20310
Vol. 5
REPORT OF THE
RESEARCH ADVISORY PANEL
ON
INDUSTRIAL DEVELOPMENT AND TRADE
May 1, 1979
FILE COPY


Table of Contents
Page No.
Foreword
Chapter   I      Bank Research: A General Discussion                         1
A.     Why  and  for  whom? .................................
B.     Making the import, production, application and
dissemination of research more effective?......... 8
C.     Research implications of alternative development
strategies........................................  14
D.     Final  comments  ...................................  17
Chapter   II     Research on industry and trade: A_summary............... 19
A.     Review  of  past  research...........................  19
B.     Overall  evaluation.  ..........................   24
C.     Future research priorities........................ 28
Appendix  I      Individual field suggestions and evaluations
I.1    Incentive policies and economic integration....... 36
1.2    Comparative advantage, trade patterns and
economic  growth...................................  47
1.3    Export promotion policies in less developed
countries and access to markets in the developed
countries.........................................  55
1.4    Small enterprises, credit markets and public
enterprises.......................................  62
1.5    Capital utilization, capital-labor substitution, and
technological  change..............................  72
1.6    Programming in the manufacturing sector........... 77
Appendix  II     List of projects and papers reviewed..................... 87


- ii -
Foreword
We hereby submit our Report on World Bank research in the areas
of industrial development and trade. Though we are collectively responsible
for the entire report, the drafting of the various sections of Appendix I has
been undertaken by individual members of the group.
Edmar L. Bacha                Gerardo M. Bueno         Juergen B. Donges
Professor of Economics        Mexican Ambassador       Professor of Economics
Pontoficia Universidade       to the EEC               The Kiel Institute of
Catolica do Rio de Janeiro    Brussels                 World Economics
Jae-Ik Kim                    Assar Lindbeck           Richard R. Nelson
Director-General              (Chairman)               Professor of Economics
Bureau of Economic Planning   Professor of Inter-      Yale University
Seoul                         national Economics
University of Stockholm
Kirit Parikh
Professor of Economics
Indian Statistical Institute
New Delhi


CHAPTER I
BANK RESEARCH: A GENERAL DISCUSSION
The overall general quality of research in the field of industry
and trade in the World Bank is, in our view, very high--compared both to
university research and to research activities of non-university organiza-
tions, including organizations connected with the UN system. Thus, the
basic problem of research within the Bank in this field is usually not the
quality but rather the type of research produced by the Bank, in particular
when looking ahead, and the use of research within the Bank.
When addressing the issue of Bank research in this field, it is
important to remember that "research", as defined by the Bank, is only a
small fraction of general analytical work going on in the Bank, and that
research on industry and trade is only a small part of the total research
effort. Broadly speaking, analytical work of various kinds--including the
assessment of economic trends and policies in various countries, sectors and
markets--comprises approximately one sixth of the administrative budget of the
Bank. About one quarter of this analytical work seems to be formally classi-
fied as "research" , of which approximately one seventh, covering about 10-11
manyears of research annually, is on industry and trade.
It is useful to start the discussion in the report with some general
principles that may be applied when planning research activities within the
Bank. In particular, we shall take up the issues of the reasoas for Zank
research and the audience for the research. We shall thereafter make some
suggestions as to how the efficiency of various research activities of the
Bank - import, production, application and dissemination of research - may be
improved. In the final section of the chapter, the implications for Bank
research of alternative strategies for economic development are considered and
applied to research on industrialization and trade. Against the background of
the discussion in this chapter we shall, in Chapter II, give a general review
and evaluation of Bank research on industrialization and trade, as well as
some recommendations for future research priorities of the Bank in this field.
More detailed discussion on these issues are presented in Appendix I, where
the previous and present research of the Bank on industry and trade are
classified into six main areas.
A. Why and for whom?
An important point of departure when assessing Bank research is why
the Bank is, and should be, engaged in research, and who the audience for Bank
research is supposed to be. A hint on these issues is provided by a formula-
tion in our terms of reference according to which research objectives of the
Bank include the task "to support all aspects of Bank operations ..." and "to
broaden our understanding of the development process". According to these
statements, the audience for Bank research should be both Bank representatives


