SWP344
Januarv 1985
Intemational Technology Transfer:  M!>-;B,RY
Issues and Policy Options
,a r 3C w- 
Frances Stewart
WORLD BANK STAFF WORKING PAPERS
Number 344
2  f Pry h



WORLD BANK STAFF WORKING PAPERS
Number 344
International Technology Transfer:
.Issues and Policy Options
Frances Stewart
The World Bank
Washington, D.C., U.S.A.



Copyright (i1979
The International Bank for Reconstruction
and Development/THE WORLD BANK
1818 H Street, N.W.
Washington, D.C. 20433, U.S.A.
All rights reserved
Manufactured in the United States of America
First printing July 1979
Second printing January 1985
This is a working document published informally by the World Bank. To present
the results of research with the least possible delay, the typescript has not been
prepared in accordance with the procedures appropriate to formal printed texts, and
the World Bank accepts no responsilbility for errors. The publication is supplied at a
token charge to defray part of the cost of manufacture and distribution.
The World Bank does not accept responsibility for the views expressed herein, which
are those of the authors and should not be attributed to the World Bank or to its
affiliated organizations. The findings, interpretations, and conclusions are the results
of research supported by the Bank; they do not necessarily represent official policy of
the Bank. The designations employed, the presentation of material, and any maps used
in this document are solely for the convenience of the reader and do not imply the
expression of any opinion whatsoever on the part of the World Bank or its affiliates
concerning the legal status of any country, territory, city, area, or of its authorities, or
concerning the delimitation of its boundaries, or national affiliation.
The full range of World Bank publications, both free and for sale, is described in the
Catalog of Publications; the continuing research program is outlined in Abstracts of
Current Studies. Both booklets are updated annually; the most recent edition of each is
available without charge from the Publications Sales Unit, Department T, The World
Bank, 1818 H Street, N.W, Washington, D.C. 20433, U.S.A., or from the European
Office of the Bank, 66 avenue d'I&ia, 75116 Paris, France.
When this paper was first published Frances Stewart was a member of the
Development Economics Department of the World Bank.
ISBN 0-8213-9219-0



ABSTRACT
This paper surveys the issues raised by technology transfer to
developing countries. Technology is a vital part of the development
process, a necessary input into all activity. Because of the
historical domination of research and development by developed
countries, a very large part of the technology used is transferred
from developed countries. The transfer raises four major issues:
those of the costs of the transfer, the appropriateness of
products and techniques which are transferred, the effects of the
transfer on learning and technological development in LDCs, and
the effects on independence. The paper considers the consequence
of the transfer and the range of policies that might be adopted,
nationally and internationally, in relation to each of these issues.
It is argued that the appropriate policies will vary according to
the stage of development of each country, its technological
capacity and its own objectives. However, in general it is
concluded that an active technology policy is called for if
the costs associated with technology transfer are to be reduced
and the benefits increased.  While it is easy to design suitable
policies on paper, it is much more difficult to ensure that they
are effective in practice.



TABLE OF CONFENTS
SUMMARY                                                           vii
I.     Why Technology is Transferred                                        2
II.     Transfer Mechanisms                                                 11
III.     Significant Dimensions of Technology Transfer                       20
IV.     Costs of Techrnology Transfer                                       23
Technology Transfer and Independence                                37
Learning Effects of Technology Transfer                             39
V.     Policies Towards the Import -f Technology                           46
Economic Strategy and Technology Policy                             48
Policies Towiards The Terms of Transfer                             52
Promotion of Technological Accumulation                             61
Administration                                                      73
Aspects of Political Economy                                        75
VI.     Thc Characteristics of Technology                                   78
Policies to Secure More Appropriate Technology                      86
Infonnation                                                   94
The Devel.opment of Appropriate Technology                    97
VII.     Some Conclusions                                                   101
LIST OF REFERENCES                                                        109



- vii -
SUMJARY
Technology is knowledge of how to do all those things associated
with economic activity. Thus, the acquisition of technology is as important
to the development process and as essential an aspect of the process of
investment as the accumulation of financial resources. Moreover, because
technology is continuously changing, with new products and processes being
developed year-by-year, the need to acquire technology is also a continuous
one, and one that expands with expansion in economic activity.
Two basic facts underlie the process of international technology
transfer. First, the origin of the vast majority of technological develop-
ments is concentrated in a few developed countries. Although some developing
countries are beginning to promote local technological development, they
remain dependent on tlhe developed countries for most of their technology.
Secondly, a large part of the market for technology has been commercialized
with proprietary rights acquired in the technology. Since technology has many
of the characteristics of a public good, such that the marginal cost of
communicating it to others is very low com- ared with the initial costs of
development, the commercialization of the technology market has involved
imposing restrictions - legal and other - on the free communication of know-
ledge. The result is that the technology market is highly imperfect, the
price charged for technology tends to be oligopolistic, and consequently there
is considerable scope for abuse,and potential for bargaining on the part of
buyers. The transfer of technology consists, in part, in the transfer of the
required knowledge; in part, in the transfer of various marketing rights
associated withl the knowledge. These include the right to use trademarks,
access to specified markets, and so on.  In many cases, the two are  inextricably
connected, so that if a buyer wishes to purchase the knowledge he also has to
purchase the m.arket rights, and vice-versa.



- viii -
Teclhnology is transferred in a number of different forms, ranging
from formal (technology contracts) to informal (meetings, people changing
jobs). The process is often differentiated according to the degree of
packaging involved in the transfer with the most packaged transfers consist.r.g
of the bundle of services associated with direct foreign investment, and the
least packaged in the direct purchase of machinery and other inputs. There
are various intermediate forms including joint ventures and contracts involving
associated management services *and knowhow. In any particular case, the
mechanism of transfer depends on the willingness of the supplier to supply
the technology in different forms and the desire and ability of the buyer
to buy it in a particular form. There is considerable variation in the way
different countries acquire technology reflecting differences in national
policies, in national capacities and in the industrial composition of the
technology transferred. In some countries, e.g., historically Jartn and
currently the Andean Pact countries, the most packaged forms of technology
transfer, notably direct foreign investmene, are severely restricted. The
degree of development influences the extent of packaging because less developed
countries may lack the capacity to put the package together for themselves. In
general, more sophisticated (technologically) industries show a greater
degree of packaging than less sophisticated. Despite recent support by some
LDCs for less packaging, no definite worldt-ide trends towards depackaging can
be discerned empirically.
Developing countries' objectives towards technology transfer have
evolved. Immediately following independence, most countries were keen to
industralize quickly, and to maximize the technology inflow.  Hence, they
introduced a host of tax incentives to encourage teclnology inflow.  As time



- ix -
proceeded, it emerged that the indiscriminate import of technology nad involved
high costs, had led to dualistic patterns of development, had accentuated
dependency and had inhibited learring effects.  Consequently, new objectives
evolved: in particular, many developing countries wished to regulate the
inflow of technology to reduce the costs of transfer, to increase the in-
dependence of decision making and to promote local technological innovation.
In addition, in response to employment and other problems, some countries
recognized a need for more appropriate technology.
Costs of Technology Transfer
The nominal declared costs of technology transfer need bear little
relationship to actual costs because companies which supply more than one
service may choose the form in which they receive payment. For example,
payments for technology may appear as royalty payments or as profits or as
overinvoiced import prices. The problems of disguised nayments arise in
most acute form for the most packaged forms of transfer, but are also present
to some extent in other forms, such as joint ventures.  The high proportion of
intra-firm trade and of intra-firm technology flows gives rise to the possibility
of disguised payments. There is considerable evidence of overinvoicing of imports
and underinvoicing of oxports. Wlile the evidence suggests that this
occurs to the greatest extent in pharmaceuticals, there is strong evidence of
similar practices in many other industries. Evidence of such practices has been
collected in a number of Latin American countries, and in India and Greece.
The industries covered include pharmaceuticals, clhemicals, metal products,
minerals and rubber products. It is difficult to know how much of the
disguised payments should be attributed to technology transfer and how much
to other facLors.  The need for technology transfer normally gives rise to
th!c po;sIbility of such paymuent:s, but their extent is influenced by recipient



- x 
governments' economic policies as well as by lack of compet;tion in the
world industry.
The process of technology transfer imposes additional costs as
the result of restrictive clauses associated with it. In the case of direct
overseas investment, the restrictions are implicit. For technology con-
tracts between independent parties, they are an explicit part of the contract.
There is evidence of widespread restrictions embodied in technology contracts.
The restrictive clauses include limitations    (or total bans) on exports,
tied purchase of inputs, machinery and parts, limitations on competing supplies,
and constraints which limit the dynamic effects of transfer by req iring excessive
use of expatriate personnel and discouraging local r. and d.
UNCTAD estimated the overt costs of technology transfer to be 5%
of non-petroleum exports of developing countries in 1968, and estimated that
this proportion would rise steadily over the subsequent 10 years.  Miile
other evidence suggests that the UNCTAD estimates (particularly in relation
to the rate of increase after 1968) may be overestimates, adding some .stimate
of .he implicit and indirect costs would substantially increase the costs.  for
example, to add 10 of the costs of imported capital equipment and chemicals
and one-fifth of the declared profits on foreign investment to the UNCTAD
estimates would more than double them.  The high costs do not necessarily mean,
of course, that the price is not worth payirg.  But they do suggest that there
may be considerable gains to be made from bargaining, especially in view of the
non-competitiveness of muchi of the market.
Independence
Technology transfer limits the indepenidence of decision making.  This
is most obvious in the case of direct inx'tstment, but occurs to a considerablc
extent witlh respect to transfers between independent parties where technology



- xi -
contracts include very heavy restrictions on freedom of decision making.
It has been claimed that technology transfer has come to replace ownership
as a mechanism of control.
Learning
The ability to make independent technological choices, to adapt
and improve upon chosen techniques and products, and eventually to generate new
technology are essential aspects of the development process. Relatively little
is known about the process of accumulation of technological capacity. Broadly,
three stages may be identified: in the first, the ability to make independent
technol'rical choices is developed; in the second, minor innovations occur
locally; and in the third, the capacity to create new technology is developed.
In general, the stages occur sequentially. There is a complex relationship
between the import of technology and the development of local technological
capacity, and one that encompasses both complementarities and conflicts.
Some import of technology is required as a basis for learning. But highly
packaged imports tend to limit the development of the ability to make
independent choices. Minor technological change may occur on the basis of
imported technology - this seems to respond to a mixture of technological
capacity and competitive and other pressures. The third stage - local de-
velopment of technology - may be inhibited by indiscriminate imports - partly
because of restrictive clauses involved in technology contracts which limit
the dynamic effects; partly because the foreign technology associated with foreign
trademarks tends to have a strong market advantage over local teclhnology; hence, loca
technology, even when developed, is often not used, and thus the incentive to develop
it Is much reduced. The most successful examples of technological development
have combined selective import of foreign technology with positive measures
to promote local technology. Thus the local developments are provided
with the foreign knowhow they need and the protection and promotion necessary.
Promotional measures include provision of infrastrtucture, tcchnology plans,



- xii -
subsidies and incentives to r. and d. and government procurem2nt practices that
favour local technology. Japan's technological development is a very success-
ful case of this sort of policy: technology was imported, then adapted do-
mestically and promoted in both public and private sectors.  Other successful
examples, involving a similar comlbination of policies, have been observed in
a few cases in Mexico and India.
Policies Towards Technology Imoorts
Relevant policies vary according to objectives, although there is
considerable overlap with some po:Licies serving more than one objective, and
the objectives themselves being mutually reinforcing.  For example, policies
which eliminate restrictive practices will also increase independence, reduce
costs and may contribute to the learning effects of the technology.
The set of relevant policies varies with the stage of develcpment, and
particularly the technological and administrative capacity of the country con-
cerned. Policies aimed at improving the terms of technology transfer are rele-
vant to most countries. Although they involve administrative costs, the potential
gains are lar-e, as shown by the experience of countries which have followed them.
Iowever, policies of selectivity towards imported technology require a greater
degree of technological sophistication and administrative judgment. These
policies are only worth pursuing where there exists fairly considerable local
technoJogical capacity; they may best be introduced on an industry-by-industry
basis.
The general economic strategy of the country is of relevance to the
effectiveness and outcome of technology policies.  The terms of technology
tranisfer are likely to be more competitive, the more competitive the general
environment is, and the less government policies (e.g. of protection) give rise
to quasi-rents.  But non-competitivc elements in the world industrial structure
me.nn that s;pecial technology policies are required in addition t,o general



- xiii -
competitive policies. As far as local technological development is concerned,
the relationship between general economic policies and local technological
developments is less clear. While it is often argued that heavy protection
and "distorted" factor prices remove incentives for local adaption, there
are a number of examples of local technological'innovation occurring in a generally
protected environment. Thus, India and Argentina have developed technology
locally,in a few cases sufficiently to export the technology to neighbouring
countries.  The sort of pressures which lead to local technological innovation
are of many kinds, e.g., absence of particular raw materials, and not confined
to cost pressures associated with pri.e and exchange rate policies.
A number of countries have tried to reduce the packaging element in imported
technology. A major reason is to secure greater independence of decision
making: this is desired in itself, and also as an essential part of the
learning process so that countries may develop the capacity first o make
independent technological choices, and subsequently to develop their own
technology. The policy is also related to the question of costs, since it
appears that costs tend to be higher the more packaged the form of transfer.
The policies may be mandatory - e.g., forbidding majority foreign owner-
ship in some or all industries - or rely on incentives. An essential
complement to policies towards depackaging is the introduction of some
sort of regulation over technology transfer between independent parties;
otlherwise, as indicalted by empirical evidence, many of the problems associated
with direct investment re-emerge via technology contracts. A number of countries
have introduced National Registers of Technology, permitting registra-
tion only if the cbntracts fulfill various criteria with respect to terms and
restrictive clauses. In general, restrictive clauses (e.g., with respect to
export bans and Lying inputOi) are outlawed.  Investigations of the results of
thesO policics in two countries - Col.ombia and Mexico - have slhown very



- xiv -
substantial  foreign exchange gains, and the outlawing of a large number of
restrictive clauses at relativelyl low administrative Cost.   But it has not
yet been thoroughly establislhed how far such policies have had de facto
as against de jure success, as many of the explicit and formal restrictions
may have been replaced by informal implicit ones.
Changes in the tax system may be designed to improve the terms of
technology transfer. In particular, if corporate taxation is made independent
of the degree of underinvoicing or overinvoicing (either by formula apportion-
ment or by a uniformity of rates approach), the foreign exchange and revenue
loss cause by these practices will be r,much reduced. The removal of tax
incentives towards overseas investment, which have been shown to be ineffective,
would also improve the terms of technology transfer.
Policies to protect and promote local technological developments involve
the selective import of foreign techno'ogy as well as promotional measures towards
domestic developments. Industries in which there is substantial local potential,
and some likelihood of long run comparative advantage in technological develop-
ment, should be selected; in these industries technology imports should be
permitted where they are compiementary with and conducive to local technological
development and restricted where they are competitive with local developments.
Teclhnological development is an infant industry, with strong externalities,
wlhich accordingly requires protection.  There are numerous - and fairly well
known - promotional measures, including government subsidies, training and
education and so on. None of them have proved to be particularly effective on
1/
their own, but at both macro and micro levels a combination of protection and
promotion does seemn to have been effective.
Tuie system of patents and trademarks in effect gives stronig protection
to foreign technology as against local technolngy. By so doing, it also tends
I/  See the I.R.D.C. study into Science and Techniiology Policy IlistrumentLs
in  10 Countrie-i (Sagasti, 1978).



- xv -
to raise the cost of the foreign technology. Five-sixths of patents are held
by foreigners in LDCs, and of these, over 90% are unexploitec.  International
discussion and negotlation about the reform of the system is currently under-
way. Trademarks act as a powerful protective device to foreign technology because of
the marketing power they bestow. LDC trademarks are relatively undeveloped; they
formed half of new registrations in 1974, but this was substantially less than some
ten years before. Various reforms are possible, including complete outlawing
of trademarks, outlawing foreign trademarks onlv, requiring joint foreign and
local trademarks, and imposing a heavy tax on trademarks. The main problem is
to combine consumer protection with a r.duction of the role of trademarks.
Trademarks tend to be heavily concentrated in a few industries (notably
pharmaceuticals) and action could be restricted to one or two sectors, initially.
Any restrictive policies need to be accompanied by consumer education and pro-
motion of alternative methods of guaranteeing quality, e.g. by developing government
standards.
Local technological capacity in tiird world countries is of critical
importance: it is a vital part of the development process, it is necessary for
independence , to improve bargaining power in relation to the import of
technology, and to generate appropriate technical change. While
there appears to be a certain amount of technological choice today, the continued
concentration of technical change on advanced country technology is likely to
result in increasingly inappropriate techniques. Unless developing countries
undertake r. and d. in alternative directions, the choice of technology available
in the future will be increasingly circumscribed and irrelevant to the needs of
the world's poorest, The development of a continuous process of technological
change - new techniques and products - in an appropriate direction in the
developing countries is essential if the choice is to be widened.



- xvi -
Appropriate Technology
Technology recently developed in advanced countries tends to be in-
appropriate, in many respects, for many LDCs. This is because it is designed
to meet the needs of the advanced countries; it tends for example to be in-
creasingly capital intensive. Its use in poor countries involves a concentration
of investment resources in the modern sector, exacerbating differentials and
contributing to the problems of unemployment and underemployment. Products
designed for advanced country consumers are ill suited to meet the basic needs
of poor people. However, the high productivity of much modern technology and
the efficiency and economies of scale a- sociated with advanced country products
may make the technology the best choice, given the absence of efficient alterna-
tives. Appropriate technology consists of technology more in line with develop-
ing country needs and resouices; it consists of more labour intensive processes
in the modern sector, and the development of new and improved techniques and
products for the traditional sector. The promotion of more appropriate tech-
nology includes the selection of more approp--iate technology among known tech-
nologies (including "old" techniques from ad\anced countries and secondhand
machinery) and the development of new appropriate technologies.
The development strategy a country adopts strongly influences its
technological options. If a country's consumption patterns are similar to those
of developed countries, and if its main trading orientation is with developed
countries, then the main body of technology it uses in the modern sector will
need to be that of the developed countries. Consumption patterns are likely
to be similar to those of developed countries among countries with an unequal
income distribution, an "open" policy towards technological imports, and lack
of indigenous cultural factors which inhibit the demonstration effect.  Some



- xvii -
modification of technology is clearly possible - and indeed hbppens - even in
this sort of situation, but modifications are likely to be largely a matter of
time lags in adopting the latest advanced country technology, more labour
intensive ancilliary activities, and modifications rendered necessary by
differences in the size of the market.
With more equal income distribution and with obstacles (cultural,
economic or legal) preventing the domination of advanced country products, the
potential for appropriate and self-generated technology will be greater. Trade
in manufactured products with advanced countries has generally involved the
import of technology from advanced countries, at least in the initial stage.
Trade between developing countries is more likely to be consistent with the use
of locally generated appropriate technology. Appropriate technology consists
both in modification in "modern" sector products and techniques and in up-
grading "traditional" sector technology. Even in countries with advanced country
oriented modern sectors, there is potential for appropriate technology in the
traditional sector to help raise productivitv and incomes and meet basic needs.
Policies to promote appropriate technology consist in policies
determining the demand for (or selection of) different products and techniques,
and policies determining the supply of technologies. On the demand side, the
policies concerned are chiefly national; they include determinants of income
distribution and consumption patterns, tradint strategy, policies determining
control over investible resources by different types of decision makers, and
relative factor prices. On the supply side, they include the collection of
information about different technologies and its diffusion, and research and
development and the creation of new technologies.  Potentially, there could be
an important role for international institutions on the supply side assisting in
informatiJon collection and diffusion, and in fundinig appropriate research.



- xviii -
Technology transfer belLween developing countries could have an
important role to play in increasing developing countries technological
independence and bargaining power and in promoting appropriate technology.
There are some indications that such trade is expanding; third world MNCs
are developing as is third worlcl consultancy and trade in capital goods.
As with similar developments, there is a danger that new imbalances and
dependencies within the third world will arise as a result.
There is a complex system of relationships between past policy
towards technology, policy makers and policy making. Countries which have
been heavily dependent on foreign technology, particularly in the form of
foreign investment, find it more difficult to regulate it than those which
have relied on it to a lesser extent and have maintained an arm's length
relationship with foreign suppliers. Similarly, patterns of production and
consumption which are broadly inappropriate set in force strong forces making
for similar inappropriate choices in the future. The extent of current free-
dom of decision making depends in large part on past policies. Today's policies,
then, will partly determine future opions. tn this context, radical delinking
might be a tactica]. and temporary move to strengthen a country's position in
future linked relationships.



International Technology Transfer: Issues and Policy Options l/
Technology is knowledge - knowledge of how to do and make useful things.2/
At the heart of any form of economic activity from the least to the most sophis-
ticated lies the technology or knowledge of how the activity is carried out. Be-
fore initiating any economic activity then the first prerequisite is the acqui-
sition of this knowledge. But technology is not randomly and freely distributed
throughout the world, as would appear from some economists' models. Rather, as
part of the process of historical development, the development of technology
has been heavily concentrated in certain parts of the world, notably the developed
countries. Moreover, much of this technology has been commercialized with pro-
prietary rights acquired in it so that it is not freely transferred but is sold
commercially. The international transfer - in many cases involving the international
sale - of technology is thus an essential precondition for economic development.
Since technology changes continuously, with new or improved products, new mater-
ials or new uses of old materials, and new techniques of production, the need to
acquire technology is not once for all, but a continuing one. Thus the question of
technology transfer - how it is transferred, the terms of transfer and the effects
of the transfer - is at least as critical to economic development, as the ques-
tion of transfer of cap4tal resources. In recent years this fact has been widely
recognized, as shown by the many international conferences, resolutions etc. on the
subject.  This paper attempts to summarize sume of the main issues that arise in
connection with the international transfer of technology to developing countries 3/
1/ I am grateful for comments on an earlier draft to a number of people
in the World Bank: I would especially like to mention Howard Pack
and Larry Westphal.
/ Merrill defines it as 'skills, knowledge and procedures for making,
using and doing useful things.'
3   There are many problems that arise in connection with internal technology
transfer-i.e. the transfer of technology or its diffusion within a nation,
but these are not considered in this paper.



- 2-
I. Why Technology is Transferred
The international transfer of technology takes place when knowledge in
one country - which may have been developed there or acquired from somewhere else -
is communicated to people in another country, for use there. The communication
may occur freely outside the market, or may be a commercial transaction. Effective
communication of knowledge for economic activity is not normally just a matter of
communication of design sheets and formulae, but also involves an essential soft-
ware element, the communication of how to use the information, which may require
the transfer of skills, managerial knowhow and so on.l/ The sale of technology
occurs when some essential part of the package which constitutes the technology
is in the commercial possession of some agent, who will only part with it for a
price.  This 'possession' may take the form of monopolization of the desired
knowledge, which may occur naturally when it is first developed, and before others
have caught up with the technological developments, or may be the result of legal
restrictions protecting the owners of the technology, allowing them to sell it, and
preventing imitators, as with the patent laws and those related to trademarks.
In either case a monopolistic element enters the market enabling the owners of the
technology to earn some monopolistic rents - i.e. to charge a price in excess of the
2/
actual costs of communication.- The monopolistic element may pervade a much wider
area than that specifically covered by legal protection, because of the bundled as-
pect of much technology, so that some element of the bundle which is apparently free-
ly available without legal restriction is in practice protected by restrictions af-
fecting some other element in the bundle. The actual costs of communication are
1/ "Technology, used in this study, refers to the package of product designs, pro-
duction and processing techniques and managerial systems that are used to manu-
facture particular industrial products." (Baranson, p. 13).  This paper adopts
this definition, but does not confine attention to industrial products.
2/  The justification for commercialization of the market for technology is that it
provides the necessary appropriability to induce private R. and D.  Part of
the excessive price (over and above costs of communication) constitute normal
returns to R. and D.  and should not, therefore, necessarily be classified as
monopoly profits.



