. However, observations of
the naked eye will not guarantee that replicated labor-intensive
methods are as efficient as large-scale methods. The latter may
conceivably be the most labor intensive of the efficient techniques that
are available. Unfortunately, the lack of detailed technical knowledge
again impedes a more definitive statement.
In studies of the potential range of input substitution, special
attention has been given to the agricultural sector. Here there are
dramatic examples of diversity in the relative intensity with which
various inputs can be used: in labor and machinery use, in fertilizer, in
irrigation water, in land leveling and improvements, and so on. The
existence of an important range of technological alternatives would, in

fact, appear obvious. The question of whether the observed alterna-
tives all use resources efficiently is typically answered by pointing to
evidence suggesting that farmers respond to the incentives of output
prices, even in traditional peasant agriculture, and thus that the
markets in which they operate are competitive.6 This evidence, how-
ever, is not definitive. Responsiveness to price is a necessary, but far
from sufficient, condition for the existence of reasonably competitive
markets, and there is other evidence of less than the full participation in
markets required for economic efficiency.7
Another approach to studying technologies is to carefully build up
the necessary information from engineering sources. In a variation of
this approach, engineering and economic information is accumulated to
design "synthetic" plants. The research methods are potentially
among the most powerful, since they require fewer assumptions about
the physical efficiency of observed technologies. However, the re-
search approach generates contrived production methods that may not
have faced a market test. The evidence generated by this
"engineering-economic" approach is again equivocal. A wide range of
efficient substitution possibilities appears to exist in some cases and
not in other cases.8
This survey indicates that the existing information on the range of
economically efficient technologies is not conclusive. But some impres-
sions do emerge. For example, in the light-manufacturing and agricul-
tural sectors in which most effort has been concentrated, the results
seem typically to suggest that, by the criterion of employment, there is
clearly a range of alternative technologies in many sectors. Most of the
case studies of particular industries demonstrate this, as does much,
though not all, of the evidence based on statistical investigations of
census or other data collections. The latter type of evidence may, to
some extent, reflect differences in labor absorption possibilities in the
manufacture of somewhat different products or similar products of
somewhat different qualities. As noted, it has generally been difficult in
the statistical investigations to ensure that the information used per-
tains to only a precisely identified product.
Yet there should be no surprise that the empirical tests show wide
ranges of employment possibilities. The studies of the substantial
differences between actual and best-practice techniques have often
focused on the differences in labor productivity among firms. That
implies differences in potential labor absorption, but not necessarily
that the alternatives are either physically or economically efficient.
Virtually all the empirical studies of technological alternatives have
focused on inputs and their relationship to output. To the extent that

Technological Opportunities 59
the other criteria of technological appropriateness discussed in Chapter
4 are different from efficiency or employment standards, there is little
evidence about the existence of a range of alternatives.9 Examples in
the handbooks on cost-benefit analysis that use the criteria of eco-
nomic growth or the maximization of the availability of consumption
goods indicate that by these standards there are differences among
projects.10 But the examples are not usually directed toward the
analysis of different methods of producing the same product, nor are
they actually (nor intended to be) surveys of the range of technological
A great number of "intermediate" village-level technologies have
been proposed; however, for the most part, these proposals consist of a
specific design intended to demonstrate technical feasibility. That is,
the proposals characterize the technologies in terms of required inputs
and related outputs. Only a few of the proposed intermediate
technologies include an economic cost-benefit analysis, and virtually
none have been analyzed according to other criteria of appropriate-
For the goal of income distribution there appears to be no systemati-
cally accumulated and analyzed evidence on technology other than
those effects that operate via employment. There may be a number of
other relations between technology and income distribution such as the
potentially important differential skill requirements of different
technologies. There may also be systematic differences in the average
age of the labor force using new technologies as compared to the labor
force using older technologies, and that would have distributional
effects.12 However, no systematic information of this sort is now
As noted previously, correcting the balance of payments is typically
not seen as a goal distinct from the goals of output maximization or cost
minimization for particular output levels or overall growth. Nonethe-
less, it has been argued that different technologies have been employed
in the pursuit of policies to promote exports as compared to those
designed to substitute for imports. This, it is claimed, is the result of
differences in effective incentives.
The fact that technology influences political development has also
been recognized. It is easy to fall into the belief that modern
technologies of communications and control make it easier for govern-
ing elites to establish their influence at all levels of their societies. Yet,
such influence has been achieved even in societies with only relatively
primitive technologies,13 but these have been traditional rather than
modernizing societies. It has also been pointed out that modern

technologies increase the effectiveness of contention by dissenting
groups within a society.14 The possible relations between the degree of
emphasis on large-scale industry and its associated technologies and
modern political development have been discerned but not detailed.
So, again, few generalizations about the relations of particular
technologies to political development now appear to be warranted.
Finally, choosing technologies to improve the quality of life raises
not only the question of different tastes in life-styles but also questions
about the objective relationships between particular technologies and
specific aspects of the quality of life. It is the latter set of issues that
will be taken up here. It is clear that for many products smaller-scale,
less-polluting and more labor-intensive technologies can be used in-
stead of those now widely employed. What is almost totally lacking in
the recommendation of such technologies is an overall appreciation of
the implications of their widespread adoption. For example, animal-
drawn conveyances are smaller scale, more labor intensive, do not
emit sulfur or nitrite pollutants, do make use of "renewable" resources,
and are more "natural" than vehicles run by motors dependent on
fossil fuels as energy sources. Yet the large-scale use of horse-drawn
vehicles was a major source of urban pollution before the advent of the
automobile, and large concentrations of farm animals are a continuing
water-supply-pollution problem in some rural areas. Similarly, ther-
moelectric central generating stations can be obvious and concentrated
sources of pollution. Yet, they do burn coal and oil more efficiently
than the small boilers that would be used in small-scale and more
dispersed generating stations or steam-producing units.
With respect to the potential for technological substitution, the
advocates of village-level intermediate technologies promise that major
research efforts will increase the range of techniques that can be used
with "reasonable" efficiency to improve the quality of life. That
remains to be demonstrated.
As a research agenda, the village-level intermediate-technology ap-
proach is not substantially different from the motivation of much of the
work now being done in many places. Many of the developing areas
have their own scientific and engineering research institutes engaged in
efforts to find technologies better suited to the local resources, and
there are a few international institutions of this type. They are attempt-
ing to extend the range of known technologies into unknown territory,
to discover more appropriate methods. The Agricultural Engineering
Department of the International Rice Research Institute provides an
outstanding example of organized efforts to develop appropriate me-
chanical technologies for rice cultivation in Asia. The standard of