-2-
who are responsible for general policy issues and staff members who are en   .
in more narrowly defined Bank operations. It is obvious, however, that Bank
research may also be of great relevance for people outside the Bank.   In zac-
the Bank has regarded it as a duty to stimulate research in the less devel-
oped countries as well as to produce research knowledge for people outside
the Bank, including the community of scholars around the world in the field
of development economics. In addition, research may be stimulated in these
countries when the Bank fulfills its advisary role.
But before looking into the implications for the research policy z
the Bank of the need to support Bank operations, and to stimulate researcI
the less developed countries, it is interesting to explore the implications
for Bank research of its comparative advantage as a research unit, ignoring
for the moment the issue of for whom Bank research is supposed to be perforrI.-.
It would seem that this approach to the issue of research priorities 'would
follow from a research philosophy according to which the Bank, in the most
efficient way possible, tries to contribute to the accumulation of research
knowledge in the world as a whole. Thus, the Bank would be regarded as a
producer of "public goods" in the form of scholarly knowledge in the field of
development economics. The choice of research priorities of the Bank would
then be determined solely by the Bank's ability to produce research, and not
by its internal demand for (use of) research knowledge.
By applying the notion of comparative advantage we could conceive of
a ranking list of research areas in terms of falling relative advantage of the
Bank as a producer of research - a list to be cut off at the point where the
research budget of the Bank is used up. In other words, for what kinds of
research is the Bank, given its lending and policy advising activities, a
particularly good location when the Bank is looked upon solely as a producer
of research?
On the basis of this approach, the following aspects of Bank research
seem to be particularly relevant:
(a) The research of the Bank should concentrate on fields where
particularly competent researchers are available within the Bank, or (in a
longer perspective) can be hired by the Bank. Thus, the skills currently
available amongst Bank research staff should perhaps be regarded as a short-
term constraint rather than as a factor that determines the long-term com-
parative advantage of the Bank.
(b) Bank research should exploit the skills and information that
are acquired within the Bank in its operational activities as lender and
adviser.
(c) The Bank should concentrate on large projects and projects
where a continuity of research effort is important.
(d) The Bank should exploit its information and understanding
of facts and problems in a large number of countries to make comparative
studies of national economies.


-3-
(e) The Bank should try to provide statistical data and other
information, where such information is more readily (cheaply) available
to the Bank than to other organizations. (It is a somewhat controversial
issue whether this should be called "research".)
Aspect (a) would imply that the Bank continue with roughly the
same kind of research as it has successfully pursued so far (assuming that
leading researchers within the Bank are not likely to leave), but also that
it moves into areas that are suitable for highly competent researchers who
can be hired by the Bank. In some cases, of course, senior researchers of
the Bank may be willing to change their research fields. The choice of the
Bank's research topics will probably always reflect the interests and back-
ground of dominating researchers in the Bank. This has certainly been the
case in its previous research activities, as illustrated by the research on
trade policy incentives, with emphasis on effective protection and domestic
resource costs (the field of Bela Balassa), growth patterns and sources
of growth (the field of Hollis Chenery), and investment programming and the
range of technical choice (the fields of Ardy Stoutjesdijk and Larry Westphal).
In fact, a good forecast could have been made some years ago about research
priorities of the Bank revealed in the recent past simply by looking at the
background of the economists who had been hired by the Bank! But it is
equally true that the interests of researchers have evolved in response to
their experience in the Bank.
By contrast, the Bank has not been very successful in contracting
outside scholars and in doing research in circumstances where there has not
been strong leadership exerted by researchers within the Bank. It would seem
that projects dominated by outsiders have usually not succeeded as well as
those dominated by leading researchers permanently employed by the Bank.
Perhaps open competition when choosing outside researchers could, in some
cases, improve the quality of such research. In this connection we would also
recommend that the Bank avoids getting into a straightjacket by subjecting the
appointment of researchers to nationality criteria. The experience of other
international organizations points to the dangers of such practices.
Aspect (b) suggests research in fields where knowledge is acquired
(rather automatically) by the Bank in connection with its operational and
policy advisary activities. In fact, the Bank offers to researchers the
incentive to work in fields close to policy formulation and implementation,
which makes it a natural location for policy-oriented research. One example
is research on investment project evaluation. Studies of investment projects
that are relevant for several countries at the same time may be particularly
suitable for the Bank, as compared to other (usually national) organizations.
Another example of a field where knowledge useful for research activities is
acquired by Bank operations is probably the study of government policies and
regulations. A third, related example are studies of incentive systems and
the consequences of alternative institutional arrangements in markets for
products, services, credit and labor - and the consequences of these circum-
stances for rates and patterns of economic development.