3-
by no means negl±giblef, Technology transfer may command a Drice to cover these
costs in the absence of any monopolistic elements in the market for technology.
Technology was defined above as knowledge of how to do and make useful
things. But in practice examination of the market for technology suggests that
in the process of commercialization of this market, the content of technology
transfer has become more complex than this. A major element in technology trans-
fer is the acquisition of the right to use certain trademarks and/or access to
certain markets and inputs. For shorthand we may describe the acquisition of
trademarks, and privileged access gained to markets and/or to inputs (which may
include an assurance that the firm will be kept up to date with later technical
developments) as marketing rights. These may be highly valuable to individual
firms in helping gain markets or inputs.  While they are often associated with the
communication of useful knowledge, and form a significant part of technology con-
'2/
tracts,- they are not themselves accurately described as consisting in-the communi-
cation of useful knowledge. Nonetheless, in the context of discussion of inter-
1/  Teece (1976) estimates the costs involved in transmitting and absorh-
ing all the relevant unembodied knowledge, and finds that these costs range
from 2 to 59% of total project costs.
2/ Trademarks were an aspect of 48% of technology contracts examined in Mlexico
and 58% in Argentina. But the estimates are very sensitive to the methodology
used. In a study of Brazil where each contract was divided into as many agree-
ments as there were contractual elements, trademarks formed 13.2% of the total.
But in Brazil trademark licenses are not allowed between parent and subsi-
diary company.  See UNCT.D (1977b) (paras. 1Z6 and 127)



national technology transfer, the acquisition of marketing rights forms an im-
portant element of costs and is aLso a significant aspect of motivation.  Dis-
cussion of the international transfer of technology thus covers both the communi-
cation (or sale) of knowledge and the sale of marketing rights.
The internationnl transfer of technology takes place when three ccnditions
occur. First, decision makers in one country wish to use a certain technology;
secondly, that technology is not available locally; thirdly, they believe it is
cheaper for them to transfer the technology than to reproduce it locally. These
three conditions explain the quite substantia. amount of technology transfer be-
tween developed countries.-/ But they are present to a much greater extent ir.
relation to investment decisions in developing countries, such that a very la-.ge
proportion of investment in developing countries involves international technology
transfer chiefly      from developed countries.  The three conditions apply to a
definition of technology which encompasses the marketing element just described.
Condition one: investors wish to use a certain technology. For some observers it
appears almost axiomatic that countries should wish to, and would benefit from,
use of the latest technological developments: "Whatever the source, the increase
in the stock of useful knowledge and the extension of its application are of the
essence of modern economic growth....No matter where these teclinological and social
1/ See the analysis of eg. Posner (1961 and 1970) describing the 'technology
gap' conditions which give rise first to trade, and subsequently to technology
transfer between developed countries. A full explanation of international
technology transfer should go further than the above, first explaining why
a decision is made to produce a particular product in a particular country,
as against importing it. Technology gap, product cycle, and import substi-
tution cum jumping tariff barrier hypotheses provide explanations of the
initial decision to produce. See eg. Posner op. cit., Hufbauer (1966) Vernon
(1966) and Elirsch (1967), and (1§76) SPRU (1972)



- 5 -
innovations emerge - and they are largely the product of the developed countries -
the economic growth of any nation delpends upon their adoption." (Kuznets, 1966)-
From the point of view of the soc4al. interest and long run development
prospects of developing countries, there are three reasons why one might questicn
or at least qualify this view.  First there are biases in the development of
technology: the characteristics of any technology are heavily influenced by
the economic and social conditions in the economy in which it is developed.-/
Thus technology developed for advanced societies often has characteristics which
are ill suited to much poorer economies: for example, technological advances in
rich societies tend to be increasingly capital-intensive, of increasing scale
of production and. designed. to produce products intended for high income consumers.
As we shall discuss more belcw, if imported unadapted into poor societies, these
characteristics have various undesirable effects. The Kuznets view of the un-
questionable superiority of the latest technology would cnly be invariably cor-
rect if this bias did not exist and technical advance were neutral with respect
to factor use and product design, raising the productivity of all types of tech-
niques equally, and enhancing the efficiezcy of all types of products equally.
aonetheless, recent technological advances do involve great gains in productivity
in resource use and this fact mar compensata for biases in characteristics, but it
does not do so Invariably. Secondly, there are sigaificant learning effects.
3I
The Japanese case-  4ilustrates the way in which restricted import of cechnology
may permit the local development of technological capacity. Again learning as-
pects vill be discussed =ore below. Thirdly, as already stated the technology
that Is transferred involves more than the mot recent technological advances; it
l1 See also Gerschenkuxof (1962) and Spencer (1970).
2   Tht  point is discussed at much greater length iX Stewart (1977)
Chapters l. and 3.
3/ Sew O:awa (1966) 'and UNCTAD (1978a),



- 6 -
also involves the transfer of marketing rights. While the knowledge itself may
be worth acquiring from a national point of view, marketing rights may not be.
These qualifications are made from the point of view of the national in-
terest of developing countries. They do not apply directly at the level of
the individual firm, where most technology decisions are made. It is not their
concern that amalgamatirg all the micro-decisions, the free import of advanced
country technology may cause biases in development patterns; individual firms
respond to this sort of consideration indirectly if differences in conditions
affect factor prices and consumer tastes. Long run learning effects are
externalities not allowed for in individual decision making while considerations
related to marketing rights enhance the individual decision maker's assessment
of the value of the technology in question rather than the reverse. For the
most part then, individual decision makers wish to acquire recent technology -
not necessarily the latest, the exact vintage depending upon technological
alternatives and the factor price/market situation in the country in
question. At a national level too, governments, keen to industrialize, wish
to have the 'best' technology, which is often identified with the latest.
Condition 2: the technology is not available locally. During the past two hun-
dred years technological innovations have been dominated by a handful of countries.
A study by O.E.C.D. (1970) identified 110 significant innovations in the twentieth
century. All emanated from developed countries with the U.S. responsible for
60%, the-U.K. 14% and German firms 11%. As'these figures indicate innovations are
not evenly spread among developed countries, but largely concentrated on a very
few countries. Only the U.S. and the U.K. are net creditors in terms nf rnyp1ty
paywents. Th U.S. accounted for nearly two thirds of the gross receipts of royalty
payments among the major recipientsl/ in 1977; the main European countries accounted
1/ From the data fund of the I.M.F. Countries included are the U.S., Japan U.K.,
Austria, Belgium, France, Germany, Italy, Netherlands, Sweden, Australia,
S. Africa, New Zealand, Nonay, Finland and Spain.  As a group) these countries
are in credit Mo the rest of the world, in terms of royalty payments, to
the tune of 1,065 million SDRs.  The U.S.' own net credit Is far greater
than this - 3,662 milliosn SRDs.



for nearly 30%, and Japan for 3%. "hese figures reflect the past history of
technological domination by a few developed countries. But much of this domi-
nation remains. The developed countries are responsible for 97% of world re-
search and development expenditure-,.  6 nations-/ employ nearly 70% of the
world's research and development manpower and spend nearly 85% of R and D funds;
only 6% of an estimated 3 1/2 million patents issued in 1972 were granted by
developing countries, and less than one sixth of those issued by developing
countries were owned by developing country nationals.  -/ With the exception
of Brazil and India, developing countries import a high proportion of their
4/
capital goods- ; over 90% of LDC plant and machinery imports come from developed
countries. In recent years there has been some increase in the technological
capacity of developing countries - an increase which, as is to be expected, has
been unevenly spread among developing counitries. This is indicated by rising
expenditure on research and development and, by some evidence of incipient ex-
5/
ports of technology by some developing countries; and by increases in the export
6/
of capital goods by some developing countries.-  But while these developments
are potentially significant, particularl, in relation to policy, they are of
relative insignificance in the general picture. The developed countries retain
a massive preponderance in technological innovation which is a fundamental
fact which must underly any discussion of technology transfer: it is this pre-
ponderance that is sonetimes described as technological dependence.-/
1- Annerstadt (1978). Expenditure figures are a bit misleading because of dif-
ferences in salary levels: developing countries account for 13% of world
scientists and engineers involved in research and development.
2/  USA, USSR, Japan, Federal Republic of German, France and the UK.
3/  UNCTAD, (1975b).
4/  See Maizels (1963) and Stewart (1977), Table 5.2.
5/  Described by Wells (1977), Diaz-Alejandro (3977) and Lall (1978).
6/    rt<; of machinery and transport equipnent were 37% of developintg market
economy cxpor1t; in .1976.
7/  See UNCTAD, (1976)



- 8 -
Despite the fact that most innovations occur in the developed countries, there
are - to an increasing extent - local LDC sources of technology. In the first
place, although small in relative extent, R. and D. in the third world is in-
creasing, and third world innovat:ion (which extends, as in all economies, well
beyond formal R. and D.) and adaptation is rising.-/Secondly, when a certain
technology has been transferred once, internal transfer - from the initial re-
cipient to other users - could, in theory replace international transfer. How-
ever, in practice both these potential sources of internal technology tend to
be underutilized in many countries. Much LDC R. and D. tends to be misdirected,
from the point of view of generating usable technologies.- /A considerable por-
tion of it is basic research, which nlay be necessary to build up a scientific
capacity, but which many observers believe has been overemphasized. Moreover,
there are weak links between R. and D. and local productive activities, with a
notoriously small proportion of total R. and D. being conducted by firms.- /It
is noteworthy that of five countries which have developed Technology Plans, four
specified explicitly that a significant weakness of existing efforts was weak
links with the productive sector.-4  Evei. where local research efforts do gen-
erate viable technologies, there is a strong/ tendency for these technologies
to be rejected in favour of foreign sources.  This rejection is largely due to
1/  The five countries which have Technology Plans are all planning to raise the
proportion of GNTP devoted to R. and D; significantly, to around 1%.  Most re-
cent figures collected by Annerstadt show a significant increase in R. and D.
expenditure in LDCs in 1973 as compared with earlier figures,  Adaption and
innovntion have been identified in Korea and Taiwan (see Ranis 1972 and 1973),
and in Argentina(see Katz (1978) and Maxwell (1978)).  See also other sources
cited in Lall (1978). While a certain amount of LDC innovation has by now been
established its quantitative significance is not clear.
2/  This was the view of the USAID Report (1970); Cooper (1972), Ilerrera (1972)
and Reddy in Bhalla (cd) (1979).
3/ See eg. Subrahmanian (1972), Carlsen (1975).
4/  UNC1AD (3978 b). The countries were India, Mlexico, Pakistan and Venezuela.
5/  See Kidron, (1965) Subrahmanlian (1972), Sercovich (1975), Lall (1975). Aurorni ~.nd
Morehouse (1974).



the market power bestowed by foreign trademarks, which is in turn engendered by
consumer belief that foreign 's best, and bv some bad experiences with local
technclogies.  Foreign technologies mav elso involve a greater marketability on
world markets, and in some case access to markets which have been cartelized by
foreign sources.   The second potential source of internal technology is the
internal transfer of techniques which have been initially imported. Both Japan
and China have pursued this as a conscious policy. In the inter-war period,
Japanese technology policy was summarized by the slogan:  "The first machine by
import, the second by domestic production".    In China, Baranson reports that
a UL.S. "process team was shown an ex.ct duplicate of its own - albeit, an eight-
year old model - catalytic cracker and platforming refinery of a 3,000-barrel-
a-day capacity apparently copied from one that UOP had previously sold to Cuba".
(p. 118). But Japan and China are exceptions. Until recently many countries
haVe followed a laisser-faire policy towards the import of techn-logy permitting
the duplicative import of technology. Restrictions in technology contracts limit
the possibility of internal transfer, while it is often in the interests of the
individual firm to acquire a foreign technology from the original foreign source,
despite the fact (or indeed because of it) that some other local firm has already
acquired it. Thus empirical investigations find that very similar technologies
are sold to the same country a number of times, and sometimes that the identical
2/
technology is sold a number of times. A study of DanislTforeign investment in
India found that every single technique transferred was already in use somewhere
.1 UNCTAD (1978a)
I/  See Neers9  (1975)



- 10 -
in India. The potential internal sources of technology a-e thus underused and the
extent of international transfer exaggerated, as a result of a laisser-faire
policy towards the transfer of technology, and the market power bestowed by
foreign technology.
Condition 3:  decision makers believe it is cheaper to import the technology
rather than reproduce it locally. Despite the real costs involved in technology
transfer, costs of local reproduction - particularly at the level of the in-
dividual firm - are likely to be greater. This is in part due to the lower
levels of technical sophistication and of R. and D. capacity in LDCs as com-
pared with DCs, and in large part to economies of scale associated with R. and
D., and particularly with its use.  Thus it is not surprising that once a parcicu-
lar technology has been developed, the reduplication of the development process
should be. more expensive than acquiring it. Moreover, given that many inves-
tors wish to acquire the marketing rights as well as the knowledge, and that lo-
cal R. and D. cannot generate this, foreign technology may be bought even when
local reproduction would be as cheap.
The three conditions necessary for the international transfer of.tech-
nology are thus very often present in investment decisions in third world countries
This is particularly the case when the market situation is such that foreign
trademarks enhance profitability, and governments pursue a passive policy
towards technology imports.



II. Transfer Mechanisms
Defining technology I/ very broadly to include all knowledge related
to econorm2ic activity naturally means that it encompasses a very wide range of
types of knowledge, and that there are, correspondingly, very many mechanisms
of transfer associated with it. In content, technology includes knowledge about
infrastructure and services, agriculture as well as industry; it includes speci-
fication of what is produced as well as how it is produced. The how of pro-
duction is not merely a matter of technical specifications but also includes
managerial techniques, forms of organization and so on.
Much technology is transferred informally - through reading books,
journals, sales literature, through personal contacts in meetings and conferences
and so on and through th2 movement of trained people from one job to another.
Within more formal categories of transfer, a distinction has been made2/ between
direct and indirect mechanisms: direct mechanisms are those used when the
recipient enterprises are in direct cont&ct with the suppliers of technology;
indirect transfer occurs when a company in an advanced country plays an
intermediary role packaging the technology for the developed country. In
practice, this distidction is too firm: there are a large variety of types of
transfer involving varying degrees of packaging.
1/ The literature abounds with definitional discussions of exactly what is
involved in the process of technology transfer, and of the various elements in-
volved, e.g., the distinction into "general technology," "system-specific"
technology, "firms-specific-technology" and so on (Quinn 1969).  See the dis-
cussion in Cooper and Sercovich (1971). Here we avoid this type of discussion.
2/ Cooper and Sercovich.



- 12 -
Direct forms of transfer include direct contracting of individual experts
and consultant companies, engaging engineering design and plant construction
enterprises, training nationals for specific production prolects, technical
information activities and transfer of the process technology embodied in capital
goods bv importation of equipment purchased directly from machine manufacturers.-/
Indirect mechanisms range from the completely packaged in the form of dire(
investment overseas in a wholly-owned subsidiary, through joint ventures, turnkey
arrangements and license and management contracts between independent parties.
License agreements encompass a considerable range of contracts which vary
in what they cover, in the restrictive provisions involved, and in the form and
rate of payment. Some license agrer.-nents contain provisicns fcr some equity
participation; some include management contracts, or the right of the licensor
to appoint managers and/directors; licensors may sometimes appoint quality control
experts, and occasionally control marketing throubh wholly-owned subsidiaries.
License agreements often contain restrictive clauses in relation to the rights
of the licensee to export, to conduct and/or use independent research, and tie-in
clauses, whereby the licensee has to purrhase inputs from the licensor, and so on.-l
The determination of the mechaniism of transfer in a particular case
is the outcome of the willingness of the supplier of technology to supply the
technology in a particular form and the desire and ability of the recipient to
acquire it in a particular form. Generally, indirect mechanisms tend to be
adopted where a country lacks the capacity'to undertake direct purchase, where
proprietary technology is involved which *ill not be released, or where (for
marketing or other reasons) the recipient wishes to acquire trade marks.-3  But,
1/ Cooper and Scrcovich.
2/ Cooper and Scrcovich provide a comprehensive discussion.  A considerable mnuount
of researclh has been done examining the nature of transfer mechanisms in par-
ticular countries - see e.g., UNCTADI(3975a), Asian Productivity Organization
(1976), S.P.R.U. (1972)., IJNCrAD (1974), Mtarga Institute (1975), Sercovich (1974),
Vaitsos (1974).
3/ S.P.R.U. (197'), pp. 19-20.



- !.3 -
as just suggested, there remains considerable choice as zo the nature of the
indirect mechanism.
W4here the interests of the rec4pient country differ from those
of an individual recip-ent enterprise, the outcome may depend on who is responsible
for the negotiations - government or enterprise - and how far enterprise freedom
is restricted by government regulations.
The quantitative signilficance of different mechanisms varies as
between countries, as between industries, and over time. Country variation,
illustrated in the Table below for 1970, is partly a matter of national policy
towards technology transfer and private overseas investment, and partly of the degree
of development, and in particular the technological and managerial sophistication
of the country concerned. The less developed - in terms of this type of sophistica-
tion rather than per-capita income - the more packaged the transfer is likely to be.
because the country may lack the capacity to put the package together itself, and
because it may lack the bargaining power to insist on so doing.-/ But independently
of stage of development, the question of aLational technology strategy is of critical
importance: some countries, such as Japan and many of the socialist countries,
only permitted license agreements or direct transfers; many others have encouraged
1/ But, taking the ratio of intra-firm tQ total technology payments to U. S.
companies as a guide to the degree of packaging, in 1976 developed countries'
packaging (with the exception of Japan) was not dissimilar to that of the
developing countries (see footnote, p. 18). This may in part be due to the
non-regulatory/interventionist policies of most developed market economies
permitting packaged transfers. It also may be due to the fact that as
countries become more developed, the technology they need to acquire from
overseas becomes more sophisticated, and the tendency for packaging in
sophisticated industries to be greater than in less sophisticated outweighs
the effects of greater technological sophistication in reducing packaging.



- 14 -
direct foreign investment; I/ while a few follow an intermediate policy
(e.g., India).21 The content of licensing agreements also varies between
countries: as is to be expected, countries with little independent managerial
capacity are likely to have agreements which include management contracts.
This was illustrated in a study of technology transfer to Ethiopia,-/ which
found that the majority of agreements contained management contracts.
The nature of the transfer mechanism adopted varies with the industry.
The more technologically sophisticated, the more difficult it is for countries
to rely on direct transfer. The extent and nature of property rights over
technology vary with the industry, as does the significance of trademarks.    UNCTAD
(1975a)  classified industries into "modern" (requiring heavy research and develop-
ment and with a recent record of tetchnological innovation), "traditional" and
"other" - both of which use well-established technologies.  They found that for
13 countries, 57% of the contractual arrangements were in the "modern" classifica-
tion in manufacturing, but there was a wide dispersion between countries. Relatively
sophisticated countries, like South Korea, had as much as 80% of contractual
arrangemeents in the modern industries, presumably indicating that direct transfer
occurs in mature industries with well-established technologies; but in Dahcmey,
80% of contracts were outside the research intensive sectors.  A comparative
study of pharmaceuticals, chemicals and electrical engineering in Latin America 4/
found significant differences in preferences of supplying firms for different
mechanisms of transfer. Broadly, these differences in preference were reflected
in the actual arrangements. In the pharmaceutical industry, there was a strong
preference for transfer via wholly-owned subsidiaries, and also for patent
1/ As shown by the massive tax incentives provided.
2/ See S.P.R.U. (1972) Ozawa (1966). Subralmanian (1972), Neers6 (1975), UNCTAD
(1978), Mytelka (1977), UNID0 (1978).
3/ UNCTAD (1974).
4/  S.P.R.U. (1972)



- 15 -
iable 1
OWNERSHIP CHARACTERISTTCd OF TIE CONTRACTING ENTERPRISES
IN THE TECHNOLOGY JIZECEIVING COUNTRY, EITD 1970
% of Contracts in Enterprises with
Majority Foreign   Minority Foreign       Wholly Nationally
Country                Ownership           Ownership               Owned
Cyprus                     48                  44                     8
Colombia                   45                  12                    43
Brazil                     36                                        64#
Sri Lanka                  29                  42                    29
Peru                       28                  19                    55
Rep. Korea                 13                                        87#
Pakistan                   12                   8                     79*
India                       3                  12                    85
Yugoslavia                  -                   5                    95
Source: UNCTAD (1975) TD/B/AC.ll/OjRev. 2, Table 3.
# Minority foreign ownership treated as wholly national.
* Specifier, as "100% government controlled."
protection of the technology transferred. In chemlcr.ls, the majority of firms
preferred joint ventures or pure licensing arrangements - the preference being
greater the smaller the size of the supplying firm. In electrical machinery -
which on the whole is a mature technology which has been widely dispersed - by far
the largest nuamber of preferences was for licensing agreements. In a study of metal
working and chemical firms in three Andear Pact countries, Mytelka (1978) distinguis-:.
between technology transfer via licensing and direct acquisition through experience
of personnel, non-negotiable means (e.g., copying), the purchase of machinery and
processes, journals end professional meetings.  "Ownership structure, product



- 16 -
sector and firm size...correlated highly with the decision to license."
Foreign firms clustered in industrrEs with complex and volatile technologies; they
together with mixed firms overwhelmingly licensed their technology, even wilere
the technologies were simple and mature.  Nationally-owned firms were domninant
in simpler (technologically) industries: 61% of nationally-owned firms acquired
technology through direct means, as compared with 17.6% of foreign firms, 12.5%J
of State firms and 25% of mixed fi-ms.  The firms were asked why they decided
to license:  51% gave braLvd name acquisition as a reason; 47% the complexity of
the technology, 30% prior L elationship to the technology supplier, 26% cheapness,
and 17% 'other' which in(ludes advanLageous bargaining conditions, reduced costs,
few restrictions. Returning to the earlier distinction between that part of
technology transfer which eonsists in the communication of useful knowledge,
and marketing rights aspects, it appears that for this sample,of the total
reasons given, one-third v!.-e of the knowledge communication typ' (complexity and
cheapness), while twe- th:, i, were other reasons, much of which could be classified
as marketing rights.   Iziis does not mean, of course, that licenses obtained for
e.g., brand name reasons, did not bring with them useful knowledge and quality
guarantees which it would have been costly to obtain in other ways.
A study of sources of technology in S. Korea -/ found that formal
mechanisms of transfer - licenses and te.hnical agreements with foreign
technology suppliers - accounted for ornly a small proportion of tne total.
Form,al mechamisms from foreign suppliers were of greatest significance in the
modern sector - accounting for 21% of all important technology sources. This is a
smaller proportion than other sources of foreign technology, which included tech-
nology embodied in Korean labour anad rNanagement (18.6%), foreign suppliers of
capital equtipment or raw materials (8.2°') and foreign buyers (6.6Z). In
1/ Pursell and Rhee (1978).



- 17 -
the traditional and resource based sectors formal mechanisms played a sig-
nificantly smaller role.  While there may be special factors in S. Korea
accounting for some of the importance of informal mechanisms - for example,
buyers can only play a significant role in an export-oriented economy - the
study suggests we may often o"leremphasize the role of formal mechansims in
technology transfer.
In recent years, there have been various forces making for less packaged
forms of transfer - both in recipient and supplying countries. Growing industrial
sophistication among recipients has increased their capacity to provide elements
of the package for themselves; but they have also moved into more sophisticated
(technologically) industries, where packaging is more dominant. Increasing
political and economic sophistication, combined with the accumulation and spread
of knowledge about the technology transfer process, has increased countries'
intervention in technology transfer, sometimes involving a search for unpackaged
alternatives. -/ In some of the supplying countries, government regulations
curtailing investment overseas, and fears of the riskiness of overseas investment
in the light of moves among recipients to reduce equity participation, have
contributed to more arms length forms of technology transfer. 2/
The evidence, however, is rather mixed. Baranson claims that "the
most dramatic of the new strategies has been the adoption by some U. S. firms
of an explicit policy to shift from equity investment and managerial control
of overseas facilities to the sale of technology and management services as a
direct means to earn returns on corporate assets." 3/  He cites examples in five
major research-intensive industries. In the parts of the world where policies
have been most explicitly directed against packaged forms of technology inflow,
1/ See UNIDO (1978) and Mytelka (1977).
2/ Baranson (1978) discusses the reasons, particularly from the point of view of
the supplying countries, at greater length.
3/ Baranson, op. cit., p.5.



- 18 -
e.g. India and the Andean Pact Countries, there is evidence of a move towards
less packaged forms. 1/ But as Table 2 shows for the world as a whole, and
for developing countries other than Latin America, there has been a significa-nt
2/
increase in the proportion of U. S. technology receipts which are intra-firm.
Tlih  U. S. accounLed for 7:?2 of gross receipts for technology among the U. S.,
Japan and Europe in 1969; the proportion has declined to 68% in 1977. 3/
It appears that U. S. technology supplies are more often intra-firm than supp:ies
from other countries. 4/
No definite conclusions can be drawn on the basis of this evidence -
which is not surprising, perhaps, given the many factors that contribute to the
overall breakdown, including recipient country policy, industrial composition,-5/
6/
country source of supplies, etc. But it does appear that there are a range or
alternative mechanisms for technology transfer in many industries.
l/ See Mytelka (1977).
2/ As disctussed in detail below, royalty payments are not necessarily an accurate
guide to payments for technology transfer.  Over time, changes in the envirron.:ent
may leac' companies to change the form (proportionately) in which they receive
differel t forms of payment. This may be one explanation of the declining iatra-
firm raLio for Latin America, whlere the Andean Pact prohibited the payment of
royalties between subsidiary and parent companies.
3/ These figures are for receipts from technology transfer to all countries.
Evidence for payments by particular developing countries also suggests the
declining role of the U. S.:  for example, in South Korea, the U. S. accounted
for 40% of technology imports for the years 1962-66, and 217 in 1973-77; 'apj7
accounted for 33% in 1962-66 and 53% in 1973-77. In Peru, the U. S. and Canaca
accounted for 60% of the royalty payments in 1971 and 35% in 1974. A study of
license agreements in the Philippines for agreements hefore 1970, showed that
67% were with U. S. firms, 8% with Japanese firms; for the period 1971-73, 34%'
were with the U. S.  12% with  Japan. (UNIDO, 1978).
4/ This is suggested by the fact that intra-firm payments for particular countries
to all sources are substantial'ly lower than those shown for U. S.  payments:
e.g., for Brazil intra-fi-n payments aere 52% of total payments in manufacturing
1965-70; in Argentina they were 42% in 1972. (Chudnovsky 1978).
5/ See Chudnovsky  (1978) ,Table 2 for variations in the ratio of inter-firm
to intra-firm royalties according to industry in Argentina.
6/ Tlhe proportion of total. payments of royalties and management fees to U.-S.
companies, w7hich were int-ra-firm, rose sharply over the period 1960 to 1976
in the de.veloped countries, as well as the 'other developing':  In Japan,
it rose from  13 to 50%, in the U.K., from  56  to 84%', in Western Europe,
from  44  to 82%, Canada, 80  to 94%, in Australia, New Zealand and South
Africa fromn 65 to 84Z.



- 19 -
Table 2
PROPORTION OF TOTAL U.S. RECEIPTS FROM ROYALTIES
AND MANAGEMENT FEES WHICH ARE INTRA-FIRM
World              Latin America         Other Developing
1956                     64                      88                     n.a.
1960                     62                      83                      72
1972                     76                      85                      80
1976                     82                      82                      90
Source:  1956 - Hymer (1976), Table 2.2.
Rest  - Chudnovsky (1978), Table 1.



- 20 -
III. Significant Dimensions of Technology Transfer
Objectives of LDCs in relation to technology transfer have evo)vpt4
in response to experience in relation to technology transfer over the last 2c years.
Initially, keen to industrialize rapidly and thereby, as they thought, acquire
wealth and economic independence, most countries main aim was to maximize the
quantity of technology transferred. Hence, they introduced a host of incenti:.es,
such as tax incentives, duty drawbacks, protection, and the provision of infra-
structure, designed to encourage the inflow of technology from overseas, main.l
in the form of private overseas investment. -/  In other respects, a broadl-y
laisser faire policy was followed with respect to technology transfers so that it
was left to individual firms to decide the form and nature of the technology
transferred. However, as the consequences of unrestricted inflow became apparent,
new objectives evolved: the laissf,r-faire policy has therefore given way to a
much more interventionist strategy at both national and international levels.
First, the high and rising cost of technology transfer drew attention
to the neet to introduce policies which would improve the terms of technology
transfer.  This was initiated by the Andean Pact group of countries, then UNC:A.D, a,:
by now the need for policies in this area has gained very wide acceptance amo=g
LDCs.
Secondly, in many countries a dualistic pattern of development was
associated with the unrestricted import of advanced country technology. The
capital intensity of the investment, the heavy underemployment, and the elitist
1/ See Lent (1967), Reuber (1973) and U.N. (1972) for a description of the ir.-
centives.



- 21 -
consumpcicn patterns  :n many counttie3 led to emphasis on the need for appropriate
technology:  i.e., tecinology In line with the needs and reso'irces of poor
countries.  Although the need for appropriate technology was largely 4nit±aced
in the advanced countries, it has achieved increasing support among LDCs.
Thirdly, the dependency school of development suggested that neither
political independence nor industrialization was sufficient to achieve economic
independence, and that at the root of the dependency relationship exhibited between
periphery and center lay technological dependence. A more modest approach. which
does not rest on acceptance of the whole dependency analysis, shows that techno-
logical dependence did indeed impose restrictions on the freedom of decision
making that were scarcely compatible with economic independence. Hence, the
creation of independent technological capacity became an aim.
Fourthly, it was increasingly felt that unrestricted imrort of techno-
logy inhibited the development of local technological capacity which in turn
reduced long-run development prospects and engendered further technolcgicaL
dependence.
The new' objectives towards technology transfer, which have been
accepted and pursued with varying degrees af enthusiasm, are in many ways in
conflict with the initial objective of maximizing the inflow.  /  They each
suggest a much more interventionist policy on the part of governments, and in
general a more restrictive and selective policy towards technology transfer from
abroad. They each support (from different points of view) the need to develop
1/ See the findings of the feasibility study into "A New International Mechanism
for Appropriate Technology." (1978).
2/  There is a (non-coincidental) parallel between chpnpin. 0j'4NC    3 S tr%rards
economic development and those towards technology transfar.   Itc  the
dethronement of COP has gone, after a timelag, a corresponding detrocking of
associated (instrumental) objectives such as maximizing technology transfer.



- 22 -
Third World technological capacitv. In other respects, the new objectives are
sometimes in conflict, and often d±ffer 4n emphasis and policy,r conclusions.
In the rest of chis paper, we examine the policy opcicns indicated by the new
objectives.  The paper takes four d±mensions of technology transfer:
1. the terms of the transfer, or the cost;
II.  the effects of che transfer on 'ndeoendence of d4ecision nak4ng;
III.  the effects of the transfer on local technological capacity, or the
learning effects of the transfer; and
IV. the characteristics of the transfer.
Each of these corresponds with one of the 'new' objectives described.  The
paper considers past experience in relation to each of these dimensions, and
policy options suggested (and in some cases adopted).  The next sections consider
the first three dimensions, which are grouped naturally together in terms of
policy. The question of characteristics raises somewhat sep"rate issues and
is dealt with in Section VI.