Technological Opportunities 61
appropriateness in this and similar institutions appears to be lower cost
or larger output for the same cost. Only seldom is another standard
used to evaluate the new methods generated in the various research
organizations. That is primarily because, for new technologies as well
as old, it is difficult to establish clear relationships between a particular
technology and the other standards of appropriateness that have been

Sources and Costs
Information in developing countries about technologies may have its
source in new research and development, transfer of knowledge from
other areas, and/or adaptation from existing methods.
Information about existing technologies can be transferred from a
number of sources through a variety of channels.15 Technical informa-
tion can be obtained via sources that are virtually "free" in the sense
that little, if any, cost is involved in obtaining access. This is essentially
the case for the information found in technical books and journals.
But using that information requires a processorr," an engineer or
scientist who understands the literature. Since trained scientists and
engineers are relatively scarce in the less-developed countries, this
source is less readily accessible than in the industrialized countries.
However, once someone with the appropriate training is available, the
information contained on the printed page is available with only small
charges up to the "processing capacity" of the technical interpreter.
There are other sources of free information in addition to the printed
page. For example, the knowledge that a particular type of production
is feasible in a developing country is free and cannot be hidden if there
is any production at all. While obvious, it can be quite important.
Educational institutions often provide information which, if not free, is
of relatively low cost. Casual conversations and structured technical
meetings will similarly be occasions for virtually free exchange of
How much technological knowledge is freely available or at low
cost, and how significant is it? How much of that is general background
information, and how much is "hard" technical data? There is no basis
for arriving at an assessment.
Domestic research and development is carried out in both private
and public institutions in developing countries, but typically on a small
scale except for agricultural research. The accomplishments of the
industrial research and development organizations, while apparently
numerous, have seldom been evaluated for their significance.16

Another repository of a great deal of technological knowledge is
people. The scientists and engineers who learn formally in the schools
and universities are perhaps the most obvious examples. But the
foremen and craftsmen and even the unskilled laborers "carry" a good
deal of detailed information, ranging from specifically technical
methods to general knowledge of work organization. It may be ac-
quired "on the job" or in "vocational" courses. Some is job specific,
but some is to a degree transferable. Formal education is acquired in
many of the developing countries by sending students abroad, particu-
larly at the university level. Internal mobility is also a way of spreading
technological knowledge within a country. However, in developing
countries the kind of technological knowledge required by industrial
foremen and lower-level workers is not easily acquired. Such workers
have few opportunities to travel and gain experience abroad and only
limited opportunities to learn from experience in their own countries
because of the relatively small amount of industrial employment.
In many cases, the necessary knowledge for a production process is
protected by foreign patents that are protected by international patent
agreements. In this case, to use the patented information the enterprise
in the developing country must pay royalties or license fees. These fees
are determined in bargaining processes in which the representatives of
the developing country are likely to be at a disadvantage in facing the
monopoly power of technology sellers. The results are often
agreements that restrict how the technologies are used and/or where
the output is sold.
Quite apart from patented technology, some information is available
only with foreign participation of some kind, often the direct participa-
tion of a multinational corporation. These organizations, which have
attracted a great deal of attention in recent years, are undoubtedly
carriers of much information, some of it protected by patents, some of
it unpatented know-how embodying the technical and managerial ex-
pertise accumulated in the organization. Given the special international
operating capabilities of the multinational organizations, they have
been able to exercise important types of influence in many of the
developing countries. Assessment of their decision-making powers and
their influence will be discussed in the next section.
As noted, much of modern technology is embodied in particular
machines. The machines themselves may not even be patented if they
operate on well-known principles. But to have the production
capacities these machines can provide, it is necessary to have the
machines, and those may be available in turn only from a restricted
range of producers, perhaps all of them abroad. Even when the general

Technological Opportunities 63
principles on which the machines operate are well-known, there are
likely to be many specific designs to resolve specific problems, and
these designs may be best obtained from experienced producers.
That developing improved methods by adapting technology is wide-
spread and important can hardly be doubted both from anecdotal
evidence and systematic case studies. The classic "learning-by-doing"
example of the Horndal Iron Works in Sweden is frequently cited.17
But the improvisation of the Punjabi mechanics and engineers is
equally classic.'8 It might appear reasonable that the adaptation pro-
cesses in production establishments would be especially subject to the
incentives of local wages, capital, and material costs. By this reason-
ing, the adaptation process should generate methods especially likely
to "minimize" cost by more intensively using relatively abundant
resources.19 There is evidence, again anecdotal, that this happens.20
Yet it need not follow that adaptation always moves toward increased
use of the relatively abundant factors such as labor in developing
countries.21 Engineers should be interested in cost saving whenever
and wherever possible. There is little evidence that adaptation in devel-
oping countries is more likely when it results in the use of more labor
and less capital, rather than the reverse, or when the result is the use
of less of both.
This brief review of the alternative sources of technological informa-
tion suggests the difficulties of arriving at an overall assessment of the
importance of any one channel, though evaluations have been under-
taken with respect to specific technologies.22 There are no statistics of
knowledge flow; there are no measures of the relative importance of
production based on indigenously generated information compared to
foreign information. There is no way of measuring the significance of
knowledge obtained from books and journals, or knowledge gained
through the education abroad of scientists and engineers, or knowledge
embodied in purchased equipment or in patents or other special pro-
duction techniques licensed from abroad, or knowledge brought to the
developing countries by multinational corporations. Thus, we are left
with only general impressions of the importance of foreign technologi-
cal knowledge. For example, agriculture, which is still by far the
largest sector in most of the developing countries, is still the sector
least penetrated by multinational corporations in most of these coun-
tries and, probably, the sector that generates the smallest amount of
patent fees or royalty payments. Only a limited amount of farm
equipment is used in most developing countries, and that equipment is
usually not technologically sophisticated enough to command impor-
tant royalty fees. That may, however, not be true for the fertilizer and

insecticide plants, which will become increasingly important in supply-
ing the agricultural sectors of the developing countries. And biological
as well as mechanical innovations have been of enormous importance
in many of the less-developed countries.
With some exceptions, such as transportation and power generation,
the service sectors are not major importers of foreign technologies. It is
in manufacturing and mining sectors where foreign technologies em-
bodied in equipment or paid for by patent fees and royalties are likely
to be the most important. Since these are also likely to be among the
fastest-growing sectors in the course of development, their expendi-
tures for knowledge will be a growing share of total costs of acquiring
knowledge in the developing countries.
Unfortunately, there is no organized evidence and little conventional
wisdom on either the potential for substituting different sources of
knowledge or the costs of the various sources. While monopoly power
controls some knowledge flows, the significance of such power cannot
be fully assessed for any developing country.