Considerable research capacity has in fact already been built up
within the Bank in several of these areas. The research efforts on investmen.
programming and trade policy incentives have already been mentioned. However
a potential for research on institutional and policy-oriented problems has
also been acquired by way of various research efforts of the Bank in a large
number of countries with different institutional arrangement and policies* A
modest start in studies of this kind has also been made in sector reviews and
studies of small-scale enterprises, state enterprises and financial interme-
diaries. Ongoing analytical work within the Bank on trends and problems in
international markets and in national economies could also be a foundation for
more systematic research efforts within the Bank.
Aspect (c) would suggest a heavy emphasis on large projects and
research activities where updating and follow-up research is important.
Aspect (d) is a recommendation concerning the mode of carrying out
research rather than about a research area (or problem area). The fact that
both the research activities of the Bank and its operations refer to a great
number of different countries is a strong reason why the Bank should have a
comparative advantage in comparative studies.
Aspect (e), finally, might suggest increased efforts by the Bank to
collect, process and publish data that come out naturally from the operational
activities of the Bank. However, to make this activity a main task of the
Bank - i.e. to turn the Bank into a predominant data bank in the development
field - would require truly enormous resources. We know how great such a task
is even on a national scale. To do the same thing for about 100 countries
would therefore require a formidable effort. It may thus be reasonable to
limit demands on the Bank in this field to taking greater responsibility for
the data it collects for use in its own research and surveys, and to making
these data to a large extent available to outsiders. It may be noted that
this is not only an issue about publication of research results, but also
about improving and controlling the quality of data, which may very well be
regarded as a research task. However, resource constraints make it necessary,
we believe, to take a rather selective approach to the collecting, improving,
controlling, processing and publication of data.
A rather special reason why the World Bank may have a comparative
advantage relative to universities in research activities such as (b)-(d) is
that these pursuits, to be efficient, often require a rather strong and
permanent research organization, which the social science departments of
universities often lack.
A strict adherence to the principle of comparative advantage in
the production of research would imply that types of research knowledge that
are not effectively produced by the Bank - regardless of how important they
are for Bank operations - should be imported. It is obvious, however, that
such a comparative advantage approach is not a sufficient criterion upon
which to base Bank research strategies.


-5-
Firstly, the need (demand) for scholarly knowledge in Bank operations
cannot always be satisfied by importing research results (in fields where the
Bank has not a comparative advantage in the production of research knowledge).
One reason is the specificity of required knowledge, another is the lack of
interest outside the Bank in certain types of research knowledge that the Bank
needs (demands). Thus, in order to satisfy its own requirements for research
knowledge, the Bank certainly has to carry out research that is particularly
useful for Bank operational needs, and that is not done elsewhere. In other
words, the Bank has also to perform a role as residual supplier of research in
some fields.
What would be the concrete consequences of following this principle
rather than the principle of comparative advantage in production? In the
light of the roles of the Bank as lender, investor and policy adviser - and
considering the often highly distorted relative prices in many LDCs - the
heavy emphasis on studies by the Bank of effective protection and domestic
resource costs could certainly be defended on the basis of the principle of
residual supplier of research. In fact, it would seem that these studies have
been regarded as particularly useful by operations people, though some com-
plaint emerges to the effect that the techniques may have been somewhat
"overelaborated" for the purpose of Bank operations.
Another implication of the principle of residual supplier of research
is probably to justify research concerning large investment projects (in par-
ticular where the Bank is involved as a lender), projects with considerable
externalities (such as learning by doing), returns to scale and/or great link-
ages between sectors. The research within the Lank on investment planning
and programming is an example of this type of research for which an impor-
tant point of contact seems to be the Industrial Projects Department (IDP).
It would appear, however, that the operational staff of the Bank, particularly
in the regional offices, have not yet made a thorough assessment of the use-
fulness of this work; nor have they undertaken to manage its application in
an efficient way. In fact, it would seem that they usually do not believe
that research done in this field is very helpful.
However, we would argue that an adherence to the principle of the
residual supplier of research also motivates studies concerning problems in
which Bank research so far has not been strongly involved, but which in our
judgement reflect severe bottlenecks for economic development in many LDCs.
This brings us back to an earlier recommendation that was motivated by the
Comparative advantage principle, i.e. research on the consequences of (i)
government policies and regulations, (ii) institutional arrangements and
(iii) incentive regimes not only in product markets (where the Bank has
made considerable research efforts) but also in factor markets. The argu-
lent for more research efforts of this type would probably be particularly
strong should the Bank choose to concentrate research more than before on
the least developed countries and on the consequences of industrialization
for the least favored group within individual countries.
Thus, the residual supply of research approach and the comparative
advantage approach both seem to imply a strong emphasis on studies of the