23 -
IV. Costs of Technology Transfer
(i) Actual costs: It is difficult to ascertain actual costs of
transfer. Apart from the fact that data is weak and sparse, the packaged
aspect of much technology transfer and the indirect costs imposed raise
problems. Where there is any degree of packaging the supplying firm is,
by definition, supplying more than one service. This means that the firm
has some choice in how it receives its return for each element of the package -
e.g., whether it receives payment for technology as royalties, as profits,
or through the transfer pricing mechanism. Declared payments nominally
attributed to any one item may therefore bear no relationship to the real
charge levied. To get at the total charges levied is lifficult; but with
rough and ready means, estimates have been made. But this does not solve
the problem of how to allocate the items among the elements of the package.
This is peculiarly difficult in connection with technology transfer because
there is often some monopoly element involved, enabling the total charges levied to
exceed the competitive charges. But the source of monopoly is not always
the possession and sale of the technology; it may involve access to markets,
or raw materials or so on. However, if we define the technology transfer
to include marketing rights, as we have above, then it can be claimed that
in large part monopolistic charges may be attributed to the technology transfer.
But we should not include, as many seem to have done, competitive payments
for other services which form part of the package, as costs of technology
transfer.
These problems, that arise from the packaged nature of technology
transfer, are most acute for the most packaged torms of transfer - viz., that
of the TNC to wholly owned subsidiaries. But joint ventures also have



- 24 -
considerable scope for manipulating payments. Wholly owned national
companies often receive technology and inputs from the same source -
again it is not possible to know to what extent particular forms of
payment represent the payment for technology.
A further problem is that restrictions on activIties form a
normal part of many technology contracts. But these impose costs, which
can be substantial.  For example, tied sources of inputs is similar to
tied aid in increasing costs, but worse because with aid tying normally
takes the form of tying to a particular country whereas with technology
transfer the tying is to a particular firm. Even for aid, the costs have
been estimated to be as much as 30% of the cost of the equipment. -  The
costs are likely to be greater for private technoloRy transfers. Other
forms of restriction also impose costs which are difficult to estimate, such
as restrictions on exports and on local innovation. The extent of these
restriction has been established by a number of studies, but not their
costs.
In view of these problems, it is perhaps not surprising that none
of the (quite numerous) studies on technology transfer to developing countries
have come up with satisfactory estimates of costs.
The most systematic estimates of direct costs were mrade Dy UNCTAD
(1975). For 1968, they estimated -that direct costs to LDCs of technology
transfer were around $1,500 million for royalties and consultancy payments,
which was equivalent to 5% of non-oil exports and around 0.5% of GDP. They
1/ Bhlagwati (1968).



- 25 -
estimated that thiese payments were likelv to grow et around 20% p.a. on
the basis of questionnaires to recipient countries. However, the technology
receiving countries expected a lower rate of increase. If we assume that
in fact technology payments grew by 15-Zo p.a. from 1968-1977, they would
amount to  i5,Z75m. in l977.-   Table 1 shows the UNrUfAD estimates togerner
with technology related payments. As suggested above some element of these
related payments are likely to be due to technology.
In addition to these direct costs, overinvoicing of imports and
underi.voicing of exports represents an additional cost. The very substantial
proportion of international trade that occurs within the transnational corporation
gives scope to these practices.
1/ The data are derived from questionnaires to recipient countries. More
recent data collected by the IMF is too incomplete to provide much guidance.
It excludes management/consultancy and the coverage is very incomplete.
The estimates for royalties are far lower than those of UNCTAD (generally
of the order of 1/2% to 1% of exports). It is likely that a good deal of
intra-affiliate payments have been omitted in the IMF data.  However, the
Fund data suggest that the UNCTAD estimates overstate the costs. According
to the Fund data, gross royalty receipts among the main technology suppliers
from all countries grew by 99% p.a. between 1969 and 1977. Net receipts
grew by 10.9% p.a.



- 26 -
TABLE 3
Direct costs ot transrer ot Iccltnology in comparison w ith ofiler relevant
forcign cxchange fl,j%N s of devclopinig countirics, 1968a
Proporti-bn of i,ect
PVatte          pay,,,lcra lop Irinilef
(miltios of           of tri,01j,l 'ry
Flows                                   do/Irls)              (per e(nt)
Outflovs
1. Direct  paYji, nnts  for  transfer  of  techinology
(palents,  licensws, know-how, tradenmarks, and
mana,:emcnt and othcr teclinical scrvices)                    1,500                   100
s. Tcchniology-rclatcd paymcrnits:
(aJ  imports (c.i.f.) of machincry and equiplicnt
(excluding passcger vcihicles) and ol chemiicals        18.420                     8
(bJ P'iofit on direct forcign investment (excluding
oil-producinig countrics)b  .................             1.721                    87
3. Service paymcnts ont cxternal public debt                     4,022                     37
Inflowr
4. Non-petrolcum  cxports (f.o.b.) .29,350                                                  5
5. Total oflicial flows.                                         6.710                     22
6. Direct foreign investmencit (including rcinvested
carninl2s) .2,700                                                                       56
Zou  es:
Line 1: 1 .ZCTAD secretariat estimates (sce text).
Linc 2 (a): United Nations, .Mlonthly lBulletin of Statistic . vot. XXVI. No. 7 (July 1972).
Line 2 (b): "'the outfloNv of financial resources fronm  developing countries: note by the UNCTAD
secrctariat" (TDIl I 8/Supp.S), loc. cit.
tine 3: lBRD/IDA,.AtrnrialReport, 1972.
Lincs 4, s anrd 6: UNCI'AD, Ilaindbook o,f Itiertiarrional Trade anid Decilopnmenr St,:trltic:. 1972
(Uniited Nations pubhication, Sales No. Lt/F.72.tt.D.3).
D Data do not include Southern Europcan countries.
b Including oil-producisng countries: $4,934 millIon.



- 27 -
It is estimated that roughly one-third of total US exports were
on an intra-firm basis in 1970.1/ In manufacturing three industries accounted
for 70% of that trade: transport, equipment, non-electrical machinery and
chemicals. Fifty three percent of manufactured exports to developing countries
was intra-firm, and for developing countries other than Latin America the
proportion was 65%.2/  Intra-firm trade is here defined as trade between
a parent and a majority-owned affiliate. If joint ventures were included
the proportion would be higher. A detailed study of Brazil and Mexico showed
that the intra-firm share of exports was high and growing: in 1960 it was 68% and
54% respectively; in 1972, 73% and 82%.-/ Imports from the parent company
accounted for 50% (Brazil) and 58% (Mexico) of the affiliates' total material
imports in 1972. Again, intra-firm trade was found to be particularly concentrated
in technologically sophisticated industries. For the UK, a survey suggested
that about one-quarter of all exports were intra-firm, and one-quarter of
recurrent imports of affiliates in developing countries were supplied by other
units within the corporation.  Data for Norway and German firms also show a
significant proportion of trade is intra-firm; generally speaking the proportions
(of both imports and 2xports) are lower than for US firms.-/
1/  Lall (1973) and UiCTAD (1977c) summarize the evidence.
i/ Table 1 UNCTAD (1977c).
3/ Newfarmer and Mueller (1975) quoted i.n UNC'TAD (1977c).
4/  See la!CTAri (1077c).



- 28 -
The evidence thus shcws that a very significant proportion of trade
is intra-firm, giving rise to the possibility of a quantitatively significant
element of costs of technology transfer being disguised through the manipulation
of transfer pricing.   Intra-firm trade tends to be highest in the
technologically sophisticated firns where the costs of technology transfer
are likely to be greatest. Intra--firm trade appears to be of substantially
greatest significance for US companies. However, the US is by far the largest
source of technology transfer, although its predominance is declining.l/
The existence of a high proportion of intra-firm trade is necessary but not
sufficient to indicate the existencL of transfer price manipulation as a
sizeable source of income transfer between countries.  There are obvious
difficulties in getting hard evidence on the existence and extent of such
practices. But a growing body of evidence has now acctimulated2! showing that
in some industries, it forms a significant element in internaticial income
flows.
One approach has been tc, try t:' establish the extent of such practices
indirectly by regression estimates. Kopits (1976b) found he was able to explain
the breakdown of income transfers as between royalties and transfer payments,
statistically, by the tax incentives for the different types of remittances.
But he was primarily concerned with developed countries.  An econometric analysis
of export prices in Latin America suggested underpricing of exports of, on
average, 40% of the value of exports.3/ Other indirect evidence has been less
conclusive.
_/ See page 18.
2/  Summaries are contained in Lall (1973), Knnits (3q76P) and 1TNCTAD (1077c)
3/ Morgenstern and Miller (1974).



- 29 -
Direct estimates have been obtained by comparing J.ntra-firm prices with
-market prices. Detailed studies have been done for Colombia and Greece - when
the government instituted machinsry to check on transfer pricing practices of
international firms. In Colombia between 1967 and 1970, overpricing was great-
est in pharmaceuticals - "the absolute amount of overpricing for the foreign
firms studied amounted to a figure of six times their royalties and 24 times
their declared profits':-/ overpricing was greatest for the pharmaceutical
industry, (on average 155%) where the fereign exchange losses due to overpricing
of inputs were estimated to be as great as the total industrial royalties paid
by all industrial sectors for technology. In electrical goods overpricing was
estimated to be 54%, in rubber goods 44%, and chemicals 25%.
In Greece-/ two groups of products were studied.  In metals, metal products
and minerals a sample investigation showed overpricing of imports of between 5
and 88%, with a weighted average of 19.4%; for chemicals the range was between
12 1/2 and 229%, with a weighted average of 34.5%. In the first group of prod-
ucts 95% of the underpricing was the responsibility of foreign owned firms; in
the chemicals groups it was all in the foreign owned firms. For the mineral
-product group the total foreign exchange cost was estimated to be about 2 1/2
times the size of declared profits. The Greek study also investigated
three export products and found underpricing ranging from 8.3% to 16.9%. The
extent of underpricing was equivalent to 35% of the preceding year's profits for
a 90% foreign owned firm, 26% for a joint venture and 13% for a locally owned
1I  'aitsos (1971).
I/ Roumeliotis and Golemif (1978).



- 30 -
firm with foreign connections.
Evidence of substantial underpricing of imports in the pharmaceutical in-
dustry has been obtained in India, Ecuador, Peru and Chile.-/  The evidence for
extensive use of transfer pricing to secure international income flows has been
long established in the petroleum industry. -2
The evidence collected so far indicates that transfer pricing forms a
significant source of international income flows in intra-firm transactions -
at least as great as declared royalities and often as great as declared profits.
While the evidence, in terms of cotutry coverage and of magnitude, is grea,est for
the pharmaceutical industry, the practices are not confined to this industry but
extend (in lesser amount) to all irLdustries examined, such as rubber and mineral
products, as cited above. In view of the strong incentives that firms face to
3/
price in these ways (which include the tax system and restrictions on remittances-
the results are not surprising.
It is difficult to assess with fairness and accuracy how far these prac-
tices inflate the costs of technology transfer - partly because the evidence
though growing is still very incomplete, and partly because, as alre..dy suggested,
it is difficult to know how much of' the additional cost should be attributed
to technology transfer and how much to other factors. However, without trying to
form an accurate estimate certain conclusions follow:  first, the declared
royalty payments cannot be taken as a guide to the costs of technology transfer.
Some element of the income transferred as overinvoiced imports/underinvoiced
exports should also be attributed to technology transfer. In so far as it is
technological dominance (including marketing rights aspects) which gives rise
1/ See 1UNCrAD (1977c) paras. 94 ro 98.
2/ Jenkins and Wright (1975).
3/ The various motives are summarized in Vaitsos (1974), Lall (1973) and Kopits
(1976a).



- 31 -
to the market domination of a few firms and the possibility of transfer price
manipulation, then it is correct to -attribute the inxcome so transferred in
large part to the process of technology transfer. Secondly, the form or mechanism
of the technology transfer influences the extent to which this sort of inter-
national income transfer is possible. Manipulation of transfer prices can only
occur, in fully fledged form,-where trade is intra-firm. But some possibilities
exist in less packaged forms than the wholly owned subsidiary. Joint ventures,
indeed, may have as great a possibility, and a greater (at least from the point
of view of the foreign partner) reason for transferring income in this way. But
the (limited) evidence suggests that joint ventures pay slightly less for inputs
than wholly owned subsidiaries.-   .ae possibilities are clearly less for locally
owned firms, but even here where they have strong connections with overseas firms,
or where purchases of inputs (and/or access to markets for output) are effectively
tied to the foreign firm, part of the costs of technology transfer may appear as
inflated prices of imported inputs. However, both a priori reasoning, and the
limited empirical evidence on this, suggest that the more packaged the transfer
mechanism, the greater the flow of interiational income that occurs via transfer
pricing. Thirdly, irrespective of the total sums involved, the existence of
sizeable international income flows associated with technology transfer, in the
form of transfer price manipulations, suggests a number of directions for policy.
These will be discussed below.
Indirect costs   as already indicated the process of technology transfer often
imposes restrictions on the activities of recipient enterprises which impose real,
albeit indirect, costs. In the case of wholly owned subsidiary these restric-
tions are often implicit. But in transfers between less closely associated
enterprises, they are explicit elements in technology contracts.-/
I/ Stopford and Wells (1972) p. 161 - 162
21  UNCTAD (1975a) found that in both India ana the knilippines provision for
tied purchase of inputts was substantially greater in technology transfer
between indcpendcnt enterprises (20Z in India, 58% in the Plillippines) than
between associated enterprises (10% for botlh wlholly owned subsidiaries and
minority equity in India; 9X for a wholly oxrned subsidiary and 25% minority
equity in Llte Philippines); (Table 6). Vaitsos (1971) found similar differencea.



- 32 -
Table 4 provides a useful summary of the sort of restrictions to be found.
TABLE 4
Patietri or tinit-violsm o)n ;acvvss tui icchniiigXy by dlevntillingi coultirics
A' r;'p*s at I, wv  rije,    """" Ii"t': v J IIr ' .1 csfteid li,m,it,gltij
Type of fit f,i*fhin                     Yes                               No
1. Tied pI iuhjscS or imapoitcd    Argenirn;a. Cthile. C"yprus. Fcu.io  Requblic of Kurca
in put, *cetimpinvolI .a ad pairc dewf . G   r;c   *rI  .\l Ml 1.1. Mc ic3.
parts                         Njigrij. 1.'k;S uI, l'eru, Sri
Lanka, Turkey
2. Rcstii0jun of ex iorts (totil  Afgj nlin.a. Chile. Cyprru.     Sinjn porc
prolubilion. piarliIl lilill-  I Wuldor. Greece. I r at. Ma II
ta(Iml. gCoFrapI:cal           . Mexico, Niw-rar. Vikiun,
constraint)                   I'ctu, Sri Latika, Turkcy
3. Requirnem.nt orguirinaces   Cylprus, Nigeria, Turkuy            Grcccc, Iran, MIalta, .'lcxico,
3apinit clhmjtae% in taxes,                                     SingJPorc
tariffs nI\d c\chanec ratCs
aiffectig prolits. royaltics
and icinitt;anecs
4. Limitation of comepeting
supplies by:
Ja) rc!tuiction or            Cyprmss. Grcccc. Mexico.           Irani, Malta, Pakistan, Republic
competinllg t:npor%.       Ni;rCri3, Peru                    of Korea, Singapore, rurkCy
(bJ preventitig com,-ctition    Greece, %Ijita, Mexico           Iran, Nigcria, P.akiNtan,
for loctl rxsources                                          Rcpublic of Korea, Singaporc
(cj obtainirng ocni putenits    Ecuador, Malta, Nigeria         Greccc, Iran, Singapore
to clinil:eC competitors
5. Constraints limifing thc
clynaniic effects of the
transfcr
pJi CXCcSMiVC usc of           Argcntina, Malta. MIexico         Silngaipore
expatriJte personnel       Nigeria, Plcru, Turkey
(b) discouragCement of thc    Argentina, Ecuador, GrecCC,
development ol Iocal       Ntaltw, Mexico, Migeria,
tcchni:jI and ruswarch     Turkey
and dcvAlopmcnt
cap.ibtlities
The first two items are of greatest relevance to the question of costs
of technology transfer.
Detailed studies of technology transfer have revealed that restrictions
1/
are             widespread.  The SPRU-study of Latin America found export restrictions
1/   SPRU  (1972)



- 33 -
in 12 out of 19 firms in the pharmaceutical irdustry, 8 out of 10 companies
in chemicals and the majority of firms in electronics.
A study-/ of the Andean Pact countries experience showed that of contracts
examined 77% in Bolivia, 77% in Colombia, 75% in Ecuador and 89% in Peru con-
tained a complete export prohibition. For nationally owned firms the per-
centage which prohibited exports was 92%.- No major differences were found be-
tween sectors.
The effect on countries' export potential varies with industrial strategy
and the type of technology transfer as well as the strategy of the technology sup-
plier. While these studies found significant export restrictions, technology
transfer associated with multinational companies provides access to export mar-
TABLE 5
Provisions for lied inputs in contracl'aal agreerients
for thc transfer of itchnoIo&y
?rrcen,rage of ugircc,,ict,ts
Ctountry              containing sIIcI provisiolns
Bolivia .83
Colombia ...............                     77
Ecuador .67
Peru .62
Pilippines. .26
India:
April 1961 - March 1964 ....              IS
April 1964 -- MaLreli 1969 . .-. -          S
Solt res: Colombia, p)cpartaniento N3cion3l de Plarncaci6n, 7rans.
ferelncia de fcnolote1g(o (IBogoti. Junie 1970); C. V. Vaitsos, 7The
PrOctsS of Co"lnlzercidlianri,, eof Technzologv in the A ndran Pact
(OAS, %VWashingrorrl D.C., 1971); and Restrictive businzess practices:
Jlyteriin rcport be [lie IINCTAD secretariat (United Nationis publi-
cation, Sales No. 1.72.11.D.1 0). table 2.
NOTE. For Blolivia, EcuaJor and Peru the perccntacs rcelale to
tihc total numilber of agrecine,ts contarninin references to inpluts.
For India. the perceilttge for 1961.1964 refcrs to the total number
of effective arreenenits; U1e prccvit.-e fur 19o4.1969 refers to
those agrecicilt5 obtailjiog ;Gover nment approvjl and eventually
comiig inito force dtiring the period.
1/   Vaitsos  (1971)
2/   In  Japan,  In  1962,  53%  of  technology  contracts  contained  export  restrictions,
with lnuch  a higher proportion  in  some  industries  (90%  in  metal  products,  85%  in
non-tranrsport miaclinery, 80%  in  transport  machinery  and  65%  in  paper products):
O.E.C.D. , (1967).



- 34 -
kets in other cases, as for exampLe with the processing technologies associated
1/
with export zones.-  For this sort of technology transfer the wholly owned sub-
2/
sidiary is more likely to gain access to international market than joint ventures.-
The Table, on the previous page, based on answers to an UNCTAD questionaire
snows tne prupurr..oa us contracLb wnicn iuciuueu a cie-iL clause iur iilputs.
The study of the Andean Pact group of papers showed that tie-in clauses
were experienced in all sectors st:udied, but were greatest in the pharmaceutical
industry.
Adding up the costs: as already stated it is impossible to be accurate
because of lack of data and conceptual and theoretical problems in estimating
and attributing costs. However, it does seem clear that the overt charges for
technology transfer - royalty paymer,s etc. - are way below the total costs of
tie transfer.  For axdwple, to au6 13% of thie cubts uf importetu capital
equipment and chemicals and one fifth of declared profits on overseas investment to
the LTFCTAD estirates of rechnology costs would more than double the costs in IT6°
Fo 12 1/,r7 of non-petrolctTm exporits).  Yet in .-rew of the evidence on intra-firms
trade, transfer pricing practices, tied purchases of inputs, and *xport restrictions,
these additions are extremely modest.  For example, suppose we assume that tied
purchase of capital goods adds 30% to cost:s of capital equipment, as might be sug-
gested by the evidence on tied aid, then for countries with around three quarters
of their inputs tied by technology contracts, the additional cost would be over
twerty percent of the cost of capital equipment.
It would be interesting and useful to have accurate estimates of total costs,
but even without these, it is clear that the total cost is very significant, and
LDCs could make important foreign exchange savings if they were able to reduce
the cost of technology transfer.
1/ See eg. Helleiner (1973).
/ "One of the important differences between the two types of equity arrangement
appears to lie in the different access to export markets. The foreign con-
trolled subsidiary is more likely to be offered access to the global channels
of distribution of the -foreign firm than is a joint venture" Chudson (1974).
Some reasons and evidence is contained in Stopford and Wells (1972).



- 35 -
(ii)  The Appropriate Irice:  The market for technology is a
peculiar one - even within developed countries; peculiarities aze greater
in relation to technology transfer between developed and developing countries.
The marginal cost of supplying information that has already been developed,
are low:
"once new knowledge has been created it has the character of a public good,
in the sense that the use of such knowledge by one person does not preclude its
use by another so that optimality requires that it be made available to all poten-
tial users without charge" (Johnson)
However, since the costs of the initial research and development may be
substantial, marginal cost pricing would not provide enough incentive for its
development by the private sector. To this dilemma, two solutions are possible:
first, that the public sector should be responsible for, or subsidize, the pro-
duction of knowledge.  This raises problems in securing the necessary links with
users (problems widely encountered in developing countries, where much research
is publically financed); although experience with defence research suggests the
problems are not insuperable. The second solution is to provide leg..l and other
protection to the owners/developers of technology so that they may acquire some
monopolistic control over their technology, and consequently sell it at a price
above its marginal cost. This has been the system adopted for much technological
development in developed countries. It is often justified as necessary to secure
a continued flow of research and development. Essentially then with this system
the price charged for technology is a monopolistic/oligopolistic one. The non-
competitive pricing of technology can be viewed as a mechanism adopted by devel-
oped country firms to generate returns to innovative activity in an area where
competitive appropriability is low.
Developing countries have control over how they wish to finance or pro-
tect their own research, but in acquiring technology from abroad, they have to
accept the system as they find it. In this context it makes little sense to



- 36 -
talk of the 'right', 'optimal' or 'appropriate' price. Two extreme positions
have been put forward. On the one hand, it is sometimes suggested that the develop-
ing countries are so marginal in relation to the technology sales of advanced
countries, that they should make a minimum or zero contribution to the research and
development costs of t.±e technology they acquire.-fOn the other hand, others have
suggested that even a small inroaLd into the current price of technology, will re-
duce the flow (the flow being variously interpreted as the flow of research and
development and/or the flow of technology transfer).2/ Both common sense and
evidence suggest that the latter position is Loo extreme. Developing countries,
taken as a whole, do provide a considerable portion of the expenses of technological
development, as evidenced by the costs summarized above. But they are normally
quite a way down the line, as users of new technology, frequently acquiring it not
from the first developers hut subsequent owners,-3  after a considerable time lag.
Sales to developing countries, then, probably form only a small part of the moti-
vation behind the development of most technologies. Another question is how far the
transfer, as against the development of technology would be affected by a reduced
price.  Tne answer to tiis 3s unknown - -here can be no single answer covering
all countries anu all industries, but the experience of countries which have tried
to bargain toughly sVggests a low supply elasticity.4/
An essential feature of the market for technology then is that it is non-
competitive; consequently there is considerable (extent unknown) area for bargain-
ing, as the price may vary between an upper limit determined by costs of repro-
ducing the technology and a lower limit determined by costs of imparting the in-
formation. The important potential role of policy arises from this indeterminacy,
1J Vaitsos (1974)
2/  Carlsen and Hufbauer (1978)
_| Vaitsos (1974)
4/  See evidence of Mytelka  (1977); experience of Japan.



- 37 -
Technology Transfer and Independence
Independence of decision making is valued in itself by some - hence
the resentment of foreign ownership of a large proportion of a country's assets.
But it is also necessary for the achievement of other aims. Restrictions on
freedom of decision making tend to increase costs and reduce the firm's ability
to be selective about technological choice, and to learn and to generate an
independent technological capacity.
Independence of decision making is obviously absent in the case of wholly
owned subsidiaries of foreign companies. Nationalization of assets, joint ven-
tures and restrictions on the shares of equity that foreign compi-lies may
hold are policies that have been adopted in large part in order to preserve
independence of decision making. But close scrutiny of technology contracts of
joint ventures or national firms has revealed a very considerable degree of
restriction on independence of decision making. A typology of restrictions
was shown in Table 4.
Vaitsos' research in the Andean Pact showed - in addition to iestrictions
on exports and tie-in clauses already discussed - restrictions on prices, on volume
of sales, on quality control and on the sale of technology.-  A study-/ in Argen-
tina showed that licensing agreements with wholly locally owned firms included
clauses such as: source of inputs of machinery, materials, spare parts; prices
and quantity of output; export outlets permitted; restraints on the dymamic
effect of the transfer - e.g. non-transference of patents, restrictions on local
R and D.  Similar restrictions were found in a study of Indian firms.-/  For
example, 23 percent of the agreements examined involved some restrictions on local
sales (type, quantity, price) while over half involved restrictions on exports.4/
1/ Vaitsos (1971) p. 20.
2/ Sercovicli (1972)
3/ Subralimanian (1972)
4/  See als.o M;aovelle (1974); Oxman and Sagnsto (1972).



- 38-
Changing the ownership structure is thus insufficient to secure
independence of decision making. Indeed technology contracts have been
described as "a mechanism of control of the recipient firms." -/ Direct
purchase of technology, through, e.g., use of the experience of personnel,
purchase of machinery, copying imported products, avoids this sort of re-
striction. But for most countries, this is a possibility only for the
simpler products and techniques. As noted earlier, even emong the simpler
technologies, a tendency has been observed foi foreign owned and mixed firms
to purchase technologies via technology contracts.        A different approach -
now followed by an increasing number of countries-/ - is to outlaw restrictive
clauses. But outlawing cannot always ensure de facto elimination of the practices.
1/ Vaitsos (1971), p. 21.
2/ See UNIDO (1978): e.g. South Korea, the Andean Pact Countries, Brazil -
see below.



-- 39
Learning Effects of Technology Tra sfer
The ability to make independent technological choices, to ada?t and
improve upon chosen techniques and products, and eventually to generate new
technology endogenously are essential aspects of the process of development.
The process may be described as the accumulation of technological capacity;
it is at least as important to economic development as the accumulation of
capital.
The accumulation of technological capacity is the outcome of a complex
series of forces. One significant element is 'learning by doing'; another is
an educational, infrastructural and institutional setting which both permits
and encourages the learning process. Both aspects are crucial, although in
the past policy has tended to overemphasize the institutional side, which is
most obviously amenable to government policy, and underestimate the 'learning
by doing' side. Weak links between an elaborate institutional structure and
industrial and learning by doing activities have tended to 'marginalize' the
1/
activities of the scientific institutions.-
The accumulation of technological capacity is not a simple, easily
described activity. It cannot be measured in a straight forward way at either
macro ur micro levels. At a macro-level, residual methods of measuring technical
change, suggest the substantial significanice of increases in productivity
which are not explained by increases in inputs. But these methods are highly
dubious because of the probleins of attributing increases in output to increases
in input. Moreover, in a developing country context they exaggerate the extent
of indigenously generated technological change, since in part the technical
1/  See Hierrera (1973), Cooper (1973) and Sagasti (1978).



- 40 -
change identified is due to improved imported techniques; on the other hand,
the measure in no way captures the development of independence of decision-
making which is an important element in the accumulation of technological
capacity. Micro-case studies throw more light on the process. Again, precise
measurement is not possible, but rmicro-studies of technological change do
pinpoint the significance of such change and also suggest the myriad causes.
Maxwell's study of a steel plant :Ln Argentina, for example, showed that
(endogencusly generated) technological change, was responsible for over half
the increase in output capacity. The causes of the technological change he
attributed to exogenously determined changes in demand (leading to the need
to diversify output), exogenously determined changed in operating conditions
(e.g., deterioration in the quality of available scrap), and endogenously
generated changes arising from routine activities of engineers, etc. The
exogenous causes were of the greatest significance. The form the technical
change took varied - from major changes in process, to minor changes in procedure.
The agents if the changes included people transfer from elsewlhere in Argentina,
the employment of foreign consultants and the firms'own engineering department.
There is a complex relationship between the import of technology
and the development of local technological capacity - one that encompasses
both conflicts and complementarities. The import of technology is a necessary,
or at the very least helpful, part of the learning process in many countries
and for many industries - either by providing an essential input into the
learning process or by pennitting the country to bypass the slow process of
reinventing the wheel. On the other hand, the unrestricted inflow of foreign
technology may severely inhibit the learning process.