Role of Engineering Education in Generation and
Transfer of Technology
It has been argued that one of the major sources of bias in the
generation and choice of technologies used in developing countries is
the character of the education received by their engineers. Many of the
engineers in developing areas still receive their education in the univer-
sities of the industrialized countries. Moreover, the engineering
schools in the developing countries, because of the sources of the
training of their faculties and the technical assistance received in their
establishment, are often strongly influenced by the intellectual patterns
of the industrialized countries. These conditions, it has been argued,
bias engineering training in favor of the technological methods of the
advanced countries.23 It is claimed, therefore, that engineers from
developing countries trained in these molds receive "inappropriate"
These are plausible arguments, but, as yet, there have been few
penetrating studies of the issues. Presumably, engineering educa-
tion in the industrialized countries focuses on the methods and ap-
proaches relevant to those countries. It would be reasonable to expect,
then, that engineers in industrialized countries come to prefer those
technologies that economize on the relatively scarce labor of those
countries and use more capital more intensively. But it would be hard
to document this expectation or its effects with any details. It is easier

Technological Opportunities 65

to establish that engineering education in industrialized countries em-
phasizes the sectors important for them, which include such "high-
technology" areas as petrochemical processing and electronics.
The argument that the "output mix" of engineers from the univer-
sities in the industrialized countries, in terms of their specializations,
does not match the requirements of the developing countries is, prima
facie, more persuasive. Yet, again, generalizations are dangerous. The
various fields must be examined separately to assess their appropriate-
ness. Aerospace technology and design of high-speed digital com-
puters, which attract a substantial number of engineering students in
industrialized countries, are of little direct relevance to the less-
developed countries. But nuclear power engineering, though among
the most sophisticated of modern engineering fields, may also be the
most relevant type of education for design and manning of the central
power stations in some of the less-developed countries. Similarly,
engineering education in petrochemical technology, though not rele-
vant for many of the less-developed countries, is highly appropriate for
The limited opportunity in developing countries for engineers trained
for high-technology sectors has been regarded as a major source of the
"brain drains" that have occurred, and these, in turn, have been
regarded as another bit of evidence of the inappropriateness of en-
gineering education. Yet this may represent a mistaken identification of
cause and effect. Students from less-developed countries who enter
fields with limited applicability in their homelands may, in fact, be
responding to international demands for engineers rather than to their
own nations' demands. The engineering institutions of industrialized
countries may be only a convenient scapegoat for the inclinations of
engineers from developing countries to work where they can earn the
highest incomes. Moreover, it is conceivable that the "investment in
human capital" in departing specialists provides, for some developing
countries, good returns in the form of emigrant remittances.

1. See, for example, R. Eckaus (1955).
2. A. Kelley et al. (1972), p. 25.
3. For a review of the estimates of the elasticity of substitution see D. Morawetz
(1975) and J. Gaude, "Capital-Labor Substitution Possibilities," in A. Bhalla
4. Two collections of such case studies are A. Bhalla (1975) and C. Timmer et al.


5. G. Ranis (1974).
6. T. Schultz (1964), Chap. 3.
7. The anthropological study most commonly cited in support of the "poor but
efficient" characterization of peasant agriculture, itself describes the limited
participation in resource markets. See S. Tax (1953).
8. V. Smith (1961) and W. Leontief (1953).
9. For example, even one of the most comprehensive analysts of intermediate
technologies considers only employment, output, and trade aspects and neglects
growth considerations. See G. Ranis (1974).
10. See P. Dasgupta et al. (1972).
11. A survey of more than 100 intermediate technologies that have been described
in relatively readily available literature indicated that only 24 of these had any
economic information associated with them, and for only a few was it indicated
that any sort of cost-benefit analysis had been done. No analysis was presented
for any of them of any qualities other than their output and employment char-
12. See R. Nelson and V. Norman (1973).
13. See, for example, K. Wittfogel (1957), p. 54.
14. See J. La Palombara in Binder et al. (1971).
15. See F. Machlup (1967), Chap. 8.
16. For example, "Both the arguments and the evidence for an active policy of
supporting the establishment of an industrial R & D effort in an LDC continue to
be sketchy." R. Nelson (1974), p. 75.
17. K. Arrow (1962).
18. E. Staley and R. Morse (1965), p. 179.
19. C. Kennedy (1964).
20. N. Leff (1968), Chap. 3 and 4.
21. P. Samuelson (1965).
22. For example, L. Nasbeth and G. Ray (1974).
23. F. Sagasti (1973).

5 Determinants of
Technological Decisions


Their Appropriateness

The decisions about technologies actually used in the developing
countries have been made within many different institutional
frameworks and are subject to a variety of incentives. In some cases,
though implicit objectives are being pursued, no explicit criterion of
appropriateness governs the decision. In other cases, the objectives of
the technological decision are quite clearly stated. To understand and
help implement appropriate choices of technology, it is necessary to
identify and assess the processes through which decisions are actually
made. The purpose of this chapter is to present an analytical review of
these processes. First, the role and consequences of private decision
making by enterprise under market influences will be evaluated. Next,
the impact of government policies on private enterprise and operating
directly through public enterprise will be discussed. Finally, the inter-
national influences operating through private multinational corpora-
tions and through bilateral and multilateral official institutions will be
assessed. In each case, the object will be to describe the criteria of
technological appropriateness that are being applied and the manner in
which they are implemented.

Technological Decisions by National Private Enterprise
Technological choices are determined solely by market conditions
only under special and rigorous conditions. Conventional economic