- 6 -
consequences of alternative institutional arrangements, technologies, gover-,
ment policies and incentive regimes. The reason is that both approaches to
research are based on the assumption that research should be related to the
fields of Bank operations - either because the Bank acquires special compete,.
in such issues by way of its practical activities or because knowledge is
needed in these fields in the lending and advisory work of the Bank (and is
not easily imported). A rather special reason why the Bank may be a suitable
place for such types of research, relative to universities, is that research
on the importance of alternative institutional setups for economic develoD2en.
in the LDCs has so far not become a high priority field in the academic world.
There are many relevant types of research on the importance of institutions i-
the developed countries, such as in the field of industrial organizations;
labor economics, and money and banking. Important research on related issues
is also pursued in Law Schools (such as research on law and economics),
Business Schools (studies of organizations and regulations of enterprises) and
Political Science Departments (for instance studies of bureaucracies and
budgetary processes). However, in spite of their relevance for economic
policy making, these contributions so far have not gained a very high status
in the world of academic economics. Moreover, they have only to a very small
extent been applied to LDCs. As it may be a long time before they make their
way towards application in the LDCs, we believe that the World Bank could help
speed up the process by taking a leading position in this research field.
We shall later on (Chapter II and Appendix I) discuss in some detail
both why more research of these types is important and how it may be conducted.
A second reason why Bank research cannot be ba sed only on the prin-
ciple of comparative advantage in the production of research is that there
are important "externalities" of having researchers within an organization.
For instance, research done within the Bank may be more visible to the opera-
tions people than research done elsewhere in the same field. Also, operations
people can be made aware of what researchers outside the organization are
doing by having a number of in-house researchers (a point to be developed in
the next section). More generally, the presence of researchers in an c-rganiza-
tion such as the Bank helps the development and propagation of new ideas that
ultimately affect operations. This important role should be borne in mind in
allocating nonresearch time to researchers.
It may also be argued that the general level of sophistication of
an organization, for instance in the field of policy advising and economic
surveying, is influenced by the general quality of the researchers vithin
the organization. Good researchers help to set performance standards within
an organization, which is particularly important in one that, like the Bank,
performs the role of policy adviser. Moreover, the reputation of the Bank as
an institution of high competence may be boosted by a high research capacity.
Such a reputation may help the Bank to hire talented people in general, and
hence improve the Bank's operational efficiency.


-7-
In other words, the research of the Bank should not only be subor-
dinate to the (relative) efficiency of the Bank as a producer of the research,
or to the immediate demands of the operational units of the Bank, but also
to long-run considerations of the general competence of the Bank in various
respects.  In particular, it is important to stimulate the ability of the Bank
to translate policy objectives into instruments and actions. An application
of this principle suggests that the Bank should build up research competence
in many fields where it operates - for the purpose of raising the general
level of sophistication within the Bank.
Thirdly, the Bank should also be interested in the externalities of
its research on the research potential of the LDCs. The adherence to this
principle is an argument for (i) choosing fields of research of interest to
scholars and research institutions in LDCs, and (ii) adopting research proce-
dures that facilitate the participation of such scholars and institutions in
the research program of the Bank. However, there is a risk that when following
these procedures the Bank will not always get the most competent researcher
for a specific project. Moreover, the Bank may also be criticized for distort-
ing research in the LDCs ("research imperialism"). To maximize the possibility
of LDC researchers' participation while minimizing its associated risks, some
of the Panel members are attracted to the idea of instituting a second tier of
research sponsorship by the Bank. Rather than first preparing detailed
research projects and then seeking LDC consultants to help implement the
research in the first stage, the Bank researchers would design only broad
terms of reference. These would then be submitted to open international
competition, including some preference schemes favoring the participation of
LDC scholars and research institutions. These schemes naturally must take
into account that the quality of the research is an overriding aim which
should not be jeopardized.
Thus, our discussion suggests four different principles for the
choice of research topics in the Bank in the field of industry and trade:
(a) a comparative advantage approach;
(b) an ambition to function as a residual supplier of research;
(c) an attempt to create externalities within the Bank in the
form of "sophistication" among Bank staff; and
(d) an ambition to help generate research knowledge and research
capacity in the LDCs.
In reality, it would appear that all these four principles for
research do prevail within the Bank. For instance, past Bank research has
certainly been addressed both to the community of scholars outside the Bank
and to staff members who are responsible for policy advising and lending
Within the Bank - in proportions that we cannot really pin down. Controversies
Over research strategies within the Bank probably often stem just from the
different weights that various staff members put on these various principles.