- 41 -
The development of local t.chnological capacity may be roughly class'fied
as a three stage process. In the first stage, the capacity for independent -E;rch
and choice is developed; in the second, minor technical changes (which may add
up to major changes in terms of quantitative impact) are generated localvy-
the third, new technology is developed indigenously. The third stage is unl'.:ely
to occur unless the f'irst two are well established. Different types of capacity
are necessary for each of the stages; in the first two entrepreneurial and
engineering capacity are required rather than formal R. and D., while formal '.
and D. may be necessary, in many cases, for the generation of new technology.
Government policy which has concentrated on formal R. and D. in government
financed research laboratories has tended to neglect the first two stages al:'.ough
they are necessary precursors to the effective use of the results of the for-.al
R. and D.
The impact of foreign teLcnology inflow is likely to be different
according to the stage reached, local scientific and technological capacity,
the form w'iich the foreign technology iniflow takes, and the industry concerned.
Unregulated technology inflow in packaged form may inhibit or delay
each stage, but part..cularly stages 1 and 3. Yet, some import of technology
is likely to be an essential input to each stage.
In relation to stage 1 - the capacity for independent technological choice -
use of foreign technology, in packaged form, tends to lidit independent choice.
In the case of wholly-owned subsidiaries, decisions about choice of techniques
remain those of the foreign company. Where technology is licensed, there is
more freedom of choice but, as is to be expected, the parent company (of mixed
firms), or the licensor determines the source of machinery to be used in mary



- 42 -
cases. -/  Limitations on decision making, which as already described form
part of many technology contracts, limit the freedom to make a choice and
consequently the possibility of learning how to do so. The effects are not
confined to the immediate decision covered by the licence but tend to carry
forward to the next set of decisions: thus, Mytelka found that firms that licensed
their technology were overwhelmingly likely to expect that they would acquire
future technology in the same form. In contrast, and not surprisingly, firms
that acquired technology by direct means -/  generally were expected to be able
to generate new technology in the siL-e way.
The second stage of accumulation of technological capacity - minor
changes and adaptations generated locally - is largely unrelated to the extent
of foreign technology inflow. It oc:curs in response to local pressures for
adaptatior and change.  Imported tec:hnology may provide the basic technology
subject to thls kind of technologicaLl change.  Local changes are generated as a re-
sult of the need to expand and diversify output, and in response to cost pressures.
Bot] Kntz and MRxb.T11 show.r s1.9nificant changes of this type occurring in subsidia-
ric!s of maultinational companies, 0.ere the basic technology was imported.  The
changes reqouired a certain initial engineering capacity, but in turn the
learning process generated through the changes created additional technological
capacity.  We have much too little evidence on change of this type to know
whether there is any systematic relationship with quantity or form of technology
transfer, any systematic relationship with competitive environment (national
l/  liytelka (19 Oh).  Of thiose licensing, 67% did ilot choose their own machinery -
these inclulded 91% of foreign firms, 100% of State firms and 44% of national
firms.
2/ Nine,-y-Seven percent of firms which licensed their technology said they would
be unable to dcvelop tLheir uwn technology for new products; 73% of firms which
nicquired technology ir; othe  .ays, said they would be able to develop their
own teclhnology for new products (defined as acquiring it other than through
3     ,:vtiI i&, .



- 43 -                                           I
ot international, although a priori. reasoning leads one to expect some),
I/
or with other variables.  Nor can we say, at this point, how far the results
are cumulative leading to acceleratir;g technical change of this type, to the
2/
next stage of technological development, or to technology exports.-
The third stage - generation of new techniques - is likely to be
inhibited by weaknesses in the earlier stages. One outcome of weakness in
decision making capacity is inability to generate technology autonomously.3/
To some extent, development of local technological capacity is an infant
industry, requiring protection from outdide competition it it is to be established.
The free import of technology inhibits its development in a number of ways: it
provides a reliable (and sometimes relatively cheap) source of technology as
compared with the risks of own development. These risks are not just scientific
ones, but include marketing risks. Techniques with a foreign trade-mark have
. m0ic certain market compared with techniques developed indigenously. Formal
restric-tions on local development of technology and/or the use of locally-
develovpc:d techniques form part of some technology contracts.A4/ Most foreign-
owned firms prefer to rely on their central research and development laboratories,
and do not conduct R. and D. in the developing countries.5/ Recently, it has been
the policy of some large U.S. companies to shift some of their R. and D. overseas, 6
but the shift has been almost exclusively to developed countries.
1/ Sercovich (1978) has suggested that the extent of such change will depend in
part on the initial design (whether 'underdesigned' or 'overdesigned') which
in turn depends on initial conditions.
2/  Katz and Ahblin  (1978) suggest a sort of product cycle in technology: the
import of technology leads to the development of local technology in respcnse
to environmental pressures, and then to the export of technology to countries
with similar environments. They illustrate with a few cases in Argentina.
3/  Again, the Mytelka study supports this showing a close connection between
freedom of choice of machinery and a belief in the firms ability to generate
its own technology in future.
4/  See UNCTAD, (1975), TD/B/AC.ll/l0/Rev. 2, paras. 83-89.
5/  The SPRU study showed that foreign firms did little or no R. and D. in
Latin America.
6/ See Ronstadt (1977).



- 44 -
Sagasti reporting on the IDRC study of science and technology policy
instruments which included research in ten less developed countries, concludes
that "Although technology imports do not always hamper the growth of domestic
science and technology capabilities, the wholesale importation of technology
without efforts to screen, control and absorb it usually stunts the growth
of domestic science and technology capabilities." (My underlining)
While a selective approach to the import of technology may be necessary
to protect and encourage local technological development, positive, promotional
policies are also required. The most successful examples of technology promotion
have combined selective imports of foreign technology with many positive promotional
measures: at a macro-level Japan's policies provide a very good example. Micro
examples include PEMEX in Mexico and fertilizer developments in India.l/
It is clear that the appropriate set of policies will vary according to
the industry as well as local technological capacity. In some technologically
complex industries, most LDCs will need to continue to rely on imported technology
for some time, and at this stage are in a position only to learn to Lhoose and to
adapt in minor ways. In less complex industries, many countries may be in a
2/
position to generate technology themselves.-
;-
Giving special promotion (plus protection) to local technological
developments may involve (as with other such promotional policies) short run
costs, in terms of reduced eff'ciency. This is in part a question of the
1/ Roberts (1973)
2/ Cooper (1976) provides a classification of industries along these lines.



- 45 -
opportunity costs of the resources levoted to technological developments,
and in part of any inferiority of the local technology as compared with the
foreign alternative. Exclusive ise of promotional policies (with no restrictions
on foreign technology) will avoid the second kind of loss, but, it seems likelv,
at the cost of making local efforts more ineffective.
There are three different types of reason why countries may consider
these costs justified: first, because technological independence is an aim
in itself. Secondly, because there are externalities of local technological
developments which are not captured by the firm, but which benefit the economy
(e.g., as a result of the learning experience, which individuals take to other
firms). Thirdly, because short run losses are believed to be outweighed by
long run gains-/- the infant industry argument.
11 It must also be assumed that government objectives differ from those
of individual firras either because of externalities or because of attitude
towards time.



- 46 -
V.  Policies Towards the Import of Technology
Relevant policies vary according to the objective, but there is
considerable overlap: some instruments serve more than one objective, while
the objectives themseJves are interrelated and mutually enforcing. For
example, policies which increase independence also tend to improve bargain-
ing strength. The effectiveness of different policies depends in part on
the general economic strategy of the country; it also depends on other poli-
cies towards technology. This is particularly true with respect to the
promotion of local technological development, where policies towards foreign
technology are only a small part of the total effort needed.-/Relevant and
appropriate policies will vary according to the stage of development of the
country, its administrative and technological capacity, and according to
the industry.
As seen above, the free and unselective import of foreign tech-
nology has involved a high cost (clirectly and indirectly), has limited
independence of decision making and, to _ome extent, the learning effects
of the technology. Policies towards the import of technology are designed
to reduce these effects.-
Many of the policies considered below are primarily aimed at
reducing the cost of imported technology:, this may be secured by changing
the form of technology inflow towards less packaged alternatives, by
1/ See the many policies described in Sagasti (1978).
2/ "Although there is recognition of the country's dependence on foreign
technology, an important objective of this legislation is to gain a
degree of control over this major import, botli in terms of cost and of
technological impact."  (UJ4IDO, 1978, p. 3).  This statement of the
Mexican Governmlent is representative of the intentions of many govern-
ments active in this area.



- 47 -
avoiding duplicated i,nports, by outlawing restrictive clauses of technology
imports, and by reducing the value attributed to the marketing aspects of
technology. To date, most concentration - at UNCTAD and in developing
countries - has been devoted to the question of the direct cost of tech-
nology transfer, although few countries have effective policies. The
balance of the discussion below on specific policies reflects this, despite
the fact that the question of accumulation of technological capacity is
probably of greater long run significance than the costs of any particular
transaction. However, many countrie- are not yet in a position to generate
their own technology to a significant extent - nor do they have the administra-
tive and technological capacity to pursue the kind of screening policies
required.  For these countries, the policies towards costs are the most
relevant. While it may be argued that the administrative costs of such
policies are high, the 'success' stories have shown that the cost savings
are also high and in most cases easily outweigh administrative costs.
However, more research is needed on the real effectiveness of regulatory
policies, in the long run, including an investigation of the extent to which
formal restrictions replace formal ones.
Some of the policies aimed at reducing the costs of technology
imports will also contribute to increasing technological capacity: for
example, less packaged alternatives tend to involve less costs, and also
more freedom of decision making. Outlawing of restrictive clauses normally
also outlaws restrictions on learning and research. Policies reducing the
significance of foreign trademarks will help protect local technology. But
effective policies towards promoting technology capacity extend well beyond
those related to cost reductions.   Selectivity towards foreign technology imports
designed to protect local developments requires an assessment of local potential



- 48 -
in each area, permitting technology imports where these are likely to induce
local innovations but preventing them where they are likely to inhibit them.
This clearly is a delicate task involving considerable judgment on the part
of policy makers. Many countries lack the administrative capacity and
technological awareness to pursue this type of policy across the board.
However, screening policies can be adopted on an industry-by-industry basis,
starting with the industries where there is most local potential. Policies
to promote local technological developments must also contain a strong
positive promotional element, pursued simultaneously with screening poli-
cies towards foreign technology. Thus, as indicated by the varying ex-
perience in this area, policies towards technology imports and promotion
are not all-or-nothing, but rather can be pursued at a different level,
according to the objectives of the country and its administrative and tech-
nological capacity.
Economic Strategy and Technology Policy
The general economic strategy -f a country strongly influences
technological developments.-  This is most obvious - and most significant -
in relation to the question of appropriate technology, where policies
towards income distribution, trading strategy, factor prices and credit
availability are critical in determining the choice of products and tech-
niques. These relationships will be discussed more below. But the general
economic strategy is also of significance in relation to the other dimensions.
1/ "Explicit science and technology instruments are those intended to
affect directly the decisions having to do with the growth of local S
and T capabilities; implicit ones are those that affect decision making
indirectly through second-order effects. The great weight, both in
number and influence, of the latter limits the potential impact of the
former." (Sagasti).



- 49
Thus, the potential for charging excessive prices for techrology is much
greater in a non-competitive en'rironment.  Hence the worst examnples of
under-invoicing tend to occur in industries which are heavily protected,
and in which, as a result there is considerable potential for making ex-
cessive profits. Lack of competition is partly a national problem, partly
an international one. Industries which are technologically sophisticated
often lack effective competition on a world basis, and in such cases
national policies alone cannot be effective Ji eliminating the source of
non-competitiveness and, therefore, the excessive prices. An important
source of non-competitiveness occurs via trademarks, advertising and product
differentiation. In economies, in which this type of market structure is
predominant, excessive costs of technology imports are likely. Technology
policies may be designed to offset these effects (e.g. by regulating the
prices paid, investigating invoicing and so on), but so long as the funda-
mental causes remain, the policies may well prove to be only nominally
effective.
The general economic and technological environment is also of
significance in determining the rate of accumulation of technological
capacity. But simple generalizations about the relationship between trade
and pricing policies and indigenous technlcal change are not possible at
this stage. It is necessary first to distinguish between protection/
competition in the market for goods and that in the market for technology.
As already suggested, some protection in the market for technology may be
necessary to protect local technology, and this will extend to those goods



- 50 -
(notably capital goods) in which technology is embodied.  But this is the
reverse of the form which protection normally takes, where capital goods
and technology are subject to very little or no protection, while consumer
goods are subject to hcavy protection.-L It has been claimed that "in general,
import substitution policy and full-scale protection of consumer goods
industry have tended to promote a passive attitude to the utilization and
development of indigenous R and D efforts during the early phase of indus-
trial development ."-This claim seems to be borne out by case studies of
Brazil and Ghana.-/ Yet micro-studies of technical change suggest that
trading policies may not be of major significance. Examples of quite
substantial indigenous technical change have been found in both open and
closed economies.-/The success of countries such as Argentina and India -
both heavily protected economies - in developing technology exports-/
challenge the view that an open trading strategy and undistorted factor
prices are necessary for the generation of technical change.
Technological developments are uiost likely where there already
exists a certain amount of technological capacity, in the form of engineer-
ing resources and an innovative attitude on the part of entrepreneurs,
1/ For example, in Indonesia the amalgamated tariff trade and sales tax on
capital goods and basic industrial inputs is 5%, that on intermediate
inputs 25% and on manufactured consumer goods,. 50-70%.  (See World Bank,
Indonesia, Basic Economic Report, 1979).
2/ Nam Kee Lee (1975)
3/ Beranek and Ranis (1978)
4/ See the work of Katz and others in the IDB/ECLA Research Programme in
Science and Technology.
5/ See Katz and Ablin (1978) and Lall (1978).



-- 51 
managers and worl<ers. Given this sort of background, local innovation is
more likely the greater the pressures from the local environment which
make local technological change necessary or worthwhile. Such pressures
include 'natural' pressures occuring as a result of differences in physical
environment, market size, resource availability and cost as compared with
the environment for which the foreign technology was designed and 'artificial'
pressures created, for example, by restrictions on technology imports.
The general proposition that local technology change responds
to such pressures in the environment is well documented in the empirical
literature.-/ This would suggest that a competitive environment (both
nationally and internationally) and factor prices broadly in line with
factor availability would be more conducive to technical change than an
environment in which firms are highly protected and are able to survive
and prosper without innovation or adaptation.-/However, internal competition,
difference in market size and the non-availability of certain parts and
materials may be at least as effective a; international competition in
creating the required pressures. This explains why technical innovation
and adaption does seem to occur in 'protected' economies as well as in
open economies. More systematic comparative data is necessary, at both
macro- and micro-levels,before further and more confident generalizations
is possible.
1/ Many studies have suggested that adaption and innovation tend to occur
in response to necessity (e.g. the absence of particular materials, the
disappearance of markets, major differences in size of market) to a
greater extent than in response to less dramatic incentives offered by
cost differences.
2/ See Ranis in Beranek and Ranis (1978).



- 52 -
Policies Towards the Terms of Transfer:
Policies towards foreign investment: direct overseas investment
provides the greatest potential for hidden costs of technology transfer, and
generally also involves the higlhest overt costs and the greatest loss of
independence in decision making. Policies which permit only a minority of
equity ownership on ntw investments, and a divestment on old are designed to
reduce these effects.
Many countries regard foreign direct investment as the only means of
securing the technology; but some have introduced restrictions on onwership
in order to meet their technology objectives. In some,majority foreign owner-
ship is only permitted in limited 'sophisticated' industries; in others it is
prohibited altogether.  In general, the aim seems to have been to regulate
permitted equity ownership without adversely affecting the inflow of foreign
technology.
U.S. DIREC-1 INVISTMENTS IN ANDEAN MANUFACTURING
in MSillions of U.S. S
7uu.
675
650
625
600                              Venezueb
575
550
525
,00
4175
450              7
425
400
375
350
325
200                              Cotomb.a
275                          >
7S0                   _4     -4-4
2250
200          A.4,
175         JA4
150 j
12S.4
I10X) . X
Z; 0*       -    *rw
50  *  ____--____               Cbib
19t4 r,,s19Z0Po- rZ IS977 M93
preeD D.cson 24  po&tI-ecision 24
So.mic: U.S.S4utll,if At a  , 4*f gh/e, Uited.4   Situ.   1972. 1973. 1974. 1975
z!XIit>l., D.C.:  Goveiiiiiicnt  lliintiitg  Office).



- 53 -
The Andean Pact, decision 24, provided for divestnent as well as minority
foreign ownership in new investments.  According to the study by Mytelka the
divestment requirements had only had very limited effects by 1975.  The
inflow of U.S. direct investment does not appear to have been markedly affected
by the decision, as shown in the chart.
Transfer pricing and foreign investment: the terms of technology
transfer may be improved by enforcing arms length pricing on the inter-
national transactions of transnational companies for the purpose of customs
and taxation. In practice, this is difficult to achieve because arms length
prices are unknown and because of the administrative burden involved.
Kopits argues that "with the exception of the United States, most countries
have neither the willingness nor the administrative machinery to enforce
arms length prices....'"-/
Taxation policy: the tax system may be used to deal with the transfer
pricing problem more directly, by introducing taxes which will be th.e same
irrespective of the prices of imported inputs or exported output. T.io
approaches are possible: one is to impute the profits a company is making
in a particular country by assuming them to be the same ratio to some known
variable (e.g., turnover or. employment) as world wide profits are to world
wide turnover or employment. To be effective, this obviously requires that
world wide accounts are reliably known, which is not always the case. A
second approach is to introduce a uniform tax on all items, so that irrespec-
1/ Kopits (1976a); the arms length solution has proved difficult to
administer in the U.S.  See O'Conner and Russo (1973) and Musgrave (1969).



- 54 -
tive of price manipulation, the same tax will be paid. A uniform tax rate
levied on profits plus imported Lnputs plus royalty payments plus management
fees less exports would achieve this.-    It would remove the tax incentive
(but not other incentives) for transfer price manipulation, wniie eiiiuillatitig
the tax cost of such manipulation, and consequently improving the tetms of
technology transfer in the case of direct foreign investment. Some countries
have prohibited subsidiary cQmpanies from deducting royalities in payment
to parent companies in calculating tax liability, which is a first step
in this direction.2/
Raising tax rates and/or removinR tax incentives (hnlidays, in:Jvest-
ment grants etc.) would also improve the terms of technology transfer. In
some cases the terms a country pays would be improved without any deteri-
oration in the terms received by the company. This is the case where the
LDC effective tax rate is below that of the DC from which the investment
comes, and the DC uses the credit method of double tax relief, and dcoes not
permit tax sparing agreements.-/ In such cases any tax loss as a result of
low tax rates/tax incentives offered by LDCs simply increases DC tax revenue.4/
Since many LDCs offer extremely generous tax incentives,-/ and since the
J/ See I.L.O. (1972) for discussion of a similar proposal. The proposed
tax base is similar to the Eur-opean value added tax, except that it
excludes payments for local labor and materials.
2/ An obvious objection to the uniformity proposal is that it wnL0I  rermn    P, enitntry'
freedom to vary its rate of tariff on different items, since the tariff
on imports would be incorporatied into the uniform tax.
3/ These arguments are elaborated in Stewart (1978).
4/ Apart from gains the company makes by deferring tax payments; these may
be substantial and long lived if tax havens are involved.
5/ See e.g., Lent (1967), Reuber (1973).



- 55 -
U.S. uses the credit method of tax relief and does not allow tax sparing
relief, this is the situation for a large proportion of investment in LDCs.
In such a situation, the LDCs should bring their tax rates into line with
those ruling in the DC country from which the bulk of their investment
comes. If they wish to give special incentives,'they should do so in the
form of subsidies (e.g., to employment) which would have the added advantage
of encouraging the sort of investment they wish for.
There have been numerous studies testifying to the extreme ineffective-
ness of tax incentives in LDCs in encouraging investment.-/ "The most
striking conclusion of the survey was that the foreign investors, almost
without exception, stated that taxation concessions played no role at all
in bringing them to Singapore". 2/ Studies in Argentina, Mexico, Jamaica,
Guatemala and Nigeria come to similar conclusions. Not a single study
suggests that tax incentives have been effective. They persist, nonethe-
less, increasing the cost of technology transfer to the countries concerned.
One reason is that each country believes it is on a kinked demand curve,
such that if it raised its tax rate and removed incentives it would lose
a large share of investment to neighboring countries. To get round this
belief (which -.ay well be a'misconception) regional agreement is required
in removing tax incentives and raising tax rates. Alternatively, countries
could act individually but maintain incentives by giving subsidies instead
of tax allowances.
/ E.g., King (1970), Hinricks (1974), Chen-Young (1967), Phillips (1968),
Hughes and Seng (1969).
2/ Hughes (1969)



-56 -
Taxation may also be used (as discussed more below) to encourage
research and development and labor intensive techniques of production.
International Codesof Conduct: recent international initiatives
have been to negotiate International Codes of Conduct both with respect to
multinational corporations and to the transfer of technology.-! These
codes are intended to set standards for behavior in the case of MNCs in
relation to disclosure, taxation, employment behavior etc.; in the case of
the Code of Conduct for Technology Transfer in relation to payments for
technology, restrictive practices and so on. A major issue is whether the
Codes should be mandatory or voluntary; if voluntary their impact is li'kely
to be minimal, although they may perform an educational role and encourage
national action where major departures from the Code are noted. Some have
argued that mandatory Codes might impede progress in LDCs "since a general
and widely accepted Code would have such a low common denominator that it
would actually retard certain changes iL the third world".-/  It has been
suggested that these Codes should form part of a wider international
investment regime; the aim of such a regime would be to provide an inter-
national framework in which foreign investment would be regulated. The
regime, it is argued, would reduce uncertainty (and might include features
such as guarantees, insurance etc. that would ensure this), would provide
11 The U.N. resolution on the establishment of NIEO (197&) called for a code
of conduct in relation to TNCs.  A draft International Code of Conduct on
the Transfer of Technology has been prepared in UNCTAD and is due to be
discussed shortly. The U.N. center for Transnational Corporations is in
the process of drawing up a Code for TNCs.
_! 7a!tsos (197!).



- 57 -
for the fairer distributioa cf benefits and costs, and would thus encourage
the regulated flow of technology and finance, benefitting all three parties
TNCS, developed countries and developing countries.
Policies towards technology contracts: the policies just discussed
are concerned with technology transfer in tae form of direct investment.
But as ownership restrictions become effective, technology transfer
increasingly becomes an arms length transaction between national (or joint)
companies and foreign suppliers. It then becomes important to regulate
this so as to prevent this form of technology transfer having the undesirable
effects described. To this end, a number of countries have begun to intervene
actively in the inflow of technology by establishing national registries of
technology, reviewing and renegotiating technology contracts, and outlawing
various restrictive practices. These policies may be designed to have the
joint effects of increasing the bargaining power of national companies,
reducing costs of technology transfer, protecting local technological
developmeits, increasing the learning effects of imported technology and
reducing the loss of independence in decision making. As suggested above,
some of the objectiF.es are mutually reinforcing.
There are National Registries of Technology in the Andean Pact
countries (established in 1971), Argentina (1971) Mexico (1973), S.Korea
(1976) and India (1977). In other countries, other parts of the government
machinery may effectively act as a national registry (e.g., Portugal).
National registries require the registration of all technology contracts
with foreign suppliers;this registration may only be permitted after



- 58 -
government approval, which may only be given if the contract fulfills
certain conditions and is on reasonable terms. The Andean Pact Decision
provided for the exchange of infornation between registries thus strengthening
the potential of each national registry.
It is worth investigating one country's experience in some detail
in order to illustrate how registration may contribute to objectives. In
Mexico the legislation prohibits technology agreements as follows:
1. when the purpose is to transfer technology already freely
available in the country (objectives: cost and protection of
of local technology);
2. when the price is unjustified (cost);
3.  when provisions are included which permit the supplier to
regulate or intervene in the administration of the transferee
(independence);
4. when there is an obligation to assign to the supplier of the
technology the patents, trade marks, innovations or improvements
received by the transferee (learning and independence);
5. when limitations are imposed on technological research or develop-
ment (learning and independence);
6.  when there is an obligation to acquire inputs from any given
source (costs and independence);
7. when exports are prohibited against the best interests of the
country (costs and independence);
8.  when the use of complementary technologies is prohibited
(costs, independence and learning);



- 9 -
9.  when chere is an obligation to sell exclusively to the supplier
of the technology (costs and independence);
10.  when the transferee is required to use  permanently,personnel
designated by the supplier (independence, learning and costs);
11.  when the volume of production is limited, or sale or resale
prices imposed (independence and costs);
12.  when the transferee is required to appoint the supplier as exclusive
sales agent in Mexico (independence and costs);
13.  when unreasonable terms of duration are established (costs
and learning);
14.  when the parties shall submit to foreign courts for decisions
in interpreting contracts (national independence).
Since the law came into effect in 1973, the Registry has examined about 4,600
agreements. 35% were rejected for excessive payments or restrictive clauses;
60% were rejected if one excludes majority foreign owned enterprises. The
table indi ates the most frequent reasons for negative resolutions.
As % of Rejected
Agreements
1.  Excessive or unjustified payments                  68.5
2.  Excessive duration of agreements                   31.6
3. Prohibition to use non-patented technology
or to manufacture goods, at the end of the
agreement                                          30.7
4. Submission of agreements to foreign laws
or courts                                          18.5
5. Grant-back clauses on innovations produced
by licensee                                        16.8
6.  Export restrictions contrary to the
interest of the country                            14.5



60 -
It is claimed that the policy has substantially reduced foreign
exchange costs of technology, operned export markets, increased tax revenue,
reduced prices of capital goods arLd produced "growing consciousness among
Mexican entrepreneurs, who have understood better the importance for
productivity and profits, of obtaining full information on technological
alternatives, conducting a careful evaluation and selection of products,
processes, and licensors, and negotiati.ng terms of contracts from the
strongest possible position."'l
The Mexican experience appears to confirm the pioneering efforts
in Coiombia where a Comite de Regalias was established in 1967. Between
1967 and 1971 the Comite de Regalias of Colombia evaluated 395 contracts
of which 334 were negotiated, modified and eventually approved and 61 were
rejected. In the process of negotiation,payments of royalties were
reduced by about 40%. Between th- latter part of 1970 and the beginning
of 1971 the Comite reduced by 90% tie-in clauses, eliminated export restrictive
clauses, eltminated by 80% minimum royalty payments, prohibited tax deduc-
tion for royalties and established maximum percentage royalty rates.2/
For most part these policies have been directed at reducing the
cost of technology transfer. Experience of these and other countries suggests
considerable foreign exchange gains can be made by national policies to
regulate the terms of technology transfer. It should be noted, however,
that the gains be in large part only nominal. Registries and other regula-
tions determine the formal contracts, but informal arrangements may
/ UNIDO (1978) p. 53.
2/ Vaitsos (1971)



- 61 -
negate their effects.-/ Effo-rts to regulate the transfer could be strengthened
in their content (e.g. some courtrias prohibit the duplicated transfer of
the same or very similar technology) and through an international dimension.
With international exchange of information on technology bargains, each
nation's capacity to regulate the terms of the inflow would be strengthened,
as would its knowledge about alternative sources of technology. Such inter-
national cooperation could best be started on a regional basis. The Andean
Pact contains provisions for such cooperation. International cooperation
would be particularly useful for countries with a weak administrative
structure, which can do little on their own.
The effectiveness of such regulatory measures depend largely on two
factors: first, on how fast technology suppliers replace their formal restric-
tive activities by identical informal activities; secondly, on how toughly the
local administrators pursue the policies. In the context of an economy which
relies heavily on the import of te hnology, local administrators very often
become embroiled in the interests of the technology suppliers and are then un-
able to trEat them with the arms length objectivity that is required.  The piece-
meal dismantling (or weakening) of some of the efforts to regulate the inflow of
technology (of which we see signs with respect to the Andean Pact and Mexico, for
example) suggest that early successes may exaggerate the full impact of regula-
tion. We need more information on these areas.
Promotion of Technological Accumulation
The main purpose of most of the policies just described - and the
way in which they tend to be interpreted by those countries which carry
l/ According to Sagasti, Registries only control the formal contract: "A
potential licensee who needs technology is in fact at the mercy of the
licensor who has the technology.  No amount of government intervention
can restrict the prOssurcs imposed by a licenser, and most licensees
submit voluntarily to the demandls, although the formal contract may not
reflect any controls." (p. 19).