analysis shows that, when the enterprises are profit maximizers and
participate fully in resources markets, they will tend to make
technological decisions in such a way that the incremental value of the
output from using more of any one input will equal the extra cost of
using that input. This description of technological decisions is a rather
abstract one and does not describe any particular technology. Nonethe-
less, it is a powerful characterization and permits another step in the
analysis. Suppose that the resource markets in which enterprises hire
labor and buy other inputs and the product markets in which they sell
are both perfectly competitive. Then, if decisions are made as de-
scribed, firms will maximize not only profits but also the total value of
output the system can produce. Moreover, under the conditions cited,
and if all capital resources were used, there would be no unemploy-
ment of labor. The wages of workers would be flexible and would
always adjust so that it would be profitable to hire unemployed labor.
Thus, if competition is perfect, and if it is always possible to use labor
productively, two appropriateness criteria would be satisfied
simultaneously-the value of output would be maximized and full
employment would be achieved.
However, even under the extreme assumptions of maximizing be-
havior and perfect competition, it is not possible to conclude that the
technological choices would be satisfactory in terms of any or all of the
other criteria described in Chapter 3. The invisible hand guiding
individual decision making to the social goals of output-value maximi-
zation and full employment will not guide even a competitive system to
goals that are not expressed and fully decided through markets. Fur-
thermore, economic growth and income-distribution goals and even the
balance of payments correction goal are not usually market deter-
mined, nor are political goals or general quality-of-life objectives.
Maximizing behavior by private firms in competitive markets will
not produce the socially desired growth rate for a developing country
for a variety of reasons. The growth rate is constrained by economic
conditions, but within those constraints it is essentially a political
decision, in the broadest sense, and depends on the resources that can
be mobilized by the public sector as well as the private sector.
The effects of a competitive system on the personal distribution of
income are not readily determined. Although there is a growing litera-
ture on the relations between market structure and income distribution.
few conclusions will bear much weight at this point. For distributional
objectives, as for growth objectives, it can be said that a competitive
system does not guarantee the achievement of the particular distribu-
tional goals that developing countries set for themselves.2 This general

Determinants of Technological Decisions 69
conclusion must also be true for the political development and the
quality-of-life criteria of appropriateness, though such issues have
seldom been considered.
The blindness of competitive systems to distributional and
noneconomic goals is a well-known feature of such systems. That does
not mean that they do not have implications for income distribution,
regional dispersion, the quality of life, political development, and so
on. But those implications have not yet been identified by a general
economic or more broadly social analysis.
At the other end of the economic spectrum-with perfect
monopoly-and still with profit-maximizing behavior, even less can be
said, except that the output value may not be maximized for the entire
economy. There would still be full employment if the other conditions
cited prevailed, but workers would not, in general, be so productively
employed as if there were competition. However, with respect to the
other economic and noneconomic goals embodied in the technological
appropriateness criteria, little if anything general can be said.
Complete monopoly may be as rare as perfect competition, espe-
cially when the potential for substitution of alternative types of goods
and services is taken into account. However, it is especially likely that
many markets in developing countries suffer from some degree of
monopoly power that controls prices by controlling levels of output
and sales. Important economies of large-scale production in many
types of manufacturing industry lead to lower costs as the level of
output increases. These economies may stem from the basic physical
transformation processes as well as from the time, labor, and equip-
ment savings associated with running continuous production lines or
large-size batches. While such cost savings do not necessarily increase
indefinitely with the level of output, the minimum unit cost often
occurs at scales beyond the size of the markets typical of many of the
developing countries. Thus, it is difficult for several firms to coexist if
they compete actively. The markets for many manufactured goods in
the developing countries, in turn, tend often to be small, not only
because of low incomes and small populations but also because of
relatively high transport costs that divide the country into somewhat
separate consumption units.
In the spectrum between competition and monopoly, many kinds of
markets are possible. Thus, few if any other generalizations can be
made with respect to output, employment, or the other criteria of
appropriateness. However, there is no reason to expect any of these
goals to be achieved by economic systems characterized by substantial
market imperfections due to important elements of monopoly power.

Yet the market imperfections associated with the existence of
monopolistic elements are not the only ones that characterize develop-
ing countries. Market fragmentation or partitioning is a common fea-
ture of many of their markets. Market fragmentation occurs when
particular groups of actual or potential buyers and sellers are seg-
mented into self-contained markets and do not interact with other
groups. An extreme example is when a producer simply does not
participate in markets as a buyer of productive resources or as a seller
of his products. Complete self-sufficiency by peasant farmers would be
such a case. The restrictions on sale or rent of land and hiring of labor
that characterize traditional agricultural practices in many areas are
examples of less-complete self-sufficiency.
The fragmentation or partitioning of markets has a variety of
sources. In financial markets it is, in part, due to the official restrictions
placed on chartered financial institutions that limit the range of their
financial activities. In agriculture and other sectors where traditional
patterns of behavior may be especially important, these behavior
patterns will restrict the kinds of transactions that are socially accept-
able. While such patterns can and do change under the pressures of
development, they are also quite persistent, and it cannot be assumed
that change will necessarily be swift.
Market fragmentation limits the effectiveness of economic forces in
achieving full and efficient use of labor and other resources. Unem-
ployed and low-productivity use of resources in one sector may exist
simultaneously with nearly full and efficient use of similar resources in
other sectors.
These observations on the pervasiveness and effects of market
imperfections are significant partly because two quite different kinds of
recommendations have been based upon them. It was noted earlier that
widespread monopoly power, which especially impinges on small and
rural producers, has been cited as the justification for the search for
small-scale technologies. But these technologies, if achieved-and
there is as yet little evidence that they can be-would permit greater
self-sufficiency, so that essentially they represent withdrawals from
market participation. But this would substitute one kind of imperfec-
tion for another. The other argument, which has come with force and
repetition from authoritative sources, is to allow the market to solve
problems of unemployment. According to proponents of this recom-
mendation, individual enterprises choose the appropriate technology in
response to prices determined in efficient markets.3 The recommenda-
tion is based largely on the apparent success of several countries in
avoiding a major unemployment problem in the course of development.

Determinants of Technological Decisions 71

Korea, Taiwan, Hong Kong, and Singapore are the usual cases cited.
The argument starts with the claim that a significant range of alterna-
tive, efficient, labor-intensive techniques exists. However, other than
the studies of substitution possibilities discussed above and finally
viewed skeptically, the evidence cited is indirect: relatively intensive
use of labor is observed. From this observation, the proponents
conclude that the standards of efficiency are satisfied. But this argu-
ment is based on the assumption that these markets are reasonably free
from noncompetitive elements and are not fragmented and that man-
agers are profit maximizers.
While we are not attempting here to evaluate the conclusions for the
countries studied, their limited validity is suggested by the contradic-
tory observations of the narrow scope for capital-labor substitution
and of market fragmentation. Moreover, other studies have generated
the paradox that, even when relatively labor-intensive technological
choices exist and are more profitable than capital-intensive methods,
the labor-intensive methods are not always chosen. In a study of
technological choices in Indonesian manufacturing, this choice has
been attributed to an "engineering bias" for modem, capital-intensive
techniques. Interviews supplemented by some quantitative data indi-
cated that decisions made on technology were not dictated by cost-
minimizing considerations alone but were heavily influenced by and, in
some cases, finally decided on technical grounds that required the
sacrifice of profits. The investigator concluded that:
While part of the drive toward capital-intensive technology may be explained by the
desire of the oligopolist to insure against risk and uncertainty, a large part seems also to
be a response to some objectives of the engineering man. These are:

1. Reducing operational problems to those of managing machines rather than people.
2. Producing the highest quality possible.
3. Using sophisticated machinery that is attractive to the engineer's aesthetics.4

Yet, such a skeptical study may be no more generalizable than the
results of the studies on which the "let the market work" recom-
mendations are based.
A review of the available evidence led one observer to conclude:
Unfortunately, we cannot sort out to what extent the use of capital-intensive
technologies can be attributed to (a) the unavailability of labor intensive alternatives .
(b) inappropriate conditions of choice (information, prices, etc.) . and (c) irrational
preferences for sophisticated methods of production. . .