Disappointments over the research activities of the Bank are bound to be felt
by those who evaluate the research efforts of the Bank on the basis of one o
these principles only.
It would seem that the management of the Bank should clarify, to
itself as well as to others, which of these principles (motives), or possibly
others, should be emphasized. A similar attempt has, of course, been made
by us.
Also the procedure of the research within the Bank should to some
extent be influenced by the principles (motives) that govern Bank research.
If the Bank is simply trying to make the best possible contribution to the
research knowledge of the world, a rather concentrated research portfolio is
suggested. It is then also important to allow a very broad freedom for the
researchers to choose topics themselves, and hence "to do their own thing".
Moreover, the more successful the Bank is in hiring competent scholars, the
smaller the need for strictly formal organization and bureaucratic administra-
tion of research within the Bank.
On the other hand, if the Bank follows the residual supply of
research strategy, it is necessary to see to it, by way of organizing and
monitoring research, that the research becomes relevant to the operations in
the Bank. A more dispersed research portfolio would then probably follow than
by adhering strictly to the principle of comparative advantage in the produc-
tion of research.
If instead a heavy emphasis is put on the idea of creating high
sophistication among the Bank staff in general, an even more dispersed research
portfolio would probably follow, perhaps with some risk of not achieving the
critical mass of resources that is necessary for a breakthrough on the interna-
tional research frontier. Thus, this research principle comes into some
conflict with the others, in particular with the principle of comparative
advantage, because of returns to scale in research.
Finally, if a high priority is given to the ambition to improve the
research capacity in the LDCs, participation of researchers from these coun-
tries becomes a crucial criterion in the design of research projects. Research
would then often have to be organized as joint ventures, with a rather concen-
trated Bank research portfolio to assure reasonable efficiency.
B. Making the import, production, application and dissemination
of research more effective
It may be useful, when discussing the role of research within the
Bank, to make a distinction between import, production, application and
dissemination of research. It is important that all of these research-related
activities are pursued in an efficient way. How that may be achieved is the
topic of this section.


9
Import of research knowledge
The issue of the appropriate use of research within the Bank refers
to research knowledge in general rather than only to research knowledge that
is produced by the Bank itself. The bulk of research knowledge that is poten-
tiallY useful within the Bank will always be produced outside it. We there-
fore suggest that the Bank strengthen its capability to import research knowl-
edge in a systematic way. One of the most efficient ways of doing this is
probably to place people with a research background in operational positions
in various units within the Bank, so that research knowledge can be imported
not only via researchers of the Bank, but also "directly" by the operations
staff. This can be achieved in several different: ways: people outside the
Bank with a research background may be hired to operational positions in the
Bank; researchers of the Bank may move over permanently to operational posi-
tions; and researchers of the Bank may take operational positions temporarily
(for instance one or a few years). Of course, operational ability is a pre-
requisite for such appointments. In fact, the biggest import of research
knowledge probably occurs when someone with a background as researcher is
hired by the Bank as researcher or operations officer; knowledge is often most
effectively imported "in the heads" of people at the time of their employment
rather than by reading research documents or doing research later on.
If the Bank is anxious that its own research is highly relevant for
its operations, i.e. if strong emphasis is put on the principle of the residual
supply of research, it is important to rely heavily on permanent staff members.
Moreover, when relying on outside consultants it may be a good idea not only
to use the most outstanding consultants who are available, but also to build
up a network of more or less permanently Bank-affiliated outside researchers.
These could then acquire some knowledge about the use, and usefulness, of
research within the Bank. It also becomes important to choose consultants who
are able to communicate with Bank staff members. In order to avoid inbreeding
and one-sideness of Bank research this network of consultants should include
scholars with different philosophical outlooks, skills and methodological
preferences.
Another obvious way of importing research is to commission people -
inside or outside the Bank - to prepare state of the arts papers, an issue to
be discussed in connection with the problem of dissemination of research (pp.
15-19).
Production of research
The most important aspect of the production of good research is,
of course, to have highly competent researchers both within the Bank and as
consultants. If the Bank, as we suggest, shifts its emphasis somewhat to new
fields of research, it is crucial that the Bank hires outstanding specialists
in fields relevant to such research. What this means in concrete terms has
to be considered carefully by the Bank. Our view is that the best research
is usually done by scholars with a strong theoretical and methodological back-
ground in a broad field of research. Suppose for instance that the Bank is
going to study problems of incentives and imperfections in factor markets,