- 62 -
them out - is to control and improve the terms on which foreign. technology
is transferred, rather than to prornote local technological developments.
Some of the policies would also contribute to this objective as shown, for
example, by the details of the M"exiLcan provisions. But more comprehensive
policies are necessary if the promotion of technological developments
locally is the main objective: indeed for countries which have the capa-
city and intention to pursue this als a major objective the cost issue - on
which so much attention is currently focussed - becomes of subsidiary importa2.ze.
There may be, as already indicated, many routes to the achieve-
ment of a sustained process of indigenous technical innovation. Contrast,
for example, the different paths pursued historically by the UK, the US,
Germany and Japan.-/ Moreover, there is considerable uncertainty as to
either necessary or sufficient conditions in developing countries today.
But the following seem to be the types of policies likely to be conducive
to technological change:
a. Policies which protect local technological develop-
ments by permitting only selective import of foreign
technology;
b.  Policies which remove legal restrictions on local
developments, and reduce the excessive market value
attributed to foreign as against local technology:
i.e. policies towards patents and trademarks. trade-
mark policies may also contribute to the cost reduc-
tion objective.
c. Policies which promote local technological develop-
ments by providing the required infrastructure,
education, institutions, incentives and so on.
/  For an excellent overview see Ranis in Beranek and Ranis (1978);
see also the individual essays in that volume and Landes (1972).



- 63 -
(a) Selectivity towards technology impor.ts. Most successful
'second comers' technologically have relied on the selective import of
technology from abroad. According to Rosenberg the U.S. was "highly
discriminating in borrowing patterns and highly selective in the uses
to which imported technologies  were put."" In 19th Century U.S. this
was achieved without deliberate government intervention, as a result of
relatively high levels of local mechanical and innovative abilities,
reacting with the very substantial differences in resource availability
compared with technology supplving countries, notably the U.K. Japan's
policy of selectivity was a much more conscious and deliberate part of
government policy. Between 1868 and the 1950s, Japan regulated and
restricted technology imports ruthlessly, while promoting local technological
capacity by education and learning by doing. The aim was to import
technology from abroad only once, and then learn to develop and adapt the
technology herself. The import of technology was "effected by diverse
means ranging from sending missions overseas to seek and purchase machinery
to more sophisticated technical licensing agreements with foreign
technology suppliers" ..."the importation of machinery ... immediately
became the object of imitation and trial manufacture, and eventually
of domestic production."2/
The net effect was a steady rise-in the level of technology
imports and a change in their composition to ever more sophisticated
industries.
1/ "The Role of Science and Technology in American National Development,"
in Beranek and Ranis (1978.)
2/  UNCTAD (1978a), para. 17.



- 64 -
In some ways, developing countries today face a more difficult
situation and have mbre need for protection of local developments than
the early latecomers. The scientific and technological gap between
their own capacities and that of the advanced countries is greater, and
the commercialization of technology and the domination of the technology
market by a small number of transnational corporations has gone much
further. Successful protection and promotion requires a dual policy of
selectivity towards imports and promotion of local developments.
Selectivity involves some form of control over technology
imports which may be achieved in a variety of ways: the centralized
purchase of all technology imports provides one-model (involving obvious
administrative problems - it is probably a better model for socialist
than mixed economies); an alternative is for some form of compulsory
registration (as with the National Registers described above), where
permission for technology imports is given only if the technology import
can be shown to be complementary io (rather than competitive with) local
technological efforts. An efficiency criterion may be added to prevent
excessive protection leading to very inefficient technology.
Mytelka's research into the Andean Pact country experience
suggests that the development of local capacity to choose and generate
technology may require unpackaged non-contractual technology transfer.
-This is difficult to induce; however, promotion of local consultancy
firms, specializing in packaging, would reduce dependency on foreign
sources    for packaging.



- C5 -
For the most part, azcording to Sagasti, screening of foreign
technology has been weak because the agencies involved lack firm and
clear criteria for selection.-/ South Korea has recently introduced
selectivity permitcing the input: of technology only if it fulfills
various criteria; technologies which duplicate existing technologies,
or which would easily be developed locally are not to be imported.
However, it seems in practice that a very liberal interpretation is
adopted and most requests for technology imports are granted.  While
the import of technology is in theory, subject to regulation, the S.
Korean Technology Transfer Centre conducts an active survey to identify
foreign technologies that are needed, and has specified 1958 technolo-
gies needed over the next five years necessary for the production of
1,500 items in the machinery, electronics, metal and chemical
industries.-/
(b) Two types of legal restriction provide an especial
handicap to local technological developments - patents and trademarks:
Patents: thLe international patent system was intended to protect
and encourage innovations.  But the system does not appear to work in the
best interests of developing countries. Of the world total, 6% of patents
were granted by developing countries, and of these 5/6 are held by
1/ Sagasti (1978)
2/  UNIDO (1978a), pp., 57-64



- 66 -
foreigners, mostly TNCs, and only one sixth by developing country nationals.
The commercial value of those held by nationals is relatively small.
Among patents issued to foreigners by LDCs, the vast majority (over 90%)
are not exploited. Patents thus, for the most part, tend to prevent
competition and local innovation, rather than encourage itT Currently moves
are under way to secure reform of the international patent system: the
proposed reforms include   the introduction of investors certificates,
exclusion of some products/processed from patentability, limitations on
patent duration, stricter provisions for compulsory licensing and revocation
as remedy for non-use, removal of restrictions on imports of articles and
products           manufactured by a process which is patented in the
country, and strengthening disclosure requirements.
International reform along these lines is preferable to individual
national initiatives, but if it fails, or proves insufficient each country
(or group of countries)    has the power to enforce their own system, as
have a number of countries including Japan, Italy and the Eastern block
countries.
1/  See Vaitsos (1972), and UNCTAD (1975b):  "patent practices of
developing countries, following international standards,
have legalized this peculiar situation wlhich has come to act as a
system of reverse preferences granted to foreign patent holders in
the markets of developing countries".



- 67 -
Trademavrks:    tradcrmarks act as a c3r.ificate of origin and
a guarantee of quality. Eut the costs to LECs are high. While the private
benefits of acquirLng a trademark to a particular entrepreneur, in terms
of market power, may be great, there are social costs.  The system encourages
the import of foreign technology from companies responsible for well known
trademarks, even where there are alternative sources of technology,
domestically or from less well known and less costly sources overseas and thus
reinforces foreign technological domination. It tends to favor the
established and well known, making it more difficult for newcomers to
compete. It leads to expenditure on advertising, some of which is infor-
mative, much persuasive, which developing countries can ill-afford.
According to one estimate a group of developing countries spen., on average,
0.77% of GNP on advertising,-L/ (which is more than expenditure on Rand D.).
Moreover, products originating in developed countries tend to be most
heavily advertised which means that inappropriate products tend to be
overpromoted in relation to traditional and appropriate products.- In
general, the question of trademarks is most critical for consumer goods,
and of lesser significance for capital and intermediate goods.
In developing countries 50% of new trademarks issues in 1974 were
registered by foreigners (88% in Africa, 65% in Asia and 34% in Latin America),
a sizeable increase over 1964 (when 27% of newq trademarks were registered
by foreigners). The vast majority of trademarks registered by foreigners
are from developed-market economies (95%); 1% are owned by nationals of other
developing countries. Chemicals, food, tobacco and textiles account for
60% of the new registrations.-J
I/  !l('rA! (1077b) 7, 5F
2/ The situation is notoriously bad in the drug industry where brand name/
advertised drugs are widely prescribed, although the costs are sometimes
50 tilnes or more those of generic equivalents.  See Lall (1975).
3/ UNCTAD (19771h).



- 68 -
Trademarks are said to be a powerful tool for the promotion of
exports. However, developing countries have not, for the most part, used
this tool to promote their own trademarks:  insofar as trademarks  have
been used, they have been mainly of foreign origin.
Policies Towards Trademarks:   some of the abuses associated with
trademarks would be eliminated by policies which eliminated restrictive
practices in technology contracts. But other action would be required
to reduce this source of monopolistic power, to protect the development
of local technologies, and to reduce other undesired effects of the
trademark  system.  The following represents a broad range of possibilities.
i. Total abolition of trademarks in all sectors, or in selected
sectors. The major problem here is the need to provide alter-
native systems of protection for consumers: these are difficult
to devise and costly to administer. One possibilit-y would be
to provide govennment 'utility' certificates to products ful-
filling required standards.
For political and administrative reasons, this does not seem to
be a likely policy for most countries in relation to all sectors. However,
selected sectors a're especially   suitable for such policies.  In particular,
it is a possibility for the drug industry and has been recommended by the
Hathi Coimnittee in India, xwhich proposed that India's basic drug needs
could be met by 116 generic drugs, as against 15,000 drugs currently
marketed.11 Since there is sectoral concentration of trade mark registra-
tion, action of this sort in one or two sectors might do much to prevent
1/  Policies to replace brarndecl drugs with generic drugs were introduced
in Sri Lanka in 1972.  The State Pharmaceutical Corporation purchased 47
drugs at an average cost of 28% of the price quoted by the former trans-
national suppliers. But after a major campaign, the policies were partially
reversed in 1977.  In Afghanistan a recent statute imposed the use of generic
names for pharmaceuticals.



- 6)
abuses.
For most sectors, however, we may assume that the system is
likely to remain, and therefore policies need to be devised to reduce
abuses:
ii. Limitations on permitted advertising.
iii. Consumer education.
iv. Alternative sources of quality control, such as government
approval certificates, and legislative standards.
v.  Compulsory licensing of trademarks such that all who wish may
produce with the trade-iark on payment of some specified fee.
vi. Raising the fees for registration of trademarks, particularly
in relation to renewals.-/
These policies are applicable to all trademarks, foreign or national,
Other policies may be devised to give special protecticka to local
trade marks:
vii. By a differential system of charging fees.
viii.  By banning foreign trademarks, totally or selectively.
ix.  By the development of joint trademarks.  In both Mexico and Brazil
policies to encourage joint trademarks have been introduced.
In NIexico joint trademlarks are nmandatory, in Brazil optional,
the option to be exercised by the licensee. Joint trademarks
permit the sharing of the benefits from good will created
by the imarks, and assist in the development of the local
trademark.  But they are unlikely to have major effects on
consumption patterns or local technological developments.
1/  Fees chargcd for regist:ration In developing countries are on average
half those charged in developed countries. In developed countries
renewal fees are as high or higher than initial fees; in developing
countries iriany -countrics charge a lowqer fee and some charge no fee
at all.



- 70 -
The question of trademarks is one of the most important issues in
relation to technology transfer in the area of consumers goods and medicines;
it is closely related to the question of costs, since consumer preference for well
known trademarks permits technology suppliers to earn quasi-rents; it is related
to the question of the developmenit of local technological capacity, as esta-
blished trademarks provide an obst:acle to the development and use of local tech-
nologies; and it is also relevant to the question of appropriate products and
techniques. Yet is is a relatively new area, in terms of policy. Few
policies have been tried and partly as a consequence many have not been
fully developed. Policies here are closely tied up with many other
areas of policy - towards consumption patterns, and international trade
as well as technology.  In this area (as in others) it is fairly easy to
see the direction in which countries should go if they are to share the
rents earned,but more difficult to see how to protect local dejelopments.
International action to develop agreed LDC product standards and systems
of quality control, independently of t-ademarks,is one possible way forward.
(c)  Policies to promote local technological innovation.-/   Policies
to promote local technological developments are clearly an essential counter-
part of policies towards technology imports. The range of policies adopted
to promote technological development are well known. On the supply side,
1/ The main focus of this paper is on the international transfer of
technology. Consequently, the sections devoted to local technological
developments are shorter than a balanced treatment would require. More
discussion is contained in the section on Appropriate Technology below.



- 71 -
they include building up scientific and techrical infrastructure, providing
general and technizal education and training, promoting local consultancy
firms and developing Technology Plans.  On the demand side, they include
government procurement policies, tax and incentive policies and industrial
policies. In addition to these 'explicit' science and technology policies
there are also - as discussed above - general economic policies which
influence the rate of local innovation.  Unfortunately, while the relevant
promotional policies are well known, there is considerable controversy
surrounding almost every one of t Cem. Sagasti's survey of science and
technology policy instruments concludes that "evidence gathered in the
Science and Technology Policy Instruments project shows overwhelmingly that
the explicit S and T instruments (with the exception-of personnel training)
have little impact on technological change particulazly at the early stages
of industrialization." The major impetus in most LDCs has been towards
building up institutions for basic restarch, seeing these are the first step
in the innovative chain: but whether this is an essential part of technical
innovation, or a sensible way of developing countries to spend scarce resources
is now subject to donsiderable controversy: "the less developed world is strewn
with scientific institutes and other expensive white elephants which contribute
neither to science nor to technology.' '1 Technology Plans are becoming
increasingly fashionable, but like general economic plans, their effectiveness
depends on the policies and projects they produce rather than the Plan document
itself, which may often prove a waste of scarce administrative resources.
1/  Ranis in Beranek and Ranis (1978).  Ilis paper sammarizes the arguments for
and against basic scienice in LDCs.



- 72 -
Many tax incentives have proved redundant:-/ the most successful are the
most specific, as in Peru and S. Korea. But as with most tax incentives,
there is a strong possibility that they will simply induce firms to categorize
expenditure as 'R. and D.' in order to qualify for the incentives, without
making any real changes. Moreover, such incentives tend to leave out the
more mundane minor innovative activities, which may have at least as much
to contribute to the innovative process, and may be an essential precursor
of major innovation.
Government procurement policies represent one of the most powerful
tools available for promoting the use of local technology.
Four conclusions appear to stand out from experience with respect to
promoting local technological change: first, that links with industry are
.essential, if promotional efforts are not simply to add to the zoo of
white elephants. Such links may be achieved through contract research,
in-house research or strong informal links between the scientific and
technological activity.-/ Secondly, that there has been some tendency to
overdo basic science as against applied science and technology. Thirdly,
that there is a criitical mass of scientific and technological activity
necessary before the work has any substantial effects. Some of the apparent
failures of many LDCs may be because this critical mass has not yet been
reached. It has been suggested that certain countries - notably Argentina,
India, Brazil, S. Korea and Mexico - have now achieved the minimal critical
1/ Sagasti (1978)
2/ Japan provided sucil links by developing 'model' government owned factories
and then selling them to the pirivate sector, when they were successful.
In S. Korea, KIST hlas manifold informal links withi industry.



- 73 -
mass for indigenous technological developmer.t -  zourthlly, countries
should specialize in scLence and technology, consciously choosing the
areas which they intend to develop, and concentrate promotional activities
and selectivity towards foreign technology imports, in these areas.2/
Administration
The discussion of policies has briefly reviewed a large number of
actions LDC governments might take, if they adopt an active technology
policy. Some of them would impose heavy a6ministrative costs
since they involve investigating techrology contracts one by one: policies
towards costs and those involving screening foreign technology imports
would be especially costly from an administrative point of view; they would
also (particularly the screening policy) require considerable technical
expertise, which is especially scarce in most LDCs.   Thus the administrative
costs involved are by no means trivial; it is therefore necessary to
establish that such resources are available, and that their use in this
respect would be worthwhile. Here one needs to distinguish between policies
directed towards improving the terms of the transfer and those towards screen-
ing imports to promote local developments.
Many countries are not yet at a stage in which very much screening
makes sense. For this there must be a certain amount of local technological
capacity. The development of such capacity requires the sort of promotional
1/ Sagasti (1978)
2/  Cooper (1976), Sagasti (1978) and Ranis in Beranek and Ranis (1978)
all come to this conclusion.



- 74 -
policies just discussed, and is likely to take place initially on the basis
of imported technology. Hence, countries which are particularly weak
administratively (especially in relation to the special skills required
of technology administrators) are also those where screening policies are
premature. They may be desirable in one or two industries but screening
for these cases would not require excessive administrative resources. But
these countries, which rely heavily on imported technology and lack much
capacity themselves, are precisely those where policies towards controlling
the terms of transfer are likely to generate the greatest gains. Moreover,
the established magnitude of the Iains are such that even quite heavy
administrative resources (which may themselves be bought from abroad,
for example from more advanced LDCs) would be justified. But where the
administrative base is weak there is much to be said- for introducing
general policies - such as taxation and policies towards protection and
trademarks - which reduce the foreign exchange loss through technology
purchase and therefore the need for aduinistrative policies to control it.
As already suggested, regional and international cooperation may ease
the administrative burden.
Countries which are more developed, in technological terms, are those
for which the screening policies are more relevant, but they also have more
techno-administrative resources to support them. Moreover, as they gain in
technological strength, the importance of policies towards the terms of
transfer diminishes because the individual enterprises become increasingly
capable of striking a satisfactory bargain for themselves and hence the



- 75 -
terms tend to improve without sc riuclh goverTLmert inte-vention1  An
effective screening policy does require considerable expettise and
judgement, but given that the policy m.ay be initially concentrated on
a few industries, the costs need not be excessive.
Aspects of Political Economy
Many of the policies described involve challenging very powerful
interests: some of these interests are outside the developing countries,
but in many cases they have been internalizad and are well represented
within the country concerned. The history of attempts to control the
terms of technology transfer is strewn with failures of 'will,' which is
a somewhat amorphous way of describing successful resistance, on the part
of the interests involved, to attempts to control them. The following few
examples provide indicationis of what can happen.
In both Sri Lanka and Brazil, plans to replace brandname drugs
with generic drugs were diluted after powerful political campaigns which
had strong foreign support.2/
In Mexico, the near monopoly position of a successful national firm
in the steroid hormone industry, which led in research and development,
was broken as a result of strong protestations by U.S. Companies, and
1/  Sagasti concludes that countries which are more technol6gically
sophisticated "suffer less from the restrictions that normally
accompanly the irmports of technology and pay less for knowledge
than do firms in countries where these capabilities have not been
developed " (p. 49). But the experience of the U.K. both in relation
to oil and natural gas technology, and to pharmaceuticals (La Roclhe)
show that eternal vigilance is required.
2/ See Aganral (1978)



- 76 -
pressure from the U.S. Government. As a result shortly after the industry
was dominated by foreign subsidiaries, which secured the plant barbasco,
on terms which meajnt that much of the rent from this rare plant was taken
out of Mexico. Later, government attempts to regain control over the
industry for Mexican firms failed partly due to political opposition.L/
Mytelka (1977) illustrates aspects of political economy with respect
to three Andean Pact countries, following their common policy of Decision 24.
Peru, with a stronger national bourgeousie, succeeded in divesting in
accordance with this policy between 1970 and 1975, while changes in Ecuador
and Colombia, where foreign interests were relatively stronger, were
negligible.
Certain conclusions are suggested by taking the political economy
dimension into account: first, from a political point of view the ability
to control the terms of technology transfer is likely to be greater the more
arms lergth (and the less) previous relationships with technology suppliers.
Thus, paradoxically, from this point of view, less developed countries with
little technology transfer may be in a stronger position than those where
interests are firmly entrenched. Secondly, certain types of policy may be
easier to secure, from a political. point of view, than others. General
promotional policies are likely to be more acceptable than specific restrictive
policies. This is not just because they are likely to be less effective
(although doubtless this is part of the story) but also because their effects
are normally more widely spread and therefore specific opposition is less
likely to be aroused. Thirdly, discretionary policies are more subject to
1/ Cereffi (1978)



- 7/ -
abuse (and to thwartirg by e.g., brit-ery) than general policies.  This
suggests that general tax changes may be more effective than particular
contract-by-contract negotiaticns.  But, although they may be more ef'ective
if properly conceived and introduced, for this vcry reason, they are likely
to be subject to very great opposition.
The political economy aspects do not provide a reason for not
trying to introduce desirable changes: they suggest that political forces
should be taken into account in formul'ating policies and the need for strong
'will' to carry them out, needs to be emphasized.



- 78 -
VI. The Characteristics of Technology
The characteristics of technology are largely determined by the
nature of the economies for which they are designed. The most significant
determinants of the characteristics of new technology are the income levels,
resource availability and costs in the society in and for which the technology
is designed, and the system of organization of production, and the nature of
the technology in use in the society. In each of these respects, societies
of advanced countries differ from those of poor countries. Consequently,
technology designed to suit advanced countries tends to be ill adapted (or
'inappropriate') to the conditions prevalent in poor countries. The transfer
of such technology to poor societies tends, as a result, to cause various
distortions and inefficiencies.
In discussing 'characteristics' of technology, one should include
all :-he relevant features which determine its resource use, productivity,
and impact on production and consumption patterns. These features include
the nature and design of the product, the scale and organizational system
for which the technology is designed, its resource use, including capital
and labor intensity, materials and fuel use, skill requirements, and the
infrastructural and complementary inputs it requires. The traditional
economists' characterization of techniques according to their capital or
labor intensity forms only one, and quite often a relatively ipsignificant,
aspect of the total characterization.
The income levels of a society broadly determine consumption
patterns, both as to the type of products consumed and as to the nature of
products within each broad category of need. Countries with the same



- 79
average income  evel may differ in c,nsamption patterns beicause of diffe-
rences in income distribution and cultural anid institutional differences.
Nonetheless, average income levels are an important determinant of con-
sumption patterns. The design of products irn advanced countr-es, then, is
broadly in line with their income levels. Wlen products are transferred to
poorer societies, their characteristics tend to be out of line with average
income levels. This results in imbalances in consumption patterns: the
imbalance partly takes the form of maldistribution of income and consumption,
so that a small proportion of the p^'pulation earn incomes similar to those
of rich societies, and have correspondingly similar consumption patterns.
Another form of imbalance is that which occurs in expenditure patterns at low
income levels: a high proportion of expenditure may be devoted to goods
designed for higher income levels, while other needs are relatively neg-
lected.l/
The view that products are desi.gned for particular income levels,
and ma. be inappropriate for societies with much lower incomes, does not
mean such products confer no net benefits on these societies and should be
rejected. In some cases, this may be the situation. But in others, the
products may extend choice and increase welfare. Given the massive research
and eevelopment devoted to new products-/ in advanced socirties, the near
l/ A rLcent study by J. Wells (1977) quotes a survey in Sao Paulo, Brazil,
which found that between 1959 and 1969/70 'there was a significant rise
in the share of household expenditure devoted to domestic appliances...
an a.ialysis of (physical) per capita food consumption shows that there
was :, douvnward trend, indicating that levels of nutrition among the work-
Ing class deteriorated absolutely.,
2/ Al. Javestigations into motives and outcome of R. and D. in advanced
countrics show that much the largest proportion is devoted to new products
or imTr'vements in old products.  See Gustafson (1962) and Mansfield (1969).



- 80 -
complete neglect-of research into products specifically appropriate to poor
societies, and the economies of scale associated with the mass production of
many of these new products, many modern products increase the efficiency of
need fulfillment, even for low income consumers, as compared with the pre-
vious range of choice. There are cases where the modern products may
decrease welfare. But the main argument about appropriate products is not
that new products from advanced countries decrease welfare absolutelv,-/ hut
that they tend to have high-income characteristics, and result in consump-
tion imbalances compared with new products specifically designed for the
conditions and income levels of poor societies.
In general, it is to be expected that new products would increase
welfare by increasing the range of choice open to consumers. But if the
new products completely displace (or raise the costs substantially) of old
products with more appropriate characteristics, or if consumer choice is
distorted by advertising exclusively devoted to one class of products, con-
sumers in poor societies may actually be nade worse off than before the new
products were introduced.-  This is particularly likely to be the situation
l/ Because of their high income characteristics, new products do tend to
benefit higher income groups disproportionately thus leading to a rela-
tive deterioration in income distribution, but not one that is normally
captured by the statistics.
2/ In my study of maize milling, I found that the most recently developed
product was more expensive, less nutritious and required a more capital
intensive and large scale technology. It was the only branded product
and the only one advertised._ Consumers were switching to it in a
substantial way.  This was not only due to advertising,  It lasted longer
and many consumers preferred the taste. Similar examples are to be
found in the drug industry where new products, which are much more expensive than
the old generic equivalent,are heavily advertised, and also have some additional
benefits (e.g. in terms of reduced side effects). See Speiglit (1975).



- 81 -
where there are major ineivisibilities, so that decisions are non-marginal.
Very often this applies to decisions involving government expenditure such
as major irrigation projects, hospitals, road and transport systems, military
systems and so on.
The higher level of incomes in advanced countries also influences
resource availability and costs. The labor force tends to be highly
educated; real incomes and the 6pportunity cost of employing labor is high.
Thus techniques designed for these conditions tend to involve high levels
of worker education and skills, and high levels of productivity. Moreover,
the high incomes generate a corresponding high level of savings per head,
and therefore of investible resources.  In comparison with average per capita
resource availability in LDCs, the technology, therefore, requires savings
per workplace far in excess of those available and makes education require-
ments greater than those the country can afford. If the technology is
transferred unmodified to poorer countries, investment resources and ex-
penditure on the development of human resources have to be disproportionately
devoted to a small proportion of the population.
The organization system of advanced countries - with large units
of production managed by a professional managerial class - influences the
type of technology that: the system produces. Inevitably, new techniques
that fit in with this structure tend to be developed, rather than techniques
designed for a radically different system. But the system is very far
removed from traditional societies in most LDCs; to adopt the technology
efficiently, the advanced country productive system has to be reproduced.



82 -
This involves dependence on imported managers, heavy expenditure on managerial
education, and leads to an apparent dearth of managerial and entrepreneurial
talent, and, often, a scale of production that is disproportionate to
market size leading to oligopoly/monopoly and excess capacity.
Another important determinant of the characteristics of new tech-
niques is the technology already in use.. This imposes demands on any new
technique, in terms of scale and characteristics of output, if the output is to
serve as inputs-elsewhere in the system, and determines the availability and
nature of inputs. The technology in use in most advanced countries demands
standardization and reliability. The net result of the many connections
between different parts of the productive system is that there is (within
broad limits) technological determinism; techniques which do not fit in
become inefficient, not because thvy are absolutely inefficient, but because
they are inefficient within the context of the technology in use. As far
as LDCs are concerned, this means that their choice is much more circumscribed
than might appear.  Having made an initial decision to adopt some part of
modern technology, this often dictates that other parts should also be
adopted. Hence, while a careful study of available choice of techniques
suggests that the choice is reasonably extensive, the real choice in the con-
text of a particular technology and market system may be much narrower.
To summarize briefly: techniques designed for modern advanced
countries tend to produce high-income products, require high levels of
investible resources per employee, high levels of education and skills, be
of a large scale and require sophisticated management techniques, be asso-
ciated with high levels of labor productivity, and be linked, through inputs



- 83 -
and outputs, with the rest of the alvanced technelogy syatem. If these
techniques are transferred unmodified to LDCs, the result will be a con-
centration of resources, of savings and expenditure on human resources and
infrastructure, on a small part of the economy. Iitcomes will tend to be
concentrated in this area, leading to markecs for the high-income products
the system produces. Resources available in the lou-income country will
tend to be underutilized, including raw materials as well as labor.
Many of the well-established characteristics of the dual economy
can be seen as following from the characteristics of advanced country tech-
nology: the capital-intensity of productive techniques, the heavy reliance
on imported managers, skill deficiencies, un and underemployment and a rela-
tive (often absolute) deprivation of the economy outside the modern sector.
Only economies which are growing very rapidly and are selective about the
choice of techniques and adept at modifying them are able to overcome this
dualism, by absorbing a growing proportion of their workforce into the
modern sector. South Korea and Taiwan provide the obvious examples. In
other economies, dualistic tendencies have been partially offset by a
deliberate attempt to protect the non-modern sector, providing it with
resources and protected markets to prevent it being undermined by the
modern sector. This is the policy of 'walking on two legs' pursued most
extensively by China (and to a less marked extent by India). In China
employment expansior. in the modern sector has been similar to that in many
other developing economies, lagging well behind growth in output, while the
technology adopted has tended to be capital intensive.-/Overall employment
}/  bee Sigurdson (1977) and Rawski (1978).