That unemployment is not a major problem in particular countries
does not in itself necessarily imply the effective working of good labor

and other resource and product markets. Yet skepticism about the
power of unimpeded markets to resolve unemployment problems in the
developing countries does not imply that markets do not influence the
use of resources and that it is not necessary to be concerned about
market distortions, whatever their source. It is important that market
as well as other incentives point in the right directions, if markets are to
be used to mediate economic decisions. Otherwise it is unlikely that the
development goals will be achieved. However, correct market incen-
tives may not be a sufficient condition for that achievement.

Influence of Government Policies on Choice of
Appropriate Technologies
Government intervention in the economies of the less-developed coun-
tries ranges from the direct ownership of some production enterprises
and direct controls over private enterprise to the use of a wide variety
of taxes and subsidies to accomplish overall objectives. These policy
tools are used to achieve overall as well as distributional and welfare
goals, in some cases by directly influencing the choice of appropriate
technologies. But in many instances, the effects on such technological
choices are unintended by-products of government policies.
From their observed actions, it appears that governments pursue
different goals at different times and, by the same or diverse means, at
the same time. Some government projects have been used to provide
employment; others are carried out and operate to maximize output or
to minimize cost. In community-development projects, governments
have attempted to directly shape the quality of life of the participants.
While it would be helpful to know the relative weight that government
decisions give to the different appropriateness criteria in particular
countries, that information is seldom if ever explicit, and no studies
have yet elicited it.
Finally, an important point from Chapter 3 should be recalled.
Government, by its direct action or control over private investment,
can impose on enterprises any criterion of technological appropriate-
ness it desires. But only if the goal is to maximize net revenues will the
enterprise have reasonable hope of viable independence from the
government budget. If any other criteria of appropriateness are pur-
sued, it is likely, and in some cases certain, that the enterprise's
revenues will not cover its costs. This is simply because pursuit of
other goals means that the enterprise is not trying its best to cover
costs. And, if the enterprise cannot cover its costs from its revenues, it
must be subsidized from the government budget if it is to continue to

Determinants of Technological Decisions 73
The direct government controls over new investment that are com-
mon in the developing countries are designed to achieve sectoral
allocations that are more effective in contributing to the desired rate
and pattern of development than if the investment decisions were
determined by private enterprise and market influences. By controlling
the distribution of investment between the consumption and capital-
goods sectors, these direct controls also help force a higher rate of
saving than would otherwise be achieved. In some cases, direct control
is exercised over the use of specific technologies as, for example, in
determining the scale of production or the type of fuel or raw materials
used. Except in some public enterprises, direct controls over output
are rarely used in the developing countries, though agreement on
specified output and export targets may be a condition for the granting
of investment licenses.
Price controls over both inputs and outputs are common. For exam-
ple, minimum wage laws are intended to offset local monopoly power
in hiring and to assure minimum incomes. Maximum prices on certain
foodstuffs have similar objectives in the subsidization of consumption.
Regulations that set maximum interest rates are intended to encourage
investment and, when they are set differentially for loans in different
sectors, to direct the allocation of investment.
In addition to these direct quantity and price controls, the developing
countries have much the same arsenal of taxes and subsidies as those
used in the industrialized countries. These taxes and subsidies extend
to foreign trade in the form of tariffs and export subsidies. In the
foreign-trade sector, various types of quota restrictions over imports
are also used. Although revenue collection is the major objective of
many of the taxes and subsidies, some are used to influence patterns of
output or use of inputs. The effects of all these fiscal and quantitative
instruments depend not only on the structure of each of the developing
economies but also on their abilities to assess and collect the taxes and
to control the payment of the subsidies.
A common claim is that the various government quantity and price
controls and taxes and subsidies "distort" the efficient use of re-
sources that would be produced by perfect markets. However, markets
are never perfect, and the "distortion" may be intended because the
operation of actual markets is distrusted and the economic patterns
they would create are rejected. Other distortions are not intended but
occur as by-products of government actions; their effects may over-
shadow the primary purpose of those actions. For example, interest-
rate ceilings on bank loans, which are intended to encourage new
investment, often set the price of funds much below the rental values of
the real capital the funds can purchase. Low interest rates may be quite

unnecessary in the developing countries, where the incentives to invest
are so substantial that low interest rates are not required to encourage
new investment. Interest-rate ceilings lower the cost of capital for that
portion of the business community with access to the funds at the
controlled rate. This, in turn, may lead to the choice of more capital-
intensive technologies than would be warranted to maximize output,
the rate of growth, or the absorption of labor or to satisfy any of the
other criteria of appropriateness.
Similarly, government labor regulations (such as minimum-wage
laws, requirements on employers for social contributions, or severance
pay regulations) may raise the effective price of labor above its real
scarcity value in terms of the goals of the economy. That, in turn, will
encourage the substitution of other productive resources for labor and
will reduce employment incentives. This is another example of the
potential, if not actual, conflict between alternative criteria for the
appropriateness of technological choice. The objective of the various
minimum wage, welfare contribution, and employment regulations is to
improve the distribution of income. The immediate side effect, how-
ever, is to create incentives to reduce employment. The final result, as
a consequence of reduced employment, may be increased inequality in
the distribution of income.
A number of other types of government policies designed to guide
and control the development process have similar, though more indi-
rect, effects. Government tariff and quota restrictions on imports and
export subsidies, for example, may affect the technological choices. As
noted in Chapter 4, it has been alleged that foreign-trade policies are
prime determinants of the choices of technology and, in particular, of
the absorption of labor in manufacturing. Overvaluation of the ex-
change rate (for example, by discouraging exports) may emphasize
industries with only limited employment potential.
The emergence of unemployment as an open and pressing problem in
the developing countries has led to government commitments to deal
with it directly, and these commitments, in turn, have become part of
development ideology. For example, job guarantees have been re-
quired to prevent workers from losing their jobs when they must leave
work for sickness or to do military service. In some cases, subsidies
have been provided for the employment of labor. These policies
encourage the inefficient use of labor and the flow of workers from
sectors not covered by the regulations to the covered sectors.6 In
practice, that stimulates migration of labor from rural to urban areas.
Thus, one result of such policies is the creation of more open unem-