- 10                                     1
or alternative institutions and government policies - and the consequences
of these rates and patterns of growth. What is then required is not mainly
narrowly defined specialists on policies and institutions in specific markets
and countries, but rather outstanding scholars in fields such as industrial
organization (including the issue of competition and entry), technical devel-
opment, financial intermediation, public expenditures and taxation, and labor
economics. If the Bank, as we believe, should analyze success and failure
stories of various countries, there is also perhaps a case for hiring some
scholars with high competence in the analysis of economic systems and modern
economic history. Again, there are good reasons for emphasizing the impor-
tance of using scholars with some dispersion of philosophical outlook on the
issue of economic development.
An important question in these fields is under what general condi-
tions (policies, institutions, systems of incentives and perhaps also cul-
tural characteristics such as the work ethic), entrepreneurship, techrlological
advance and innovations are likely to flourish. It is possible that research
in these new fields would sometimes benefit from the application of rather
interdisciplinary knowledge and methodology.
However, it is well known that interdisciplinary studies are extreme'.
difficult to pursue successfully. For some research questions something can
be gained merely by adding the understanding of a sociologist or a political
scientist to that of an economist, but for most interesting questions not much
is gained by such a simple addition of vie-wDoints. To achieve a truly inte-
grated approach to a problem requires a much more intricate intertwining of
ideas and knowledge. For a group of interacting scholars to attain the
capability of doing research together in an integrated way requires a lot of
time learning each others perceptions, and learning how to pose and answer
questions in a way that transcends the disciplinary boxes. This is not an
endeavor to be initiated with expectations of a quick payoff, and the fact
that there are so few examples of successful interdisciplinary research
suggests that even with patience the endeavor is highly risky. It is also
noteworthy that several of the more successful efforts involved integration
from different fields in the head of an individual scholar rather than through
the joint thinking of a group of scholars. The important interdisciplinary
questions raised above call for an effort at the Bank to broaden the analysis
to go beyond conventional economics. But such a broadening should be entered
upon cautiously and with patience. Much more is required than merely putting
together an interdisciplinary research team.
Application of research
In some measure, Bank research has been applied in operations, with
the modality of application depending upon the particular type of research.
However, at present the in-house demand for application of Bank research
through case studies seems to far exceed the supply. This is partly a reflec-
tion of the Bank only recently having established a research program. Most
effort to date has been devoted to doing research. It is only as research
results have been forthcoming that the demand for application has materialized
in concrete form. Consequently, the time is at hand when the Bank should