- 84 -
policies have succeeded because of the absorption of labor in the agricultural
and rural non-agricultural sectors.
There is a growing body of literature that questions the rather
simplistic technological determinist argument advan~ced above. It is argued
that in many industries a wide choice of efficient technologies has been
established by empirical research.-/Recently, Pack has argued that countries
could make significant gains at the macro-level in terms of employment, out-
put and savings, by policies leading to the adoption of the most appropriate
techniques in existence. It has long been established that there is con-
siderable potential for labor intensity in ancilliary processes, even if the
2/
core technology is more fixed.-' Old techniques from advanced countries
offer more labor intensive and small-scale alternatives than the most recent
techniques. While this sort of argument supports the view that there is
a wider range of choice than the completely determinist view implies, other
czonsiderations suggest that the potential for selecting labor intensive tech-
niques may be exaggerated by calculations based on the micro-case stixies.
In the first place, many of the *studies show that considerations of product
standards/characteristics may rule out the labor-intensive technologies.-/
Secondly, the labor intensive techriques are often only economic at small
scale.-/ Thirdly, entrepreneurs dlo not have information about the complete
1/ See e.R. Bhalla (1975), Jenkins (1975). Pack (l978a)r White (l9-7R).
2/ See Ranis (1973).
3/ See Stewart (1977), p. 199.
4/  Ibid.  See Boon C1964, 1975). Pack (1978a),



85 ^
'shelf' of technicues in existence; their access to inforimation aoout diffe-
rent techniques denends on theiz channels of information. There tends to
be a bias in chiannels of information towards technology currently in use in
the supplying countries - i.e. the advanced countries. Appropriate tech-
niques, which are often older techniques fro1n advanced countries, or tech-
niques recently developed in LDCs are less well promoted.
Fourthly, many of the studies neglect the determinants of choice
or selection mechanisms. It is often assumed that the only relevant selec-
tion mechanism is the relative price of capital and labor, and t.±at is in
the control of the government.  In fact, the determinants of choice are far
more complex.  The nature and scale of the market is one critical deter-
minant. Products sold on the international market or to high-income consumers
may need to use the most recent technology in order to compete.-L/ Because
scale of production and the nature of the market are of importance in deter-
mining choice of product and technique, the factors determining these are
significant. Income distribution and trading strategy help determine the
nature and scale of markets for different types of product. The distribu-
tion of investible resources between enterprises of different size and type
2/
is also of relevance.- The very substantial wage differentials between
enterprises of different sizes suggests that the real wage level may be
outcome as well as (partial) cause of technological choice. Moreover, a
1/ Cooper et. al (1975) show that the more labor-intensive techniques may
have to be ruled out for 'certain quality demanding segments of the market'.
Sercovich shows in case studies in Argentina that firms were frequently
forced to abandon attempts to use local technology by the need to compete
in product differentiation to maintain their share of the market. High
quality products are not invariably produced with capital intensive technology.
In Egypt, El-Karanshawy found that it was the capital-intensive methods of
producing carpets that found it difficult to find an international market.
The processing zones use modern but labor intensive techniques to produce
exports.
2/ See evidence on how choice of technique varies with size of enterprise in
World Bank (1978).



number of recent studies of technological choice have shown that the most
rational choices - both in terms of profit maximization and appropriate-
ness of techniques are sometimes rejected in favor of less profitable and
less appropriate techniques.-L/ Thus investigation of the decision mechanisms
involved is required if choice of technique is to be altered in a more appro-
priate direction.
Fifthly, the studies establishing a range of choice are essen-
tially static, depicting the situation at the point in time when they were
made. But the so-called 'shelf' is moving as technical change p.-oceeds,
and some parts of the shelf are moving much faster than others. In parti-
cular, with the current world balance of R. and D., the modern capital
intensive part of the shelf is moving forward, in te.ms of new products and
efficiency of technique, as compared with the labor intensive part; and the
'modern' end is also getting increasingly capital-intensive, larger scale,
and the products more sophisticated as incomes rise in the advanced coun-
tries. Thus while there may be scope for some push in the direction of
labor intensity and other dimensions of appropriateness in some products,
the aggregate effects are liKely to be swamped by dynamic changes in the
opposite direction.
Policies to Secure More Appropriate Technology
More appropriate technology may be roughly defined as technology
whose resource use is more in keeping with LDC resource availability, and
whose products are more suited to low-income consumers. This means that
more appropriate technology will be more labor intensive, less skill intensive
1/ See studies of Tlmmer et. al (1975), and Pickett et. al (1974),



smaller scale, use more local materials, and produce sim'ler low-income
products than most of advanced counitry technology currently being transferred
to low-income countries.-/ Technology m.y be more appropriate ir. one respect,
and no more so in others; or more in some respects, less in others.  It is
then a matter of judgment as to whether it is to be preferred. More appro-
priate technology may be less efficient, either in the economic sense that
it is less profitable, (socially and/privately), or in the stricter sense
(which has been defined as technical inefficiency) that it uses more of all
resources to produce the same output. If it is socially or technically
2/
inefficient as compared with the 'inappropriate' alternative, then the
economy would suffer from some output loss in adopting it. In such a situa-
tion it would only be worthwhile adopting if it were believed that learning
effects would be such as to outweigh initial inefficiency, or that income
distribution and similar effects are such as to outweigh the loss in output.
If we assume that these effects may be incorporated into the measure of
social efficiency (and that we can give some meaning to this concept), then
an economy should only adopt more appropriate techniques if they can be shown
to be socially efficient. In discussing ways in which appropriate technology
1/ Some of the characteristics of 'appropriate technologyt briefly described
above (for a fuller discussion see Stewart (1977), Chapter.4) have been
subject to considerable debate; probably most controversial are tvo - the
question of scale and that of 'appropriate' products. The Strathclyde
research believes that scale is irrelevant to appropriateness. Above it
is defined as one criterion among others, on the grounds that small-scale
productive units are more in keeping with the entrepreneurial and managerial
abilities and market size of most LDCs. The appropriateness of products
is also left out of the Strathclyde discussion - and of most of the case
studies examining choice of technique. It is included here for reasons
discussed above.
2/ Both these concepts present problems - which may be extremely serious in
deciding whether a particular technology is efficient in one or other
senses.  But it fs a useful. shorthand here, to indicate that efficiency
is an important censideral:ion, but one that needs to be qualified to
include income distributioni and lecrning effects.



- 88 -
may be promoted, it is assumed that what is in question is the promotion of
socially efficient appropriate technology.
Broadly, one may distinguish two types of appropriate technology:
appropriate technology for the 'modern' sector, which consists in the adap-
tion of modern sector advanced country technologies in more labor intensive
directions; and appropriate technology for the traditional sector, which
upgrades and improves traditional technologies.  Both have been neglected
in terms of information collection and dissemination, research and develop-
ment.  Promotion of more appropriate. technology in both categories may
involve improvements in information dissemination, minor or major modifica-
tion of existing techniques, or the development of entirely new techniques
and products. For completeness perhaps a third category should be included:
more appropriate systems of public services, including more appropriate
products and delivery systems.
It is difficult to establish orders of priority for the various
categories, and perhaps unnecessar.y to do so. The potential for improve-
ments in the traditional-technology area are great, given the almost com-
plete neglect in the past; this area is likely to help those in greatest
poverty more directly than modifications in modern sector tecnnology.
Successful innovations in this area should improve the productivity of the
poorest and, if widely disseminated, could reach very large numbers of
people. Appropriate innovations in the modern sector would speed up growth
of employment in this sector but are unlikely to improve the lives of most
of the very poor-, in the short run: in essence they are likely to expand



- 89 -
the numbers in aiddle-income occupaticns..  Innovations ,.n this araa have to
compete with simultaneous R. and D. in developed countries, which -ay render
the results obsolete - particulaily where there is Droduct innQoetion and
competition in trade - and may also make it difficult to ensure that the
appropriate techniques are widely adopted. On the other hand, if a process
of innovation were engendered In this sector, it could become self-sustained,
leading to learning by doing and further innovation.-/ In so far as LDCs'
ultimate future lies in the modern sector, then innovation here may be
critical.  Since the bulk of investment is likely to remain in the modern
sector in middle-income countries, innovatiDns related to this sector would
be likely to affect a larger proportion of investment. Appropriate innova-
tion in the modern sector is also vital if developirg countries are to
avoid having to follow future technological developments in the advanced
councries, which are likely to be in an increasingly inappropriate direction.
Policies to promote appropriate techrology may be roughly categorized
into those that affect the demand for different types of technology ai.d those-
that affect the supply.
A different set of policies, with respect both to demand and
supply, is relevant to promoting appropriate technology according to the
sector being considered: this is because the three sectors - traditional,
modern industrial, and public sector - contain different types of decision
maker, with different objectives, consLrsirits and capacities. Among obvious
differences, the naLule ot iuaz'ets serv(. d differs between traditional and
1/ As appears to 1-ir1e happened, for t   ;.le, in the case of the Swaraj
tractor - see. Morehouse (1978).



- 90 -
modern sectors and consequently the type of product in demand; the size of
investible resources and managerial capacities differ so that different
scales of production occur; factor prices and access to technology differs,
with the traditional sector facing low labor costs and high capital costs,
while often in the modern sector capital is subsidized and labor costs
(relative at least to the traditional sector) are high. In the modern sector,
there are normally strong links with foreign sources of technology. The
links are weak or non-existent in the traditional sector. All these factors
affect the selection of technology.
In the modern sector in many countries, as many studies suggest
as well as much casual evidence, changes are needed in selection mechanisms
as well as in information/development related to appropriate technology. In
the traditional sector, information, education and credit is requiired. The
sectors also differ in their potential capacity to generate the information
search and technology development themselves. For the most part, the tradi-
tional sector lacks the capacity for self-generation and new technologies
must be developed from outside, but with strong links with the potential
users.-/ The need for,outside institutional support applies to information
collection as well as technology development. But the modern industrial
sector in many countries is potentially capable of doing a considerable por-
tion of the search and research itself, and would'benefit in terms of take-
up of the results and learning effects by so doing. In this sector, then
what is needed is an incentive structure and support system for internally
11 As for example in the work of ASTRA in Bangalore.



- 91 -
generated activities.  Efforts of independent instituticns financed from
outside often fail because they are not nsed by local entrepreneurs.
Most of the policies affecting demand are internal to particular
countries. They include changes in income distribution, which affects the
monetary demand for different types of product; changes in trading strategy,
which similarly affects demand for different types of product; changes in
credit/tax policy affecting the control over investible resources of different
types of decision maker; and changes in factor prices. Government policy
towards consumer demand - e.g. educption of consumers, limitations on advertis-
ing, policies towards trademarks - also affect choice of products and tech-
nique; and in the very large number of decisions over which government has
1/
direct control, the govcrnment may directly affect the choice of technology.-
Mast 6f these changes, as stated, are internal: only trading strategy has
a direct international dimension. Here there are a number of options:
a. The classic import substitution strategy. This normally
amounts to import reproduction rather than import substitu-
tion, and since it involves complete replication of goods
previotsly imported, it also involves replication of methods
of production. The consequence is capital-intensity of pro-
duction and inappropriate p*roducts.
1/ Little systematic research has been done comparing state and private
sector decision making with respect to choice of technique. The limited
evidence available suggests that state corporations in mixed economies
choose at least as inappropriate technology as the private sector (see
El-Karanshawy for a study of carpet weaving in Egypt where the State-owned
firm was introducing mechanized methods to compete with traditional hand-
woven methods in the private sector). It also appears that the State
sector is more lilkely to depend on foreign sources of technonlnv than
private nationally-owned firms - see MIytelka (1978) and Morehouse (1978).



- 92 -
b. Export orientation, particularly towards the North. This,
at least in the short run, should involve a more labor
intensive technology (although not the production of more
appropriate productis) than the straight I-S policy. As
time proceeds, s0 long as innovation remains the virtual
monopoly of the North, the policy is likely to. require
continual adjustmenit and increasing capital-intensity,
but of a lesser extent than a policy of progressive I-S,
which keeps up with the latest product developments in
the North.
c.  Appropriate import substitution - i.e. the substitution
of locally produced 'appropriate' goods for those pre-
viously imported. To some extent, this was the policy of
Japan in the thirties, and of China in the 1960s and 1970s.
The technology and products should be more appropriate,
but there may be losses in inefficiency because of the
absence of modern research into appropriate productsi/and
because of restrictions on trade. It is really only feasible
for large countries.
1/ Japan was able to avoid this source of inefficiency by taking 'modernt
products and modifying them. See UNCTAD (1978a).



- 93 -
d.  South-South trading links :ombined with a polic- to promote
appropriate products and techliiques and to restrict the import
of inappropriate products and techniques. This is likely to
involve reorientation of trading strategy in manufactures away
from North-South to South-South. The option is similar to
appropriate import substitution but it i3 on a South basis, rather
than that of a single country. -/ It thus enables the South to
exploit the advantages of international trade, while maintaining
South protection, and increasing the appropriateness of technology.
The South-South trading relationship would include trade in technology
and capital goods. There are various degrees of stringency with
which this option may be pursued. At one extreme, it requires com-
plete delinking with the North, apart from primary production. At
the other, it involves greater preference on a South-South basis
but without further restrictions on North-South trading links.
Current moves to promote South-South links are in line śiith the
latter strategy.
1/ Paul Streeten drew my attention to the pertinent quotation from D. H. Robertson:
(Robertson was discussing-trade between the U. S. and the U.K.).
"The simple fellow who, to the advantage of both, has been earning a living
by cooking the dinner for a busy and prosperous scientist, wakes up one day
to find that his master has invested in a completely automatic cooker, and that
if he wants to remain a member of the household he must turn shoe-black. He
acquires a kit and learns the techniques, only to find that his master has
invented a dust-repelling shoe, but would nevertheless be graciously willing
for him to remain on an empty the trash-bins . Would he not do better to remove
himself from the orbit of the great man and cultivate his own back garden? And
if he can find some other simple fellows in the same boat with whom to gang up
and practice the division of labor on a less bewildering basis, so much the better
for him."



- 94 -
The supply of different technologies:   a false, but nonetheless
helpful distinction is often made between information/research and development.
The distinction is false because information collection merges into search,
and search into research. I/ There are weaknesses, absolutely and relative
to what is happening at the inappropriate end of the technology spectrum, in
each of these activities in relation to appropriate technology.
Information
Commercial channels provide the mair. sources of information to
entrepreneurs in the Third World: for the most part, these are developed
country sources - TNCs, consultants, machinery salesmen - all of whom are
far more familiar with, and are more likely to benefit from investment decisions
incorporating fairly recent techniques designed in and for advanced countries.-
rhere are a growing number of voluntary bodies who try to provide information
services in relation to appropriate technology, but their efforts are small,
somewhat amateurish, and they are not incorporated into the main stream of
1/  For example, the Strathclyde study in collecting information about the
existing range of technical change devoted a minimum of two people for one
year to the process. As they devised 'synthetic' alternatives to those
techniques already in existence (i.e., new combinations of existing processes),
their efforts came closer to research than search. To establish the viability
of some of the new synthetic combinations would require testing, etc., essentially
a development activity.
2/ See Roberts (1972) and Bhatt (197)).



- 95 -
entrepreneurial decision maii,, In tLhe Third 'Vcrld  i;  I'he development of
Third World consuitancy firms, sotie 1hird World TICs and machi.nery suppli rs
2/
are beginning zo o-ovide more appropriate sources of tnformaticn.-   But the
total level of these efforts are small, and while it seems that the information
provided does tend to be related to more appropriate technologies than most
of the advanced country resources, this is not always the case. -/  A few con-
sultancy firms in the advanced countries are beginning to specialize in this
field.
It is clear that there is a need for a more systematic system of
information collection and dissemination in relation to appropriate technology.
A number of proposals for such a system have been put forward. 4/  Some are of
the data bank variety. Others consist of an information cum consultancy service. It
seems the latter are more likely to be effective, as data banks are difficult to
use, and do not provide the essential software advice which is needed. The
most promising proposal for international action was that developed in the World
Bank: the technology referral service. -/ Promotion of Third World sources
of technology is another important aspect of policies to promote information
sources of more appropriate technology. Although Third World capital goods
1/ Some of these are described in Stewart (1979)
2/ See Lall (1978), L.T. Wells (1977) (1978),(1979) and Lecraw (1977). Lecraw
does a detailed comparison of Indian and other TNCs and finds that their
technology is more appropriate (i.e., more labor intensive, on a smaller
scale and produces simpler products).. Wells (1979) finds LDC firms' invest-
ment in Indonesia are considerably less capital-intensive than developed'
country multinational investment; they are also concentrated in less research-
intensive industries and in industries witlh less product differentiation.
3/ It is widely claimed that the capital goods industries in Brazil, India
and also South Korea ishow very little adaptation as compared with capital
goods industries in advanced countries. See Pack (1978b) and Stewart
(1977), Chapter 6.
4/  Sm marized in Stewart (1979).
5/ This proposal, in abeyance at the moment, linked the provision of information
to financial screening institutions in developing countries, thus providing the
important tie-up with decision making, which is lacking in most proposals.



- 96 -
industries have been castegised for failing to be 'appropriate' and to innovate,
on balance they are probably more appropriate than most advanced country
technologies; and they may become more so in the future, as innovation is
encouraged.
The promotion of Third World sources of technology is largely a matter
of internal policy of developing countries and of their trade and technology
regulations vis-a-vis other Third World countriep. Developed countries may
assiRt hv siippnrt for cpnitR1 garIAs industries and Third LTorld consultancy in
their aid policies, and by relaxing aid tying to include machinery bought from
Third World sources.



- 97 -
The Development of AppropriFtez Technology
Scme appropriace. technologies have been developed in the developed
countries, while much current R. and D. in LDCs is tnot directed at the
development of appropriate technology. Hence, one could conclude that R. and D.
in LDCs is neither necessary nor sufficient for the development of appropriate
technology. Nonetheless, there are very strong arguments favoring developing
appropriate technology Ln LDCs; first, because the learning effects would then
be generated within the developing countries; secondly, because links with LDC
entrepreneurs are easier to maintain; thirdly, because resource costs and
availability of developing countries are more likely to be incorporated in the
design of the new technologies. For example, although there was no explicit inten-
tion to develop appropriate technology, the indigenous development of an Indian
tractor involved far lower capital costs than advanced country technology. The
aim, therefore, should be to promote a process of research, development and
technical change in developing countries which will produce more appropriate
technologies. This aim is in line with t:he objective of creating technological
capacity in LDCs. While this does not exclude all efforts in developed countries,
it indicates where the main thrust should be. The best way of generating this
process of technological change will vary according to sector and to country.
Among LDCs there are very marked differences in current technological capacity.
Some countries - particularly among the human resource rich middle-income countries
(but also including India) - have many of the inputs required to generate such
change.  Others lack institutional capacity, and human resources. -  Given these
marked differences in capacity, there is a dilemma: one possibility would be to con-
centrate efforts in those areas where the resources are already present, encouraging
X/  For a discussion of obstacles to technical change for a   rr±ate So_noogy
in LDCs, see Reddy (1979) and IMAT (1978).



- 98 -
the export of appropriate technology from them to other less well endowed
countries. Unavoidably, this will form one aspect of a strategy to promote
appropriate technology; but in order to aioid new forms of dependency emerging
within the Third World, and to generate the technological capacity essential
for long-run development, effort:s to promote appropriate technological change
should also be promoted in the resource-deficient areas.
Clearly, relevant policies will differ in the two cases, as may
relevant technologies. In some ways the distinction is similar to that made
earlier between developing technology for the modern sector and that for the
traditional sector. In the middle-income human resource rich countries, for
the most part the technologies may be developed using internal resources.
International assistance may be required in a catalytic way to induce current
institutions to redirect their efforts to appropriate technology./   Much
of the earlier discussion on the development of technological capacity is
relevant here (see pp. 61-73). It was suggested there that a three pronged
po' cy was required, consisting of selectivity towards technology imports,
changing the legal and marketing system so as to favor local technology,
and promotion of local developmenits through provision of infrastructure govern-
ment subsidies, incentives and so on. Especial attention needs to be placed on
securing strong links with the productive sector preferably by getting the
research and development carried out within firms.  Specific tax and incentive
policies may be effective. For (eample, in Peru, a tax of 2% is levied on
net pre-tax income of industrial firms which is used to finance central research.
The tax is repaid to firms who use it to finance R. and D.; the net proceeds go
1/ See IXAT (1978) for a definition and description of the type of catalytic
effort required.



.1 99 -
to finance a cenwral f'und f'.r R. and D.-I A recen. suvrey!  found that 60%
of the Peruvian fl'rms sampled engaged in R. and D. compared wj.th 11% of firms
izlncuador and 23% of Colombian firms, which may indicate the effectiveness
of the incentive.
The results of such R. and D. may not be appropriate according to
all, or even sometimes any, of the criteria discussed above. But as argued
above, it Is likely to result in more appropriate technology than much imported
unmodified technology../ This is because local costs and resource availability
unavoidably influence technology design, even when there is no conscious
intention to do achieve this. But much depends on the structure of incentives
and markets within the economy. If the market is protected, market share
depends on differentiated modern advanced country products, and factor prices
are 'distorted,' then locally-generated technologies may not-be substantially
different-from imported technologies.  This is'especially likely if the import
-of foreign technology is not restricted.  -But with a more appropriate set of
-selection mechanisms, more appropriate results are likely.
International assistance may be needed to-help provide resources of
both finance and manpower to create institutions to promote appropriate
4/
,technology in the traditional sector.-
1/ Sagasti (1975)
2/ Mytelka (1978)
3/  In the case of t'he local development of technology to make a tractor in India,
the capital costs were estimated to be about half that of the imported
technology, for the same level of output.  An indigenously designed casting
process involved Rs. 1,000 in contrast to Rs. 500,000 in imported equipment.
(Morehouse, 1978).
4/ See suggestions in IMAT (1978).



- 100 -
In countries lacking irLstitutional or human capacity for technical
change, much more outside assistance will be required, initially, to generate
local appropriate technologies. In these countries, the immediate prospects
may not be very good, but the efforts are worthwhile to generate the local
learning process.
Many proposals have been put forward for international assistance for the
development of appropriate technology. -/ The most developed proposal, at this
stage, is the proposed International Mechanism for Appropriate Technology. The
feasibility study on this, defines IMAT's activities as essentially catalytic,
raising funds and providing assistance to national efforts in LD.w's.
Efforts to develop appropriate technology need to be accompanied
by improvements in the dissemination of information so that the results may
be widely commuicated.
Development and communication of appropriate technology hardware forms
only a small part of the changes needed: as already suggested, changes in
slIection mechanisms are essential, so that newly developed approprit:e technology
is used. Moreover, there is what is known as the software element in the use
of technology, i.e., all the managerial and associated services which may be
ill-adapted to the new technology, and which must be adapted or it will fail.
Social/cultural and political obstacles to the use of appropriate technology are
sometimes argued to be of greater importance than the availability of hardware.
1/ Eg. Austin Robinson's proposal to U.N.D.P. (1976)



- 101 -
VIi.  Some Conclusions
1.     In recent years, there has c'eveloped wide3pread recognition of
the need to regulate the inflow of foreign technology into developing
countries from advanced countries.  This recognition has arisen from the
growing evidence that unrestricted transfer has involved high costs,
has restricted the independence of decision makers, has limited the
learning effects of technology transfer, and has inhibited the development
of local technological capacity. An active policy regulating the inflow
of technology is necessary to mitigate these effects. Recognition of the
need to regulate the inflow has been largely independent of the question
of appropriate technology and is shared by many countries which place
little emphasis on appropriate technology.
2.     Despite some overlap, policies designed to improve the terms
of technology transfer should be distinguished from those designed to
protect and promote 'Local technological developments. Most emphasls,
to date, has been placed on policies to regulate the terms.
3.     An essential feature of the technology market is that it is
imperfect - natural and legal imperfections enable the 'owners' of
technological knowledge to appropriate returns to this knowledge. These
imperfections - whiclh are normally justified as necessary to generate
adequate returns to research and development - mean that the price levied
tends to be oligopolistic and subject to bargaining.
4.     Policies designed to improve the terms of technology transfer
may be mandatory, or rely on incentives and persuasion. There are
broadly two areas of policy: one, is aimed at decreasing the packaging



- 102 -
of technology transfer so as to increase local bargaining power, to
reduce dependence and learn how to put the package together locally,
and to make control over terms easier; the other is improving the terms
of transfer of each form of technology transfer. As policies towards
depackaging become effective, regulation of the terms of unpackaged
transfer becomes important. The tax system may be designed to improve
the terms of transfer, in both packaged and unpackaged form. National
registration of technology contracts, together with requirements regulating
the terms and outlawing restrictive practices before registration is
permitted,and prohibiting duplicative imports,improve the terms of
unpackaged transfers. Some countries have made very significanc foreign
exchange gains by such policies w,hich would appear easily to justify
the administrative costs involved. However, it is not yet clear that
the de jure regulations (e.g. outlawing various restrictions) are always
carried out de facto. Regulations do not cover the many informal transfers
mechanisms, which may be of greater total significance than formal mechanisms.
5.     Policies which regulate the terms of technology inflow will 'e
insufficient to prumote local technological development. Countries
whici have local technological potential need an active policy to promote
local technological change.  Such a policy includes positive promotional
measures - such as the provision of infrastructure, subsidies and tax
incentives, education and government procurement policies - and a selective
policy towards t'e import of foreign technology. Such a selective policy
involves the discriminating import of foreign technology in those areas
where local potential exists, by encouraging the import of technology which
is likely to generate local learning and innovation and is complementary



- 103 -
with local efforts, and restricting it where it is competitive with local
capacity and is likely to inhibit it. The combination of positive
incentives and selectivity towards foreign tachnology was the successful
strategy of Japan; similar policies are now being initiated in S. Korea,
although regulations towards technology imports appear to be liberally
interpreted. This combination of policies has been adopted successfully for
particular technological developments elsewhere. But many LDCs which have
the potential (i.e., human resource rich countries) for technological capacity
are underusing it, and continuing to rely primarily on foreign technology,
which is allowed free entry, so long as the costs are not excessive. There is
a sharp contrast in the protection given the local production of goods by
many countries with the lack of protection to local technology. Selective poli-
cies of this kind need considerable administrative and technical expertise; they
bsould be introduced on an industry-by-industry basis, where strong potential
local technological capacity exists.
6. The system of patents and trademarks tends to give protection *o foreign
technology suppliers. International reform of the patent system to avoid this
is currently-in process of negotiation.   But        the question of patents is
probably of less significance than that of trademarks. Trademarks tend to
reinforce the technological dominance of foreign suppliers arid contribute to
inappropriate consumption patterns. LDC trademarks are relatively undeveloped,



- 104 -
although they formed half new registrations in 1974. To promote local technology
it is necessary to reduce the dominant position of foreign trademarks.  A vari-
ety of policies are possible ranging from total restrictions to financial in-
centives. It is difficult to devise a system which preserves the benefits -
as quality guarantor - of trademarks, while reducing the costs they impose.
7. Although many of tht policies towards technology inflow must, in the final
analysis, be pursued at the national level, there is considerable potential for
international cooperative action. International cooperation in the exchange
of information may strengthen each country's knowledge of alternatives and
practices elsewhere. The international (or regional) harmonisation of practices
with respect to transfer pricing, taxation and regulation of technology
contracts will prevent any single country losing a disproportionate share of
foreign technology inflow, which might occur if they act alone. International
action may work to identify alternatives to well known sources of technology.
Thi- may be particularly valuable for LDCs at an early stage of development,
which lack the capacity to search for alternatives and package the technology
for themselves. The ability to regulate the inflow of technology depends
on the level of development of the country concerned. Countries which have un-
dertaken active policies have been mainly middle income, with considerable ad-
ministrative and managerial capacity and local entrepreneurship.