Determinants of Technological Decisions 75

International Private Enterprise and Market Influence on
Choice of Appropriate Technologies
The significance of international business as a source of knowledge that
affects the choice of appropriate technologies has already been noted.
At this point, the issue is the character of the technological decisions
that are made directly or indirectly under the influence of international
business or multinational corporations.
The primary objective of the multinational corporations, like other
private enterprises, must be the long-run maximization of their profits.
However, that is not necessarily inconsistent with all the goals em-
bodied in the various criteria for the choice of appropriate technology.
Neither is it certain that there is consistency with any of the criteria.
The extent to which profit maximization by the multinationals is con-
gruent with national-development goals will depend, in large part, on
the character of the regulations and incentives to which they are
subject. Yet there are some goals and technological decision criteria
that in a range of likely circumstances will be either irrelevant or
antithetical to the private-profit objectives of the multinational, as well
as to national, firms. For example, no private enterprise, national or
multinational, is interested in maximizing employment for its own sake
or reducing inequality in the distribution of income for the sake of
social welfare, even though their operations may contribute to em-
ployment and, possibly, to income equalization under certain circum-
stances. If objectives of this type are imposed to such an extent that a
multinational's relative advantage of operating in a developing country
disappears, then the multinationals will simply not invest, and what-
ever potential contributions they might make will not be realized.
Although private national and multinational corporations both may
have the same objective of maximizing their profits, pursuit of that goal
by each may well lead to different technological decisions, even in the
production of the same commodity and in response to the same set of
input prices. Such differences may even be optimal, not only for the
individual firm but for the economy as a whole. On the other hand,
other differences between the manner in which national and multina-
tional enterprises operate may not be either rational or in the interests
of the developing countries.
One difference between national and multinational firms that may
indeed be rational and optimal lies in the way they choose between
existing and new technologies. National firms may try to develop
technologies especially suited to local conditions. But multinationals
may prefer to transfer to the developing countries the technologies that

have been worked out for the relatively industrialized countries. Such
transfers may be undertaken without considering the possibility of
developing alternative production methods. However, the decisions
may also reflect, implicitly if not explicitly, the advantages of working
with well-known methods. To develop techniques more suitable to the
labor and other resources of the developing countries, research expen-
ditures and, perhaps, pilot plants and production experience are neces-
sary. By comparison, familiarity with conventional methods makes
their implementation less costly and faster, and the outcome more
certain. For the multinational corporation, the choice involves compar-
ing the costs of existing methods with all the costs-research and
development, investment and operating-of finding and implementing
new technologies. In particular cases, the cost comparison can easily
favor implementing the known technologies. That decision may also be
optimal for the developing country if any plant that gets a new
technology has to bear all the costs of research and experimentation.
Their intimate knowledge of production methods may give the
multinational corporations an advantage in the use of "second-hand"
equipment from their own or other establishments in advanced coun-
tries. While not necessarily "appropriate," such equipment may em-
body relatively labor-intensive technologies especially suited for some
of the developing countries. However, the quality of second-hand
equipment varies greatly. Thus, since multinational corporations can
be expected to be familiar with the used machinery's previous pro-
duction history, they can more effectively use this means of adapting
technology to the conditions of the less-developed countries.7 Yet this
advantage is lost if, as is sometimes the case, the importation of
second-hand equipment is restricted.
In general, the technological options for national private or public
enterprises of the developing country are different from those for
multinationals. For the former, the conventional and well-proved
technologies may not be readily accessible or would become available
only at a cost, in terms of royalty and license fees, which reflects some
monopoly power on the part of the multinational. Moreover, national
private and public enterprises in the developing countries are likely to
have a longer time horizon in the country than the multinational
corporation, so any operating-cost advantages of a new method can be
expected to pay off over a longer period. The importance of the latter
consideration is reduced by the relatively high discount rate appropri-
ate to investment in the developing countries, but it is not eliminated.
Another source of difference in the choice of technologies by multi-
national corporations is the prices they pay for their productive inputs

Determinants of Technological Decisions 77
as compared to the prices national enterprises pay. Multinational firms
borrowing in international financial markets may procure credit at
lower interest rates than national firms without access to those mar-
kets. It has also been argued that multinational corporations investing
in a developing country will often raise a significant portion of their
funds in the country itself and at a lower interest rate than is available to
national enterprise.8 It may well be rational for national financial
institutions to lend to multinational corporations at lower rates than are
available to national enterprise. Since the multinational corporation is
typically larger than most national firms and may represent a diversi-
fied group of enterprises, it is less likely to default than the smaller and
less-diversified national enterprises. Moreover, loans by national
financial institutions to multinational enterprise can represent a desir-
able diversification to their portfolios so, again, they may be relatively
willing to supply funds. For these reasons, the real cost of financing the
multinationals from national sources is likely to be less than the cost of
financing national private enterprise, unless national enterprises are
subsidized by their government or permitted to borrow at low regulated
rates. In such cases, the only advantage left to the multinational is the
lower risk of lending to it. But that may well provide enough justifica-
tion for the national financial institutions to ration credit to satisfy the
multinational's loan requests first.
On the other hand, multinationals may well face higher labor costs
than national firms. First of all, labor legislation that effectively raises
labor costs often does not apply to the many relatively small firms
typical of developing countries. Secondly, labor legislation may be
differentially enforced among the firms to which it applies in principle.
Since the multinationals are relatively conspicuous because of their
size and because they are foreign, enforcement for them is likely to be
stringent. For analogous reasons, labor unions are likely to be rela-
tively effective in organizing and gaining wage increases from multina-
tional corporations.9
If the price of capital is relatively low and the wages of labor are
relatively high for multinationals operating in developing countries,
they will be induced to adopt relatively capital-intensive methods when
choices are possible. Their choice is not the result of any intention to
do a disservice to the developing countries. It simply emerges as they
try to maximize their profits by finding least-cost methods. For these
reasons, and because the technologies used in the industrialized coun-
tries are more easily available to multinationals, it is not surprising to
find that these are the technologies the multinationals often transfer to
the developing countries. Even so, as usual, the picture is mixed; there

is evidence that, in some cases, the multinationals have been more
assiduous in attempting to adapt their technologies to local conditions
than have national firms. A comparative study of can-making tech-
niques in Africa and Thailand showed that multinationals can be both
more and less sensitive to local conditions than indigenous firms.10
However, there are also irrational and misguided reasons for the
transfer to the developing countries by the multinationals of
technologies that are inappropriate by any standards. The unwarranted
"engineering" preference for modern techniques may especially
characterize the multinationals whose knowledge of such techniques is
more intimate. The multinationals may also use the royalty payments
on their imported technologies as a means of repatriating a larger
proportion of their profits than would otherwise be permitted by law.
They may insist on using particular technologies to help ensure a
market for their intermediate products or their spare parts; or they may
justify the choice of their conventional technology as a means of
maintaining quality standards, but again those quality standards may
be misguided when applied to the developing countries."1
Multinational corporations have been criticized as proponents of the
dependent roles that the industrialized countries have forced or in-
duced the developing countries to accept. This dependency, it is
claimed, deprives the developing countries of autonomy in determining
their national policies, since employment, production, and resource-
exploitation decisions are made in the context of maximizing the profit
of institutions outside the developing country. Thus, it has been argued
that the technological decisions made by multinational corporations are
also part of the overall climate of dependency.12
The "dependency" arguments contain threads of truth that may not
be generalizable, arguments for causation that have not been conclu-
sively demonstrated and, above all, a strong assertion of national
autonomy in virtually all aspects of economic life as a value that cannot
be questioned. That the dependency arguments have some validity is
demonstrated, for example, by the multinational corporations' admis-
sions that they have used their relative financial power to gain advan-
tages not necessarily available to national firms. That is certainly an
invasion of national autonomy. When this use of financial power takes
the form of illegal bribes, national policy is violated in the same sense
that illegal acts are violations.
It is pointless, however, to criticize the multinationals because they
do not try to advance the economic and political goals of the develop-
ing countries. That is not their function, just as it is not the function of
national private enterprise. While multinationals may not choose the

Determinants of Technological Decisions 79
technologies most suitable for the developing countries, their choices
are made to advance their own self-interest rather than in conspiracy
against the interests of the developing countries. In advancing their
interests, they can be expected to take advantage of whatever
monopoly power and economic and political influence they command.
Careful regulation can prevent the multinationals from using their
influence in directions counter to national goals. But that, in turn,
requires a government capable of overseeing and influencing national
economic processes. Creating such capabilities is one of the central
problems of political development and is by no means an automatic and
easy achievement of national governments.

Influence of Bilateral and Multilateral Official Institutions
on Choice of Appropriate Technologies
Through their official relations the governments of the industrialized
countries and the international economic assistance institutions pro-
vide technological information to developing countries in the course of
sponsoring particular projects. In addition, they also exert a broad
influence over the economic policies of recipient nations, including
their policies with respect to technological decisions. This influence is
contained in part in the terms imposed on the developing countries as
conditions of loans or grants from governments and international
institutions. That the influence is not always exerted exactly in the
directions the developing countries themselves would prefer is demon-
strated by their complaints about the "strings" and "leverage" ac-
companying official international loans and grants. While the character
of this leverage has been the subject of considerable controversy, it is
now virtually officially acknowledged in general, though little, if any-
thing, is publicly known in detail.13
To ameliorate the impact of development assistance on the donors'
balance of payments, most of such assistance provided through bilat-
eral arrangements is "tied" to purchases within the lending govern-
ment's economy. As a result, the range of technological alternatives
from which choices may be made is restricted, and that can lead to
inappropriate choices. When loans and grants are tied to expenditures
in the lending/granting nation, the assistance may have to be used in a
relatively high-cost market. This was the effect of tying United States
assistance before the 1971 depreciation of the dollar. In effect, such
tying to a relatively expensive market reduces the value of the eco-
nomic assistance that is provided.14 The developing countries may
decide to choose relatively capital-intensive methods because those are

the only ones available. However, though the methods appear capital
intensive, they may not be, in reality, since no assistance would be
provided if the particular techniques were not used. The real cost to the
developing country is not the "face value" of the equipment but
something substantially less. The casual observer, seeing only the
automated textile factory resulting from assistance, could well come to
the wrong conclusion.
The precise character of the influence that official sources have on
the technological decisions of the developing countries will vary with
individual government assistance agencies and multinational lending
agencies, and the manner in which the influences manifest themselves
will also differ among the developing countries. The assistance pro-
grams of the industrialized countries have a variety of rationalizations,
ranging from humanitarian to the narrowest self-interest. Critics of
dependency have charged that the industrialized countries have con-
sciously used assistance programs to maintain the developing nations
in a technologically subservient position by refusing to supply ad-
vanced technology, by supplying only obsolete technology, or by not
supplying the technology and loans needed to develop sectors that
would increase the self-sufficiency of the developing nations.15 With-
out more information than is available on loan and grant negotiations
and influences on technical choices, these arguments cannot be
brushed aside. On the other hand, a broad range of industries and
technologies has been supported by assistance programs, including, for
example, the most modern high-speed computer and petrochemical
plants as well as technical assistance in simple agricultural and handi-
craft production. This, in itself, suggests that there is at least no overall
and systematic pattern of technological discrimination.
All of the international agencies are subject to constraints that, to
some extent, prescribe their lending terms. For example, the Interna-
tional Bank for Reconstruction and Development (World Bank) bor-
rows in international financial markets a substantial part of the funds it
lends. It is understandable, therefore, that the criteria the World
Bank applies to determine the use of such funds are similar to those
used by private enterprise. These, as noted above, can be narrower
than the criteria that policy makers in developing countries might use.
A substantial part of the World Bank's funds are subscribed by its
member governments, and governments that are the major sources of
funds and shapers of World Bank policy in the use of these funds may
also have different criteria for development from those of the develop-
ing nations.
It is also clear that the assistance agencies of the industrialized