- 1.5 -.
8.  Local technological capacity in third world countries is o- critical impor-
tance:  first i. is a vital part of the development process.    Secondly, it is
necessary for independence ane to  iAprove bargaining power in relation to the
import of tecbnology. Thirdly, it is necessary to generate appropriate technical
change. While there appears to be a certain amount of technological choice
today, the continued concentration of technical change on advanced country
technology is likely to result in increasingly inappropriate techniques. Un-
less developing countries undertake R. and D. in alternative directions, the
choice of technology available in the future will be increasingly circumscribed,
and irrelevant to the needs of the world's poorest. The development of a con-
tinuous process of technological change - new techniques and products - in
an appropriate direction in the developing countries is essential if the choice
is to be widened.
9. Technology recently developed in advanced countries tends to be inappropriate,
in many respects, for many LDCs. This is because it is designed to meet the
needs of the advanced countries; it tends for example to be increasingly capital
intensive. Its use in poor countries involves a concentration of investment
resources in the modepn sector, exacerbating differentials and contributing to
the problems of unemployment and underemployment. Products designed for ad-
vanced country consumers are ill suited to meet the basic needs of poor
people.  However, the high productivity of much modern technology and the ef-
ficiency and economies of scale associated with advanced country products may



- 106 -
make the technology the best choice, given the absence of efficient alternatives.
Appropriate technology consists of technology more in line with developing coun-
try needs and resources; it consists of more labour intensive processes in the
modern sector, and the development of new and improved techniques and products for
the traditional sector. The promotion of more appropriate technology includes the
selection of more appropriate technology among known technologies (including
'old' techniques from advanced countries and second hand machinery) and the
development of new appropriate technologies.
10- The development strategy a country adopts strongly influence- its technolog-
ical options. If a country's consumption patterns are similar to those of
developed countries, and if its main trading orientation is with developed
countries, then the main body of technology it uses in the modern sector will need
to be that of the developed countries. Consumption patterns are likely to be
similar to those of developed countries among countries with an unequal income
distribution, an 'open' policy towards technological imports, and lac'; of in-
dipenous cultural factors which inhibit the demonstration effect. Some modi-
fication of technology is clearly possible - and indeed happensl/ - even in this
sort of situation, buIt modificationis are likely to be largely a matter of time-
lags in adopting the latest advanced country technology, more labour - inten-
sive ancilliary activities, and modifications rendered necessary by differences
in the size of market.-/
1/  See evidence of Pack (1976) on adaptation in Kenya.
2/ Scale has been established as the main source of adaptation among TNCS
(Reuber (1973).



107 -
With more equal income distribution and with obstacles (cultural, economic or
legal) preventing the domination of advarLeed country products, the potential
for appropriate and self-generated teclnology will be greater. Trade in
manufactured products with advanced countries has generally involved the import
of technology from advanced countries, at least in the initial stage. Trade
between developing countries is more likely to be consistent with the use of
locally generated appropriate technology. Appropriate technology consists both
in modification in 'modern' sector products and techniques and in upgrading
'traditional' sector technology. Even in countries with advanced country
oriented modern sectors, there is potential for appropriate technology in the
traditional sector to help raise productivity and incomes and meet basic needs.
11. Policies to promote appropriate technology consist in policies determining
the demand for (or selection of) different products and techniques, and policies
determining the supply of technologies. On the demand side, the policies
concerned are chiefly national; they include determinants of income distribution
and consumption patterns, trading strategy, policies determining control over
investible resources by different types of decision maker, and relative factor
prices. On the supply side, they include the collection of information about
different technologies and its diffusion, and research and development and the
creation of new technologies. Potet;tially, there could be an important role for
international institutions on the supply side assisting in information collection
and diffusion, and in funding appropriate research.
12. Technology transfer between developing countries could have an important
role to play in increasing developing countries technological independence
and bargaining power and in promoting appropriate technology. There are
some indications that such trade is expanding; third world MNCs are developing



- 108 -
as is third world consultancy and trade in capital goods. As with similar develop-
ments, there is a danger that new imbalances and dependencies within the third
world will arise as a result.
13. There is a complex system of relationships between past policy towards
technology, policy makers and policy making. Countries which have been heavily
dependent on foreign technology, particularly in the form of foreign investment,
find it more difficult to regulate it, than those which have relied on it to a
lesser extent and have maintained an arms length relationship with foreign
suppliers.    Similarly, patterns of production and consumption which are broad-
ly inappropriate set in force strong forces making for similar inappropriate
choices in the future.  The extent of current freedom of decision making depends
in-large part on past policies. Today's policies, then, will partly determine
future options. In this context radical delinking might be a tactical and tempo-
rary move to strengthen a country's position in future linked relationships.
1/ China's recent history can obviously be interpreted in this light.



- 109 -
References
Agarial, A. (1978)  New Scientirt, (Novecibar 9th)
Agmon, T. and C.P. Kindleberger, (1977) Multinationals from Small Colntries,
(MIT).
Annerstadt, J., (1978) "Technological Dependence:  A Permanent Phenomenon of
World Inequality?" (miLmeo, Institute of Social and Economic Planning, Roskilde
University Center).
Asian Productivity Organization,(1976) Intra-National Transfer of Technology,
(Tokyo).
Aurora, G.S. and W. Morehouse, (1974) "The Dilemma of Technological Choice in
India: The Case of the Small Tractor", Minerva.
Batanson, J., (1978) Technology and the Multinationals,(Lexington Books).
Beranek, W. and C. Ranis (1978) Science and Technology and Economic
Development, Praeger.
Bhagwati, J.N., (1968) "The Tying of Aid" in UNCTAD Conference Proceedings,
Vol. IV, (New Delhi).
Bhalla, A.S. (ed.), (1975) Technology and Emplovment in Industry,(ILO).
Bhalla, A.S. (ed.), (1979) Towards Global Action on Appropriate
Technolo;j, Pergamon
Boon, G.K., (1964) Economic Choice of Human and Physical Factors in Production,
(North Holland Publishing Co.).
Bhatt, V.V. (1975) "On Technology Policy and its Institutional Frame", World
Development 3
Boon, G.K., (1975) "Technological Choice in Metalworking with Special Reference
to Mexico",ln Bhalla (ed.).
Carlsen, J., (1975) "Different Modes of Technology Transfer." in Multrintinanal
Firms in Africa (C. Widstrand (ed.))(Uppsala).
Carlson, G.N, and G.C. Hufbauer, (1976) "Tax Barriers to Technology Transfers",
U.S. Treasury Department, O.T.A. paper 16.
Chen-Young, P.L., (1967) "A Study of Tax Incentives in Jamaica" National Tax
Journal, 20.
Chudnovsky, D., (197.8) "Pricing of Intrafirm Technological Transactions",
(prepared for conference at I.D.S., Sussex, mimeo).
Chudson, W., (1974) "The Acquisition of Technology from Multinational Corporations
by Developing Countries" Department of Social and Economic Affairs: ST/ESA/12,
U.N.



- 110 -
Cooper, C. and F. Sercovich, (1971) "The Channels and Mechanisms for the Transfer
of Technology from Developed to Developing Countries" (UNCTAD: TD/B/AC. 11/5).
Cooper, C. (ed.), (1973) Science       Technology and Development, (Frank Cass).
Cooper, C. et al, (1975) "Choice of Techniques for Can Making in Kenya, Tanzania
and Thailand" in Bhalla (ed.).
Cooper, C. (1976) "Policy Interventions for Technological Innovation in Less
Developed Countries," mimeo.
Diaz-Alejandro, C.F., (1977) "Foreign Direct Investment by Latin Americans"
in Agmon and Kindleberger (eds.).
El-Karanshawy, H.A.S., (1975) Choice of Carpet Weaving Technology in Egypt,
(Ph.D. th-sis, University of Strathclyde).
Gereffi, G., (1978) "Drug Firms and Dependency in Mexico: The Case of the
Steriod Hormone Industry", International Organization 32.
Gerschenkron, A., (1962), Economic Backwardness in Historical Perspective,
(Harvard University Press).
Gustafson, W.E., (1962) "Researcn and Development, New Products and Productivity
Change", American Economic Review, Papers and Proceedings, LII.
Helleiner, G.K., (1973) "Manufactured Exports from Less Developed Countries and
M4ultinational Firms", Economic Journal, 83.
Herre l, A.-, (1973) "Social Determinants of Science Policy in Latin America:
Explicit Science Policy and Implicit Scien7ce Policy", in Cooper (ed.).
-Hinricks, H., (1974) "Tax Reform and Fiscal Harmonization: Guatamala", in
D. Geithman (ed.) Fiscaf Policy for Industrialization and-Development in
.Latin America (Florida).
Hirsch, E., (1967), Location of Industry and International Competitiveness,
(Clarendon Press).
Hirsch, S., (197 ), "An International Trade and Investment Theory of the Firm,"
Oxford Economic Papers, 2Q.
Hufbauer, G. C., (1966), Synthetic Materials and the Theory of International Trade,
(Duckworth).
Hughes, H. and Y. Seng, (eds.), (1969), Foreign Investment and Industrialization
in Singapore, (Australian National University Press).
Hymer, S., (1976), The International Operations of National Firms (MIT Press).
I.L.O., (1972), Employment, Incomes and Equality, A Strategy for Increasing Productive
Employment, (Geneva).



-ill -
IMAT, (1978), "A Fe-asibiility Study into a New Internatioal, Mecsianism for Appropriate
Technology".
Jenkins, G., (1975), Non-Agricultural Choice of Techniques, An Annotated Biblio-
graphy of Empirical Studlies, (Institute of Commcnwealth Studies, Oxford).
Jenkins, C. and B. D. Wright, (1975), "Taxation of Multinational Corporations: the
Case of the U. S. Petroleum Industry," Review of Economics and Statistics, 57.
Johnson, H. G., (1970), "The Efficiency and Welfare Imp1ications of the International
Corporation," in C. P. Kindleberger (ed.), The Multinational Corporation, (MIT Press).
Katz, J., (1978), "Creacion de technologis en el sector manufactureo Argentino,"
El Trimestre Economico.
Katz, J. and E. Ablin, "From Infant Industry to Technology Exports: The Argentine
Experience in the International Sale of Industrial Plants and Engineering Works"
IDB/ECLA Working Paper No. 14.
Katz, J. et. al., (1978) "Productivity, Technology and Domestic Effects in
Research and Development," IBD/ECLA Working Paper No. 13.
Kidron, M. (1965), Foreign Investment in India, (O.U.P.)
King, T., (1970),- Mexico, Industrialization and Trade Policies Since 1940, (O.U.P.)
Kopits, G. F.,. (l976a),"Taxation -d Mulf-inational Firm Behavior: A Critical
Survey-," I.M.F. Staff Papers, 23.
Kopits, G. F., (1976b), "Intra-Firm Royalties Crossing Frontiers and Transfer
Pricing Behavior," Economic Journal, 86.
Kuznets, S., (1966), Modern Economic Growth: Rate, Structure and Spread, (Yale).
t .
LaIl, S., (19-73), "Transfer Pricing by Multinational Firms," Oxford Bulletin of
Economics and Statistics, 35.
Lall,- S., (1975), "Major Issues in Transter cf Technology to Developing Countries,
A Case Study of the Pharmaceutical Industry," UNCTAD: TD/B/C.6/4.
Lall, S. (1978), "Third World Technology Transfer and Third World Transnational
Companies," (mimeo:  Oxford University Institute of Economics and Statistics).
Landes, D.C., (1972) The Unbound Prometheus, C.U.P.
Lecraw, D., (1977), "Direct Investment by Firm from Less Developed Countries,"
Oxford Economic Papers, 29.



- 112 -
Lee, N.K., (1975) "Technological Development and Role of R and D Institutes in
Developing Countries - The Korean Case," ILO, WEP 2-22/WP 25, Geneva.
Lent, G. E. (1967), "Tax Incentives for Developing Countries," I.M.F. Staff
Papers, 14.
Maizels, A., (1963), Industrial Growth and World Trade (C.U.P.).
Mansfield, E., (1969), The EconomLcs of Technological Change,(Longmans).
Marga Institute, (1975), "Major Issues Arising from the Transfer of Technology,
A Case Study of Sri Lanka," UNCTAD: TD/B/C.6/6.
Matovelle, A., (1974), "Notas Sobre el Processo de Transferencia de Tecnologia
en el Sector Industrial," (mimeo, Junta Nacional de Planificacion).
Maxwell, P. , "Learning and Technical Change in the Steelplant of Acindas S.A.
in Rosario, Argentina," IDB/ECLA, monograph no. 4.
Merrill, R. S., (1968), in D. Sills (ed.) Encyclopaedia in Social Sciences',
(Macmillan).
XQrehouse, W.,- (1978), "Technological Innovation and Organizational Effectiveness:
Swaraj in the Tractor Sweepstakes," (mimeo,Columbia University).
Musgrave, P.- (1969), United States Taxation of Investment Income:  Issues and Argu-
-ments (Harvard)
Myt ±Ka, L., (1977), "Regulating Direct Foreign Investment and Technology
?ransfer in the Andean Group," Journal-of Peace Research, XIV.
Mytelka, L., (1978), ",Licensing and Technological Dependence in the Andean
Group," World Development, 6.
Neers6, P., (1975) "Some Aspects of India's Policies on the Import of Technology,"
Development and Change, 6.



- 113 -
Newfarmer, R. S., and W. F. Mueller, (1975)) Multinational CorpoTations in
Brazil and Mexico:  Sitructural Sources of Economic and Non-Econoalnc Power,
Report to the Subcommittee cn Multinational Corrorations, IJ. S. Congress.
O'Connor, W. F., and S. M. Russo, (1973), "A Study of Corporate Experience with
Section 482," The European Tax Review
Oxman, G. and F. Sagasto, (1972), "Transferencia de Tecnlologia hacia los Paises
del Grupo Andino," (O.A.S.)
Ozawa, T:, (1966), Imitation, Innovation and Trade: A Study of Foreign
Licensing Operations in JapanL (Colombia UnIversity, Ph.D.).
Pack, H., (1976), "The Substltution of Labor for Capital in Kenyan Manufacturing."
Economic Journal 86
Pack, H1., (1978a), "Macroeconomic ImDlications of Factor Substitution itn Industrial
Processes," (mimeo, World Bank).
Pack, H., (1978b), "The Capital Goods Industries in LDCs: A Survey." (mimeo, World Bark)
Phillips, A. O., (1969), "The Significance of Nigeria's Income Tax Relief Incentive,"
The Nigerian Journal of Economic and Social Studies, II.
Pickett et al, (1974), "The Choice of Technology, Economic Efficiency and Employment
in Developing Countries," World Development, 2.
Posner, M., (1961), "International Trade and Technical Change," Oxford Economic
Papers, 13.
Posner, M., (1970),"Technical Change, International Trade and Foreign Investment,"
iln P. Streeten (ed.) Unfashionable Economics, Essays in Honour of Lord Balogh,
(Weidenfeld and Nicholson).
Pursell, G. and Y.W. Rhde (1978) "A Firm Level Study of Korean Exports:
Technology," Research Report No. 2, World Bank, mimeo.
Quinn, J. B., (1969), "Technology Transfers by Multinational Companies," Harvard
Business Revieu.
Ranis, G., (1972), "Technology, Employment and Growth: Lessons from the Experience
of Japan`" in Automation in Developing Countries, (I.L.O.)
*Ranis, G. (1973), "Industrial Sector Labor Absorption," Economic Development and
Cultural Change, 21.
Rawski, T. G., (1978), "Industrialization, Technology and Employment in the People's
Republic of Cbina," (World Bank Staff Working Paper, No. 291).



- 114 -
Reddy, A. K., (1979) "National and Regional Technology Institutions: An
Assessment," in Bhalla (ed.).
Reuber, G. L. et al, (1973), Private Foreign Investment in Development,
(Clarendon).
Roberts, J. (1973), "Engineering Consultancy, Industrialisation and Development" in
Cooper (ed.).
Robinson, A., (1976), Future Tasks for U.N.D.P., Report to the Administrator
of U.N.D.P.
Ronstadt, R., (1977), Researcn and Development Abroad by U. S. Multinationals,
(Praeger).
Roumeliotis, P. V. and C. P. Goulemis, (1978), "Transfer Pricing and the Power
of-Transn-tional Enterprises in Greece," (prepared for conference at I.D.S.,
Sussex, mimeo).
Sagasti, F. (1975), "The INTIC System forrIndustriai.Teechnology Policy in Peru,"
World Development, 3.
Sagasti, F., (1978) Science and Teclnology for Development, Main Comparative
Report of the Science and Technology nolicy Implements Project, I.R.D.C.
Science Polic3 Yesearch Unit, (S.P.R.U.), (1972), "The Transfer of Technology to
Latin America," (O.A.S.).
.-.covich, F. C., (1974,-)`Foreign Technology and--Gontrol in-the Argentinian
.niduB try'," (Ph.D., Sussex).
-Sercovich, F.c.;(1978) "Design Engineering and Endogenous Technical Change,"
IDB/ECLA W'orking Paper No. 19.
Sigurdson, J., (1977), Rural Industrialization in China, (Harvard).
Speight, A.N.P., (1975), "Cost Effectiveness and Drug Therapy," Tropical Doctor.
Spencer, D. L., (1970), Technology Gap in Perspective, (Spartan).
Sronek, I., (1978), "The Experience of Socialist Countries of Eastern Europe in
The Transfer for Technology to Developing Countries," (UNCTAD: TD/B/C.6/26).
Stewart, F., (1977), Technology and Underdevelopment, (Macmillan).
Stewart, F., (1978), "Technology Transfer: A Consideration of the Role of Taxation,"
(mimeo, Institute of Commonwealth Stuidies, Oxford).



Stewart, F., (1979), '"International Mechaniesms for Appropor.l.ata Technology," f.n
Bhalla (ed.).
Stopford, J. M. and L. T. Wells, (1972), Mancqja t.he Multinational En erprise,
(Basic Books).
Subrahmanian, K. K., (1972), Import o': Ca+ital and Technology, (Paoples Publishinlg).
Teece, D. J., (1976), The Multinational Corporation and the Resource Cost of
International Technology Transfer, (Ballinger).
Timmer, C. P. et al, (1975), The Choice of Technology in Developing Countries,
(Some Cautionary Tales), (!Harvard)
United Nations, (1972), Report of Experts on Tax Treaties Between Developed and
Developing Countries, (Third Report).
UNCTAD, (1974), "Major Issues Arising in the Transfer of Technology to Developing
Countries: A Case Study of Ethiopia," (TD/B/AC.ll/21).
UNCTAD, (1975a), "Major Issues Arising from the Transfer bf Technology to Developi
Countries," (TD/B/AC.ll/1., Rev. Z.j
UNCTAD, (lY7b), "lhe Role ot ttie Yatent System in the Tranter of Tecnnology to
Developing Countries," (TD/B/AC.11/19, Rev. 1).
UNCTAD, (1975c), "tAn International Code of Conduct on the Transfer of Technology,"
(TD/B/C6, 1975/AC.1/2/Supp. 1. Res. 1).
UNCTAD, (1976), "Technological Dependence: The Nature, Consequences and Policy
Implications," in Proceedings of the United Nations Conference on Trade and
Development, Fourth Session, Vol. III, Basic Documents, (TD/190).
UNCTAD, (1977a), "The International Patent System: The Revision of the Paris
Convention for Protectioni of Industrial Property,," (TD/B/C.6IAc.32).
UNCTAD, (]977b), "The Impact of Trade Marks on the Development Process of Developing
Countries," (TD/B/C.6IAC.33).
UNCTAD, (1977c), "Dominant Positions of Markct Power of Transnational Corporations:
Use of Transfer Pricing Mechanism," (ST/MD/6).
UNCTAD, (1978a), "Case Studies in the Transfers of Technologh: Policies for Transter
and Development of Technology in pre-war Japan, (1868-1937)," (TD/B/C.6/26).
UNCTAD, (1978b), "Technology Planning in Developing countries:  A irelizinary ieview,"
(TD/B/C.6/29).
UNIDO, (1978), "Recent Developments in the Regulation of Foreign lechnology ill
Selected Developing Countries," (ID/WG.275/8).
USAID, (1971). "Technology aild Economic in International Development: Report of u
Seminar"s Office of Science and Technology, Washington, D.C.



- 116 -
vpttsos, C. V., (1971), "The Process of Commercialization of Technology in the
Andean Pact," (O.A.S.).
Vaitsos, C. V., (1973), "Patents Revisited; Their Function in DeveloDing
Countries," in Cooper (ed.).
Vaitsos, C. V., (1974), Intercountry Income Distribution and Transnational
Enterprises, (O.U.P.)
Vaitsos, C. V. (1975), "Foreign Inivestment and Prodtctive XCnot-edge," in G, P, 1rb
and V. Kallab, (eds.), Beyond Dependency.
Vernon, R., (1966), "International Investment and International Trade in the
Product Cycle," Quarterly Journal of Economics, LXXX.
Wells, J. (1977) "The Diffusion of Durables in Bragil and its Implications
for Recent Controversies Concerning Brazilian DeveloDment," Cambridge Journal of
Economics, 1.
Wells, L. T., (1977), "The Internationalization of Firms from Developing Countries,"
in Agmon and Kindleberger, (eds.)
Wells, L.T. (1978) "Foreign Investment from the Third World: The Experience of
Chinese firms from Hong Kong", Columbia Journal of World Business.
Wells, L.T., (1979) "Developing Country Investors in Indonesia", Bulletin of
Indonesian Economic Studies.
White, L.J. (1978)" "The Evidence on Appropr Late Factor Proportion',for Manu-
facturf l in Less Developed Countriesi' A Survey" Economic Development and
(ultu al Change, 27.
horld Bank (1978) Employment and Development of Small Enterprises.



World  Bank                                  NEW                                  indicate the effects of these systems on
the allocation of resources, interna-
Publications                             Sub-Saharan Africa: Progress             tional trade, and economic growth.
Of Related                               Report on Development                    The Johns Hopkins University Press, 1982;
Prospects and Programs                   paperback 1984. 416 pages (including ap-
Interest                                 A review of the foregoing book two      pendixes, index).
years later.                            LC-81-15558. ISBN 0-8018-2569-5, hard-
Highlights some of the major problems    cover; ISBN 0-8018-2570-9, paperback.
that African governments have faced      Stock Nos. JH 2569, $39.95 hardcover; JH
in designing and implementing pro-      2570, $18.50 paperback.
grams of reform. Examines changes
that have been introduced in the level,  Eastern and Southern Africa:
pattern, and design of World Bank        Past Trends and Future
economic work and operations in sup-     Prospects
port of such reforms. Reviews the ex-    Ravi Gulhati
tent of donor government responses to
the increased and changed needs of       Staff Working Paper No. 413. 1980. 24
African countries for external assis-   pages.
tance.                                   Stock No. WP 0413. $3.
1983. 37 pages.
Stock No. BK 9169. Free.                 Economic Growth and Human
Resources
Accelerated Development in                                                        Norman Hicks, assisted by
Sub-Saharan Africa: An                      NEW                                  Jahangir Boroumand
Agenda for Action                                                                 Staff Working Paper No. 408. 1980. 40
Ag thendall of 1979, the African Gover-  Decentralization in Developing          pages (including 3 appendixes, bibliog-
ln the fall of 1979. the African Cover-  Countries: A Review of Recent            raphy, and references).
nors of the World Bank addressed a
memorandum to the Bank's president       Experience                               Stock No. WP 0408. $3.
expressing their alarm at the dim eco-   Dennis A Rondinelli, John R.
nomic prospects for the nations of sub-  Nellis, and G. Shabbir Cheema               NEW
Saharan Africa and asking that the       Reports on the objectives of decentrali-
Bank prepare a "special paper on the     zation. Notes that many developing       Economic Reform m  Socialist
economic development problems of         countries began decentralization dur-    Countries: The Experiences of
these countries" and an appropriate     ing the last 10 years to find ways of     China, Hungary, Romania, and
program for helping them. This report,   using limited resources more effec-      Yugoslavia
building on the Lagos Plan of Action,   tively. Evaluates types of decentraliza-  Peter T. Knight
is the response to that request.        tion and conditions and factors affect-               g
The report discusses the factors that    ing the implementation of policies       Describes the Soviet-style system of
explain slow economic growth in Af-      supporting decentralization. Provides    intra-zed plannsng as a prelude to an
rica in the recent past, analyzes pohcy  information on political commitment,     n-depth discussion of reform design
changes and program orientations         administrative support, effective de-    as experenced in Chmia, Hungary, Ro-
needed to promote faster growth, and     sign and organization of programs,       mania, and Yugoslavia. Some lessons
concludes with a set of recommenda-      and the need for adequate resources,     from the reform of these countries are
tions to donors, including the recom-    Annexes look at Indonesia's provincial   noted as possibly relevant to future at-
mendation that aid to Africa should      development program, China's "pro-       tempts at Improving the efficiency
double in real terms to bring about re-  duction responsibility" system, and      both of soialdst economies and state
newed African development and            Tunisia's deconcentration program.       sectors of mixed economies.
growth in the 1980s. The report's        Staff Working Paper No. 581. 1983. 112   Staff Working Paper No. 579. 1983. 131
agenda for action is general; it indi-   pages.                                   pages.
cates broad policy and program direc-    Stock No WP 0581. $5                     ISBN 0-8213-0229-0. Stock No. WP 0579.
tions, overall priorities for action, and                                         $5o
key areas for donor attention. Like the  Development Strategies in
Lagos Plan, the report recognizes that   Semi-Industrial Economies                The Extent of Poverty in Latin
Africa has enormous economic poten-      Bela Balassa and Associates              America
tial, which awaits fuller development.   Provides an analysis of development      Oscar Altimir
1981; 3rd printing 1983. 198 pages (in-  strategies in semi-industrial economies  Staff Working Paper No. 522. 1982. 117
cluding statistical annex, bibliography),  that have established an industrial    pages.
French: Le developpement accelere en af-  base. Endeavors to quantify the sys-    IB   -2301-.SokN.W   52
rique au sud du Sahara: programme indi-  tems of incentives that are applied in  ISBN0-8213-0012-1. StockNo. WP0522.
catif d'action.                          six semi-industrial developing econom-   $
Stock Nos. BK 9114 (English) and BK      ies-Argentina, Colombia, Israel, Ko-     Prices subject to change without notice
9116 (French). Free.                     rea, Singapore, and Taiwan-and to        and may vary by country.