Determinants of Technological Decisions 81
countries and the international lending institutions often have a differ-
ent conception of the most promising growth policies than do the devel-
oping nations themselves.
This is clearly true of the World Bank, as example and epitome of
international lending agencies. Its conceptions change over time,
however, under the impact of the arguments of the developing coun-
tries and new perceptions of the development process. Only recently,
for example, has the World Bank recognized employment, income
distribution, and rural development as worthy development goals
different from growth in output and balance of payments correction.
Whatever the sources, the conceptions of development processes held
by such agencies as the World Bank provide the intellectual basis for the
leverage that, according to authoritative sources, is exercised in giving
their assistance.16
The International Monetary Fund and other international lending
agencies have received less critical attention than the World Bank.
Yet, there is evidence that each also has its characteristic views about
the national development policies that are worthy of its assistance.17
Once it is accepted that these national and international economic
assistance agencies do use their power to influence the policies of
developing countries, the question that then arises is what determines
the policy choices of these agencies and, in particular, their criteria for
appropriate technologies. Presumably, the agencies would answer that
their preferences reflect the study and judgment of their staffs, distill-
ing whatever wisdom is available in the various relevant professions.
Unfortunately, such an answer is not conclusive, as there are strong
differences of opinion concerning the validity of the agencies' posi-
tions. Not only may each developing country claim that its circum-
stances are unique; its interpretation of development theory and expe-
rience may also lead to different conclusions. Moreover, they can well
claim that the bilateral and multilateral assistance agencies themselves
have self-contradictory rules and goals. For example, it is common
now for projects being considered by national or multilateral assistance
agencies to be judged on the basis of a well-defined system of cost-
benefit analysis whose criterion may be social cost minimization,
consumption maximization, or growth. That might point to a particular
source of supply of machinery and technological information. Yet, as
noted previously, the tying of assistance by national agencies may
force recipients to use less desirable technologies than would otherwise
be chosen. For example, although investment decisions in a ther-
moelectric station depend on a cost-benefit analysis, the choice of
equipment may be confined to relatively high-cost suppliers in coun-


tries supplying the foreign exchange. In addition, the developing coun-
tries can be expected to have different development goals than those
set for them by assistance agencies. These differences manifest them-
selves in the risks they are willing to take with respect to inflation or
balance of payments problems or other development difficulties.

1. For a classic argument see P. Rosenstein-Rodan (1943).
2. The theoretical analysis of the implications of competitive systems for the personal
distribution of income has, as yet, produced results that, at best, are asymptotic
with little indication as to rate of progress toward the asymptote. See J. Stiglitz
(1969), K. Meesook (1974), and A. Dasgupta (1975).
3. H. Hughes (1974) and G. Ranis (1974).
4. L. Wells, p. 83, in P. Timmer et al. (1975).
5. E. Edwards (1974), p. 17.
6. See, for example, J. Harris and M. Todaro (1970).
7. The comparative study of can-making techniques used in Kenya, Tanzania, and
Thailand of C. Cooper, R. Kaplinsky, R. Bell, and W. Satyarakevit, in A. Bhalla
(1975), provides interesting insights into the variety of types of behavior of
multinational firms. In some cases it is quite sensitive to local factor prices, in
other cases it is not.
8. For contrasting views see, for example. W. Manser (1974) and S. Plasschaert
9. For a strikingly implicit argument see N. Weinberg, in A. Said and L. Simmons
10. C. Cooper, R. Kaplinsky, R. Bell and W. Satyarakevit, in A. Bhalla (1975).
11. For a strong view see H. Mowlana, in A. Said and L. Simmons (1975).
12. 0. Sunkel (1973).
13. T. Hayter (1971) contains descriptions of the leverage applied by the International
Bank for Reconstruction and Development in its lending operations and allegations
of attempts to suppress their publication. This leverage was later mentioned rather
briefly in the "official" history of the World Bank by E. Mason and R. Asher
14. See J. Bhagwati (1970).
15. See T. Hayter (1966).
16. E. Mason and R. Asher (1973) provide a detailed description of the operations of the
World Bank to 1973 without, however, an analysis of the rationale.

6 Special Features of
Technological Decisions

in Agriculture

After years of receiving little analytical and administrative attention
and relatively small investment allocations, the central role of the
agricultural sector in development has become more widely ap-
preciated. The relative size of the agricultural sector in the developing
countries itself warrants particular attention to the sector. In most of
the developing countries, the share of agricultural output in total output
ranges between 30 and 50 percent and in some cases is even larger.
Since the earnings of labor are generally lower in agriculture than in
industry and government, the proportion of the work force in agricul-
ture is higher than the share of output. This means that relatively small
changes in the demand for labor in agriculture result in major changes
in the availability of labor to the urban manufacturing and service
sectors. Suppose, for example, that the total labor force of a develop-
ing nation were set at 100, with 60 in agriculture, 20 in industry, and 20
in services. If there were a 2-percent annual increase in labor
productivity in agriculture with no opportunities for increased
employment in that sector, then 1.2 laborers would be released per
year from agriculture. These would have to be absorbed by the other
sectors if unemployment were not to increase. That would mean an
annual increase in employment in both industry and services of 3
percent. If only manufacturing is expanding, employment there would
have to grow by 6 percent. But if labor productivity in the

nonagricultural and other sectors were also growing, as would be
expected in the course of development, then output in these other
sectors would have to grow still more rapidly to absorb all the labor
displaced from agriculture.
In addition to its size, the other unique features of agriculture in most
of the developing countries justify an analysis specially tailored to it.
Because of its unique characteristics, agriculture cannot simply be
"left to itself" if development is not to be frustrated by agricultural
scarcities, if urban unemployment problems are not to overwhelm the
economy, and if all sectors are to participate equitably in the benefits of
economic growth. This chapter will identify and describe briefly the
unique features of agriculture and relate them to the special problems
of finding and implementing appropriate technologies in the sector.

Diversity in Agriculture
The great diversity in the technical conditions of production in agricul-
ture has several sources. The products of agriculture are themselves
quite heterogeneous and have diverse input patterns. The requirements
for the production of the row crops of grains or legumes are different
from the requirements for tree crops of fruits, nuts, berries, or leaves
and from the requirements needed to raise animals. Some of the inputs
needed to grow staple grains, for example, may be completely irrele-
vant for raising livestock, and the land and water requirements can be
quite different. The products also differ in the degree to which they can
be modified and adjusted to diverse geographic conditions. Land and
climate conditions, of course, have great effects on productivity of
labor and other resources employed in agriculture. While labor and
capital can substitute for unfavorable natural conditions to some de-
gree, the extent and ease of substitution depends on natural conditions.
The limits on our knowledge make it impossible to say that the range
of physically efficient input proportions is greater in agriculture than in
other sectors. Such comparison is difficult not only because knowledge
of the technology in other sectors is limited, but also because even in
agriculture it is seldom possible to know in detail the range of input
proportions suitable for each locale. Nonetheless, the prevailing im-
pression is that there is considerable variety in potentially economical
input choices for important types of agricultural outputs.' Figure 1,
which compares the per-acre use of fertilizer and per-worker use of
tractor horsepower among countries, illustrates for these two factors
the wide range of intensity of use of resources in agriculture. Even
within the same locale, the range of technological choices may be
different for different products.

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

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