NEW                                    Impllemeniting Programs of                Managing Project-Related
Human Development                         Technical Assistance: The
The Effects of Corruption on               Edited by Peter T. Knight;                Lessons of Success
Administrative Performance:                prepared by Nat J. Colleta, Jacob         Francis Lethem and Lauren
Illustrations from  Developing             Meerman, and others.                      Cooper
Countries                                  Staff Working Paper No. 403. 1980. 376    Practical guidance for technical assis-
David J. Gould and Jose A.                pages (including references).              tance designers and appraisers based
Amaro-Reyes                               Stock No. WP 0403. sI5.                    on feedback from aid agencies and
technical assistance recipients in Af-
Analyzes the effects of corruption on      International Technology                  rica, Asia, and the Middle East who
the administrative performance of de-
veloping countries. Examines the polit-    Transfer: Issues and Policy               participated in successful projects.
veloping countries. Examines the polit-                        ~~~~Planners and practitioners will benefit
ical, economic, and social factors that    Options                                   from the proven advice in this report.
contribute to its incidence. Data from     Frances Stewart                           Covers identification of need, design
Asia, Africa, and Latin America form       Staff Working Paper No. 344. 1979. 178    of services and administration, imple-
the basis for discussion.                 pages (including references).              mentation and management of techni-
Staff Working Paper No. 580. 1983. 60      Stock No. WP 0344. S5.                    cal assistance.
pages.                                                                               Staff Working Paper No. 586. 1984. 104
ISBN 0-8213-0259-0. Stock No. WP 0580.                                               pages. English and French.
$3.                                           MNW                                    ISBN 0-8213-0239-6. Stock No. WP 0586.
First Things First: Meeting                                                          $5.
Basic Human Needs in the
Developing Countries                               rhf rp- .. k- k.-                     NEW
Paul Streeten, with Shahid Javed                     MM-ngn   h    ulcSrieit
Burki, Mahbub ul Haq, Norman                                                         Managing the  Public Servsce in
Hicks, and Frances Stewart                                                           aDeProspeCts
The basic needs approach to economic                                                 Slk ospedts
development is one way of helping the                                                Selcuk Ozgediz
poor emerge from their poverty. It en-                                               Identifies better ways to manage hu-
ables them to earn or obtain the ne-                                         .       man resources to meet escalating de-
cessities for life-nutrition, housing,                .                              mand for improved public services in
water and sanitation, education, and                                                 developing countries, where public
health-and thus to increase their pro-                                               service employment is growing four
ductivitv.                                                                           times faster than in developed coun-
This book answers the critics of the                                                tries. Resultant problems in personnel
basic needs approach, views this ap-                                                 management, public service training
proasich ns approgich, viepwi this evol-    --                                       programs, and the applicability of
proach as a logical step in the evolu-                                               western management practices in de-
tion of economic analysis and develop-                     -   X  L                                          a
men poiy an  penta--..-                                                          veloping country settings are consid-
.nent policy, and presents aerd
clearsighted interpretation of the is-                                               ered.
sues. Based on the actual experience of   The Japanese and  Korean                   Staff Working Paper No. 583. 1983. 1g0
various countries-their successes and     Experences in Managing                     pages.
failures-the book is a distillation of    Development                                ISBN 0-8213-0243-4. Stock No. WP 0583.
World Bank studies of the operational     Miyohei Shinohara, Toru                    S..
implications of meeting basic needs. It   Yanagihara, and Kwang Suk Kim              Models of Growth and
also discusses the presumed conflict      The special style of economic manage-       distribtio wfo       and
between economic growth and basic         ment thal: has produced the "economic      Distribution for Brazil
needs, the relation between the New       miracles" of these two countries is        Lance Taylor, Edmar L. Bacha,
International Economic Order and          analyzed. Of particular interest are in-   Eliana Cardoso, and Frank J. Lysy
basic needs, and the relation between     ferences about the power of prices and     Explores the Brazilian experience from
human nights and basic needs.             role of competition, both internal and     the point of view of political economy
Oxford University Press, 1981; 3rd paper-  external. Korean and Japanese eco-        and computable general equilibrium
back printing, 1984. 224 pages (including  nomic management are characterized       income distribution models.
appendix, bibliography, index).           as pragmatic and flexible.                 Oxford University Press, 1980. 368 pages
LC 81-16836, ISBN 0-19-520-368-2, hard-   Staff Working Paper No. 574. 1984. 98     (including references, appendixes, index).
cover; ISBN 0-19-520-369-0, paperback.    pages.                                     LC 80-13786. ISBN 0-19-520206-6, hard-
Stock Nos. OX 520368, $18.95 hardcover;   ISBN 0-821;3-0233-7. Stock No. WP 0574.   cover; ISBN 0-19-520207-4, paperback.
OX 520369, $8.95 paperback.               $3.                                        Stock Nos. OX 520206, $27.50 hardcover;
The Hungarian Economic                    Levels of Poverty: Policy and              OX 520207, $14.95 paperback.
Reform, 1968-81                           Change                                     Patterns of Development, 1950-
Bela Balassa                              Amartya :Sen                               1970
Staff Working Paper No. 506. 1982. 31     Staff Working Paper No. 401. 1980. 91      Hollis Chenery and Moises
pages (including references).             pages (including references).              Syrquin
Stock No. WP 0506. $3.                    Stock No. WP 0401. $3.                     A comprehensive interpretation of the



structural changes that accompany the      in developing countries during the         Meeting Basic Needs: An
growth of developing countries, using      post-World War II period. Analyzes.        Overview
cross-section and time-series analysis     problems, successes and failures,          Mahbub ul Haq and Shahid Javed
to study the stability of observed pat-    emerging trends, and lessons learned.         burki
terns and the nature of time trends.       Staff Working Paper No. 576. 1983. 71      Bureo
Oxford University Press, 1975; 4th paper-  pages.                                     Presents a summary of the main find-
back printing, 1984. 250 pages (including  ISBN 0-8213-0303-1. Stock No. WP 0576.     ings of studies undertaken in the
technical appendix, statistical0appendix,W$3                                          World Bank as part of a program for
hibliography, index),                                                                 reducing absolute poverty and meeting
tibliography, index).                                          ~~~~~~~~~~~~~basic needs.
LC 74-29172. ISBN 0-19-920075-0, hard-          NEW                                    1980. 28 pages (including 2 annexes).
cover; ISBN 0-19-920076-9, paperback.
Stock Nos. OX 920075, $19.95 hardcover;     Policies for Strengthening                 Stock Nos. BK 9015 (Arabic), BK 9016
OX 920076, $8.95 paperback.                 Local Government in                       (English), BK 9017 (French), BK 9018
Spanish: La estructural del crecimiento    Developing Countries                       (Japanese), BK 9019 (Spanish). $3 paper-
economico: un analisis para el periodo      Glynn Cochrane
1950-1970. Editorial Tecnos, 1978.            'k   t th        b   hih hihl           Shelter
ISBN 84-309-0741-6, Stock No. IB 0523,     centralized govewabents effectively        Anthony A. Churchill
$8.95                                       utilize and improve resources within       Defines the elements that constitute
local governments. The need for            shelter; discusses the difficulties en-
NEW                                    strengthening personnel performance        countered in developing shelter pro-
at the local level is discussed, includ-   grams for the poor; estimates orders of
Pioneers in Development                    ing an analysis of useful methodology.     magnitude of shelter needs for the
Edited by Gerald M. Meier and              Discussion of local revenue sources        next twenty years; and proposes a
Dudley Seers                               and patterns of revenue generation         strategy for meeting those needs.
Some of the pioneers in development        provides a framework from which to         1980. 39 pages.
economics-Lord Bauer, Colin Clark,         examine factors affecting government       Stock Nos. BK 9020 (Arabic), BK 9021
Albert 0. Hirschman, Sir Arthur            performance at higher levels.              (English), BK 9022 (French), BK 9023
Lewis, Gunnar Myrdal, Raul Prebisch,       Staff Working Paper No. 582. 1983. 76      (Spanish). $3 paperback.
Paul N. Rosenstein-Rodan, W.W. Ros-        pages.                                     Water Supply and Waste
tow, H. W. Singer, and Jan Tinber-         ISBN 0-8213-0240-X. Stock No. WP           Dpa         l
gen--offer a retrospective view of the     0582. $3.                                  Disposal
formative decade after World War 1I                                                   Discusses the size of the problem of
when they made their seminal contri-       Policy Interventions for                    meeting basic needs in water supply
butions to the subject. In individual      Technological Innovation in                and waste disposal and its significance
papers, the pioneers recapture the in-     Developing Countries                        to development in the context of the
tellectual excitement, expectations, and   Charles Cooper                              International Drinking Water Supply
activism of that period and provide                Cooper                              and Sanitation Decade. Examines the
rare autobiographical detail and insight   Staff Working Paper No. 441. 1980. 59       Bank's past role in improving water
into why they said what they did and       pages (including references).              supply and waste disposal facilities in
what they now think about the state of     Stock No. WP 0441. $3.                      developing countries and draws con-
development thought and policy.                                                       clusions for the future.
Commentary is provided by econo-           Poverty and Basic Needs Series              1980. 46 pages.
mists of the succeeding generation,
who reappraise their ideas with the        A series of booklets prepared by the        Stock Nos. BK 9024 (Arabic), BK 9025
benefit of hindsight. General over-        of bastc needs W dBank o  the subject      (English), BK 9026 (Frenc.), BK 9027
views of the subject have been written     o  ai  ed.Tesre  nlds(pns)  3pprak
by Gerald Meier and Paul Streeten.         general studies that explore the con-
cept of basic needs, country case stud-
Oxford University Press. May 1984.         ies, and sectoral studies.                  Poverty and the Development
About 400 pages. Hardcover only.                                                       of Human Resources: Regional
ISBN 0-19-520452-2. Stock No. OX            Brazil                                     Perspective
520452, $29.95.                             Peter T. Knight and Ricardo J.             Willem Bussink, David Davies,
Moran                                      Roger Grawe, Basil Kavalsky, and
NEW                                    An edited and updated edition of the        Guy P. Pfeffermann
more detailed publication, Brazil: Hu-     Staff Working Paper No. 406. 1980. 199
Planning in Developing                     man Resources Special Report (see de-       pages (including 7 tables, 2 appendixes,
Countries: Lessons of                      scription under Country Studies hst-       reperences, footnotes).
Expenrience                                ing).                                       Stock No. WP 0406. $5.
Ramgopal Agarwala                          1981. 98 pages (including statistical ap-
Reports that, in most developing           pendix, map). English.                      Poverty and Human
countries, planning failed to live up to   Stock No. BK 9028. $5.                      Development
expectations. Outlines steps to remedy     Malnourished People: A Policy               Paul Isenman and others
this situation. Recommends a general       V                                           Since economic growth alone has not
reorientation and emphasis on the          View                                        reduced absolute poverty, it has been
most effective areas in the initial for-   Alan Berg                                   necessary to consider other strategies.
mulation of plans.Provides an in-depth     Please refer to Food and Nutrition for      The strategy examnined in this study-
review of the experience of planning       information.                                human development-epitomizes the



idea that poor people should be           LC 79-18026. ISBN 0-19-520094-2, hard-    in public administration and manage-
helped to help themselves. Four chap-     cover; ISBN 0-19-520095-0, paperback.     ment training (PAMT) in developing
ters provide an overview of alternative   Stock Nos. OX 520094, $34,50 hardcover;   countries. Highlights various aspects
strategies; a detailed look at health,    OX 520095, $22.95 paperback.              of training and career development
education, nutrition, and fertility; les-  French: Changement des structures et poli-  programs including the impact of
sons from existing programs; and an      tique de developpement. Economica, 1981.   PAMT programs in developing coun-
examination of broader issues in plan-    ISBN 2-7178-0404-8, Stock No. IB 0540,   tries during the past 30 years, the rea-
ning.                                     $12.95'                                   son for successful programs, and some
Oxford University Press. 1982. 96 pages   Spanish: Cambio estructural y politica de  innovative training designs and meth-
(including statistical appendix).         desarrollo. Editorial Tecnos, 1980. ISBN  ods.
LC 82-2153. ISBN 0-19-520389-5. Stock     84-309-0845-5, Stock No. IB 0612,         Staff Working No. 584. 1984. 127 pages.
No. OX 520389. $7.95.                     $12.95.                                   ISBN 0-8213-0234-5.Stock No. WP 0584.
Reforming the New Economic                    NPn
Mechanism  in Hungary                        _    w                                 Tribal Peoples and Economic
Bela Balassa                              Successes and Failures in                 Development: Human Ecologic
Staff Working Paper No. 534. 1982. 56     Meeting the Management                    Considerations
pages.                                    Challenge: Strategies and Their           Robert Goodland
ISBN 0-8213-0048-2. Stock No. WP 0534.    Implementation                            At the current time, approximately 200
$3.                                       Milan Kubr and John Wallace               million tribal people live in all regions
of the world and number among the
Social Infrastructure and                 Examines the original nature of key       poorest of the poor. This paper de-
Services in Zimbabwe                      management problems faced by devel-       scribes the problems associated with
Rashid Faruqee                            oping countries and outlines proven       the development process as it affects
Staff Working Paper No. 495. 1981. ill    strategies for improving management       tribal peoples; outlines the requisites
pages (including bibliography, map).      competence and performance. Presents      for meeting the human ecologic needs
approaches to management develop-        of tribal peoples; and presents general
Stock No. WP 0495. $5.                    ment recently introduced by enter-        principles that are designed to assist
prises and institutions such as busi-    the Banks staff and project designers
ness clinics and self-development        in incorporating appropriate proce-
NEW                                   programs for management as well as        dures to ensure the survival of tribal
tailored and action-oriented programs.   peoples and to assist with their devel-
Some Temporal Aspects of                  Staff Working Paper No. 585. 1984. 120    opment.
Development: A Survey                     pages.                                    1982, 118 pages (including 7 annexes, bib-
R. S. Eckaus                              ISBN 0-8213-0254-X. Stock No. WP 0585.    liography).
Identifies, quantifies, and explains fac-  $5                                       ISBN 0-8213-0010-5. Stock No. BK 0010.
tors that lengthen project design and     Tourism-Passport to                       $5.
completion. Analyzes changes that         Development? Perspectives on
take place over time in principal areas   the Social and Cultural Effects           The Tropmes and Economic
that affect economic development-la-                                                Development: A Provocative
bor supply, social structures and their   of Tourism  in Developing                 Inquiry into the Poverty of
behavior, government support func-        Countries                                 Nations
tions, market organization and effi-      Emanuel de Kadt, editor
ciency, and capital productivity. Notes   The first serious effort at dealing with  Andrew M. Kamarck
that development has proved to be a       the effects of tourism development in     Examines major characteristics of the
relatively slow process.                  a broad sense, concentrating on social    tropical climates that are si.nificant to
Staff Working Paper No. 626. 1983. 52     and cultural questions.                   economic development.
pages.                                    A joint World Bank-Unesco study. Oxford  T2e Johns Hopkins University Press, 1976;
ISBN 0-8213-0300-7.Stock No. WP 0626.     University Press, 1979. 378 pages (includ-  2nd p,inting, 1979. i28 pages (xncludng
$3.                                       ing maps, index).                          naps, bibiography, index).
LC 79-181126. ISBN 0-19-520149-3, hard-   LC 76-17242. ISBN 0-8018-1891-5, hard-
Structural Change and                     cover; ISBN 0-195N20150--572149-3  ard    cover; ISBN 0-8018-1903-2, paperback.
Development Policy                        Stock Nos. OX 520149, $24.95 hardcover;   Stock Nos. rH 1891, $12.50 hardcover; Jl
Hollis Chenery                            OX 520150, $9.95 paperback.               1903, S5 paperback.
A retrospective look at Chenery's                                                   Spanish: Los tropicos y desarrollo econom-
thought and writing over the past two           --                                  ico: reflexiones sobre la pobreza de las na-
decades and an extension of his work          NEW'                                  ciones. Editorial Tecnos, 1978. ISBN 84-
in Redistribution with Growth and Pat-                                              309-0740-8, Stock No. IB 0529. $5.
terns of Development. Develops a set of   Training for Public
techniques forTanalyzing structural       Administration and                        Twenty-five Years of Economic
changes and applies them to some ma-                                                Development, 1950 to 1975
jor problems of developing countries      Management in Developing                  David Morawetz
today.                                    Countries: A Review                       A broad assessment of development
Oxford University Press, 1979; 2nd paper-  Samuel Paul                              efforts shows that, although the devel-
back printing, 1982. 544 pages (including  Analyzes the results of a survey of the  oping countries have been remarkably
references, index).                       trends, developments, and problems        successful in achieving growth, the



distribution of its benefits among and       LC 77-17243. ISBN 0-8018-2134-7, hard-     ica, 1978. ISBN 2-7178-0038-7, Stock No.
within countries has been less satisfac-     cover; ISBN 0-8018-2092-8, paperback.       IB 0546, $7.95.
tory.                                        Stock Nos. JH 2134, $16.50 hardcover; JH    Spanish: Veinticinco anos de desarrollo
The Johns Hopkins University Press, 1977;    2092, $7.95 paperback.                      economico: 1950 a 1975. Editorial Tecnos,
3rd printing, 1981. 136 pages (including     French: Vingt-cinq annees de developpe-     1978. ISBN 84-309-0792-0. Stock No. IB
statistical appendix, references).           ment economique: 1950 a 1975. Econom-       0529. $7.95.
Woid D nt Repvrt 1953
World Development Report
issue brings not only an overview of      Management and
the state of development, but also a                                S
detailed analysis of such topics as       Development Sen                  es
structural change, the varying experi-
ences of low- and middle-income            Decentralization in Developing
countries, the relation of poverty and    Countries: A Review of Recent
human resource development, global         Experience. WP 0581. $5. (Development)
and national adjustment, and agricul-     Development Finance Companies, State
ture and food stability. Each contains a    and Privately Owned: A Review. WP
statistical annex, World Development      0578. $3. (Finance and Debt)
A large-format series of annual studies Indicators, that provides profiles of    Economic Reform in Socialist Countries:
of 200 or more pages, the World Devel- more than 120 countries in twenty-five    The Experiences of China, Hungary,
opment Report, since its inception, has  multipage tables. The data cover such    Romania, and Yugoslavia. W)  0579. $5.
been what The Guardian has called "a  subjects as demography, industry,
most remarkable publication. It is the  trade, energy, finance, and develop-      The Effects of Corruption on
nearest thing to having an annual re-  ment assistance and such measures of       Administrative Performance: Illustrations
port on the present state of the planet social conditions as education, health,   from Developing Countries. WP 0580. $3.
and the people who live on it.' Each   and nutrition.                             (Development)
The Japanese and Korean Experiences in
Managing Development. WP 0574. $3.
World Development Report 1983          World Development Report 1978              (Development)
Profiles the economies of 125 nations.  (Discusses the development experi-        Managing Project-Related Technical
Explores the role of management in    ence, 1950-75, development priorities       Assistance: The Lessons of Success. WP
successful development enterprises.   in the middle-income developing             0586. $5. (Development)
Provides reliable up-to-date statistics.  countries, and prospects for alleviating  Managing the Public Service in
Oxford University Press, 224 pages (in-   poverty.)                               Developing Countries: Issues and
cluding maps, annex, World Development Ordering information for World Devel-      Prospects. WP 0583. $3. (Development)
Indicators). ISBN 0-19-520432-8; $20   opment Report. All editions are $8, pa-    Managing State-Owned Enterprises. WP
hardcover; $8 paperback.               perback.                                   0577. $5. (Industry)
World Development Report 1982          World Development Report 1983:             Planning in Developing Countries:
(Tackles the principal issues and obsta- Stock Nos. OX 520431 (English) hard-     Lessons of Experience. WP 0576. $3.
cles in the efforts to improve the per-  cover; OX 520432 (English) paperback.    (Development)
formance of world agriculture.)        World Development Report 1982:             Policies for Strengthening Local
World Development Report 1981          Stock Nos. BK 0086 (Arabic), OX 503225     Government in Developing Countries.
(Discusses adjustment-globai and na- (English), BK 0087 (Spanish), lB 0607        WP 0582. $3. (Development)
tional-to promote sustainable! growth  (French), IB 0606 (Japanese).              Price Distortions and Growth in
in the changing world economy.)        World Development Report 1981:             DFeveloping Countries. Wt   0575. $3.
World Development Report 1980          Stock Nos. IB 0594 (Arabic), OX 502998     Successes and Failures in Meeting the
(Discusses adjustment and growth in       g                                       Management Challenge: Strategies and
the 1980s and poverty and human de-  (Spanish).                                   Their Implementation. WP 0585. $5.
velopment.)                            World Development Report 1980:             (Development)
Stock Nos. OX 139419 (English), IB 0599   Training for Public Administration and
World Development Report 1979          (Spanish).                                 Management in Developing Countries: A
(Discusses development prospects and  World Development Report 1979:              Review. WP 0584. $5. (Development)
international policy issues, structural   Stock No. OX 502638 (English).          Please see category indicated after
perience and issues.)                  World Development Report 1978:             each title for description and
Stock No. OX 139414 (English).             bibliographic informnation.



The World Bank
Publications Order Form
SEND TO: YOUR LOCAL DISTRIBUTOR OR TO WORLD BANK PUBLICATIONS
(See the other vide of this form.)         P.O. BOX 37525
WASHINGTON, D.C. 20013 U.S.A.
Date
Name                                                 Ship to: {Enter if different from purchaser)
Title                                                Name
Firm                                                 Title
Address                                              Firm
City               State_      Postal Codee          Address_               -
Country             TeIephone (    )                 City             State_       Postal Code_
Purchaser Reference No.                              Country             e'Ilephone
Check your method of payment.
Enclosed is my O Check O International Money Order O Unesco Coupons O International Postal Coupon.
Make payable to World Bank Publications for U. S. dollars unless you are ordering from your local distributor.
Charge my C VISA D MasterCard D American Express O Choice. (Credit cards accepted only for orders addressed
to World Bank Publications.)
Credit Card Account Number            Expiration Date                     Signature
D Invoice me and please reference my Purchase Order No.
Please ship me the items listed below.
Customer Interna
Stock Number  Author/ Title                                       Routing Code   Quantity  Unit Price  Total Amount S
All prices subject to change.  Prices may vary by country.  Allow 6-8 weeks for delivery.
Subtotal Cost  $___
Total copies _____ Air mail surcharge if desired ($2.00 each)   $____
Postage and handling for more than two complimentary items ($2.00 each)  $____
Ttal    ____
IBlD.0053                                                          Thank you for your order.



Distributors of World                      Attn: Mr. Giancarlo Bigazzi                Attn: Mr, IJM. Hernandez
Via LamarmAora 45                         Castello 37
Bank Publications                          *0121                                      Madrid
ARGENTINA                                  Florence                                   SRI LANKA AND THE MALDIVES
Carlos Hirsch, SRL,                        JAPAN                                      Lake House Bookshop
Attn: Ms. Monica Bustos                    Eastern Bock Service                       Attn: Mr. Victor Walatara
Florida 165 4° piso                        Attn: Mr. Terumasa Hirano                  41 Wad Ramanavake Mawatha
Galeria Guemes                             37-3, Hongo 3-Chome, Bunkyo-ku 113         Colombo 2
Buenos Aires 1307                          Tokyo                                      SWEDEN
AUSTRALIA, PAPUA NEW GUINEA,               KENYA                                      ABCE Fritzes Kungi. Hovbokhandel
FIJI SOLMONISLANDS,                    Africa Book Services (E.A.) Ltd.           Attn: Mr. Eide Segerback
FIJI, SOLOMON ISLANDS,                     Attn: Mr. M.B. Dar                         Regerinsgatan 12, Box 16356
VANWE ATE                                  P.O. Box 45245                             S-103 2/ Stockholm
The Australian Financial Review            Nairobi                                    SWITZERLAND
Information Service (AFRIS)                KOREA, REPUBLIC OF                         Librairie Pavot
Attn: Mr. David Jamieson                   Pan Korea Book Corporation                 Attn: Mr. Henri de Perrot
235-243 Jones Street                       Attn: Mr. Yoon-Sun Kim                     6, rue Grenus
Broadwav                                   P.O. Box 101, Kwanghwamun                  1211 Geneva
Svdnev, NSW 20001                          Seoul                                      TANZANIA
BELGIUM                                    MALAYSIA                                   Oxford University Press
Publications des Nations Unies             Universitv ol Malaya Cooperative          Attn: Mr. Anthonv Theobold
Attn: Mr. Jean de Lannov                   Bookshop Ltd.                              Maktaba Road, P.O. Box 5299
av. du Roi 202                             Attn: Mr. Mohammed Fahim Htj               Dar es Salaam
1060 Brussels                              Yacob                                      THAILAND
P.O. Box 1127, jalan Pantai Baru           Central Department Store, Head Office
CANADA                                     Kuala Lurrpur                              Attn: Mrs. Ratana
Attn: Mrs. Suzanne Vermette                MEXICO                                     306 Silom Road
C.P. 85, Boucherville J4B 3E6              INFOTEC                                    Bangkok
Quebec                                     Attn: Mr Jorge Cepeda                      Thailand Management Association
San Lorenzo 153-11, Col. del Valle,        Attn: Mrs. Sunan
COSTA RICA                                 Delee. Benitc Juarez                       308 Silom Road
Librena Trejos                            03100 Mexico D.F.                           Bangkok
Attn: Mr. Hugo Chamberlain                                                               g
Calle 11-13, Av. Fernandez Guell           MIDDLE EAST                                TUNISIA
San Jose                                   Middle East Marketing Research             Societe Tunisienne de Diffusion
Bureau                                     Attn: Mr. Slaheddine Ben Hamida
DENMARK                                    Attn: Mr. George Vassilou                  a Avenue de Carthage
5anfundslitteratur                         Mitsi Bldg. I                              Tunis
Attn: Mr. Wilfried Roloff                  Makkarios Al Avenue                        TuRE
Rosenderns Alle 11                         Nicosia                                    TURKEY
DK-1970 Copenhagen V.                      Cyprus                                     Haset Kitarevi AF S
EGYPT, Arab Republic of                    (Branch offices in Bahrain, Greece,        Attn: Mr. Izzet Izerel
Al Ahram                                   Morocco, Kuwait, United Arab               469, Istiklai Caddesi
Al Galaa Street                            Emirates. Jorclan)                         3evog!u-Istanbul
Cairo                                      NETHERLANDS                                UNITED KINGDOM AND
MBE BV                                     NORTHERN IRELAND
FINLAND                                    Attn: Mr. Gerhard van Bussell              Microinfo Ltd.
Akateeminen Kirjakauppa                    Noorderwal 3S,                             Attn: Mr. Rov Selwyn
Attee Mri Kari Litmanen                    7241 BL Lochem                             Newman Lane, P.O. Box 3
Keskuskatu 1, SF-00100                     NORWAY                                     Alton, Hampshire GU34 2PG
Helsinki 10                               Johan Grundt Tanum A.S.                     England
FRANCE                                     Attn: Ms. Randi Mikkelborg                 UNITED STATES
World Bank Publications                    P.O. Box 1177 Sentrum                      The World Bank Book Store
66, avenue dleIna                          Oslo 1                                     600 19th Street, NW.
75116 Paris                                PANAMA                                     Washington. D.C. 20433
GERMANY, Federal Republic of               Ediciones Libreria Cultural l'anamena      (Postal address: P.O. Box 37525
LNO-Verlag                                 Attn: Mr. Luis Fernandez Fraguela R.       Washington, D.C. 20013, U.S.A.)
Attn: Mr. Joachim Krause                   Av. 7, Esyoana 16                          Baker and Tavior Company
Simrockstrasse 23                          Panama Zone 1                               01 South Gladiola Avenue
D-5300 Bonn I                              PHILIPPINES                                Momence, Illinois, 60954
HONG KONG, MACAU                           National Book Store                        380 Edison Way
Asia 2000 Ltd.                             Attn: Mrs. Socorro C. Ramos                Reno, Nevada, 89564
Attn: M2. Gretchen Wearing Smith           701 Rizal Averue                           50 Kirbv Avenue
6 Fl., 146 Prince Edward Road              Manila                                     Somerville, New Jersey, 08876
Kowloon                                    PORTUGAL                                   Commerce, Georgia 30599
INDIA                                      Livraria Portugal                           e
Attn: Mr D.P. Veer            ~Attn: Mr. Antonio Alves Martins             Bernan Associates
LCBS Publishers' Distributors Ltd.         RAtn Do. Antcno  Al-e4  Matns9730-E George Palmier Highway
Attn: Mr. D.P. Veer                        Ru1D200     o707                           Lanham, Marylanid, 20761
5 Ansari Road, Post Box 7015              100
New Delhi 110002                           I lbon                                     blackwell North America, Inc.
(Branch offices in Bombav, Bangalore,      SAUDI ARABIA                               1001 Fries Mill Road
Kanpur. Calcutta, and NAadrasi            Jarir Book Store                             lcwod   e   ere  81
INDONESIA                                  Attn: Mr. Akram Al-Agil                    Sidnev Kramer Books
Pt. Indira Limited                         P.O. Box 3196                              1722 H Street, N.W.
Attn: Mr. Bambang Wahvudi                  Rivadh                                     Washington, D.C. 20006
JI, Dr. Sam Ratulangi No. 37               SINGAPORE, TAIWAN, BURMA                   United Nations Bookshop
Jakarta Pusat                              Information Publications Private, Ltd.     United Nations Plaza
IRELAND                                    Attn: Ms. Janet David                      New York, N.Y. 10017
TDC Publishers                            02-06 lst Floor, Pei-Fu Industrial          VENEZUELA
Attn: Mr. James Booth                        Building                                 Libreria del Este
12 North Frederick Street                 24 New Industrial Road                      Attn. Mr. Juan Pericas
Dublin I                                  Singapore                                   Avda Francisco de Ntiranda, no. 52
ITALY                                     SPAIN                                       Edificio Galipan, Aptdo. 60.337
Licosa Commissionaria Sansoni SPA          Mundi-Prensa Libros, S.A.                  Caracas 1060-A



The World Bank
Headquarters                         European Office                Tokyo Office                           U
1818 H Street, N.W.                  66, avenue d'16na             Kokusai Building
Washington, D.C. 20433, U.S.A.       75116 Paris, France            1-1 Marunouchi 3-chome
Telephone: (202) 477-1234            Telephone: (1) 723-54.21       Chiyoda-ku, Tokyo 100, Japan
Telex: WUI 64145 WORLDBANK           Telex: 842-620628             Telephone: (03) 214-5001
RCA 248423 WORLDBK                                           Telex: 781-26838
Cable Address: INTBAFRAD
WASHINGTONDC
ISSN 0253-2115/ISBN 0-8213-9219-0