O11-
begin seriously to consider how its research should be applied to problems
identified by operational staff.
The formulation of a problem so that it can be addressed by research
requires the joint effort of operational staff, familiar with the specific
circumstances in which the problem arises, and of researchers, familiar with
comparative merits and the feasibility of alternative approaches. In some
cases, the actual case study could be carried out equally well by operational
as by research staff, provided that adequate assistance is provided by the
other.  But there will also be cases in which the specialized expertise
required to conduct the case study implies that it must be undertaken by a
researcher.
It is questionable whether Bank staff should typically be expected
to carry out the case studies that replicate past research in an operational
setting.  It may be preferable to employ researchers from the country in
question to carry out the analysis, which would simultaneously increase
awareness of the issues in the particular country and further develop its
research capacity. Bank staff nonetheless should undertake some case studies,
in order to maintain their competence and further refine the approaches, which
may often result in simplification and shortcut procedures to further reduce
the cost of such undertakings.
Good argument.s can be made for decentralizing the research staff
involved primarily in application, particularly those concerned with incentive
policies and industrialization strategies. By being decentralized, staff
would gain greater familiarity with the particular circumstances of the
individual countries within the different regions. There is thus a strong
case for the establishment of small research units within the regional offices.
A different organization may be called for where operations remain centralized,
as in the case of large-scale investment projects. Here it is probbly best to
locate an applications unit within the department concerned.
In either case, the decentralized units would function as vehicles
for application of research methods that have been developed within the
Bank.
Dissemination of research knowledge
Our next issue is how to improve the dissemination and assimilation
of research knowledge within the Bank - imported as well as produced. When
discussing this problem, it is important not to take too mechanical a view
of the issue. The task is not mainly to bring over some specific tools and
actual information to operational staff and policy advisers, but rather to
spread around a certain way of looking at things.
Good interaction between researchers and operational staff is not
easy to achieve. In fact, our interviews among people in the Bank revealed a
certain tension between the two.  This is nothing peculiar to the World Bank,
however. Researchers within an organization that deals with practical affairs
almost inevitably will be viewed by operational staff as being somewhat distant


- 12 -
and academic - a kind of research enclave. But it is important that research
maintain a certain distance from the pressures bearing on the operational
staff. Research is a highly specialized time-consuming job, which has to
be protected to a large extent from demands of practical and administrative
duties. And research needs to be organized in such a way that there is a
sense of community of scholarship. If an organization like the Bank wants
to acquire and keep competent researchers, it is necessary to let them do
their own thing to a large extent, without too many disturbances from other
activities within the organization.
Another reason for tension between researchers and operational stafz
is that the former are usually concerned with a much longer time perspective
than the latter. The production period of research is, moreover, often so
long that when results do emerge, operational staff may have lost interest in
the question. And sometimes the empirical data used in research projects may
no longer reflect existing conditions. (To some extent the studies of effec-
tive protection and domestic resource costs have suffered from this dilemma.)
Moreover, whereas researchers are usually interested in the accumula-
tion of generalized knowledge, operational staff are often more interested in
drawing on knowledge, in particular in obtaining knowledge that pertains to a
particular situation. The researchers often regards the search for the latter
type of knowledge as "information gathering" rather than research. Operational
staff, by contrast, are frequently disappointed by research results because
these do not always give concrete, easily accessible, ready-made and unambig-
uous conclusions about immediately relevant operational and policy issues.
Besides, operational staff are often not aware of, interested in,
or able to absorb results or paradigms supplied by the researchers. In fact,
usually they cannot possibly know in advance the type of research that could
help them in their work.
On the other hand, every one of the criteria discussed above signals
that research at the Bank should be guided by a thorough understanding of what
operational staff know and need to know. Good two-way communication is neces-
sary if research at the Bank is to be fruitful. It is our impression that at
present communications are not as good as they should be. It would be fruitful
if researchers were better informed about the usefulness of research knowledge
to operational staff, and if researchers could communicate more effectively
with them. Obviously it is a delicate business balancing the two organiza-
tional requirements for successful Bank research - a considerable degree of
shielding from short-run interests and pressures on one hand and strong inter-
action between researchers and operations people on the other. We have to be
satisfied with compromises between these conflicting objectives. Keeping this
in mind, several reforms suggest themselves:
(a) That the researchers write, and circulate within the Bank,
popularly written reports on research--concerning research produced both
inside and outside the Bank. To the extent that the reports summarize
outside research, the processes of import and dissemination are of course
combined. In some cases, it may be a good idea to ask outside consultants


-13-
rather than researchers inside the Bank, to make the surveys. However, when
outside research results and paradigms are summarized, it is probably impor-
tant not only to summarize outstanding academic contributions but also
to try to find out what types of research have been successfully used in other
operating organizations.
(b) That joint seminars are organized by researchers and opera-
tional staff--preferably at some distance from Washington (with disconnected
telephones!) to make undisturbed discussions possible.
(c) That more circulation of, people between research and operational
activities is brought about. Sabbatical leave for research, within or outside
the Bank, for the operational staff may be one method of achieving this. Such
circulation may be difficult to achieve in the field of methodological and
highly technical (model-oriented) research, where the rate of skill deprecia-
tion is often very high. However, in more applied fields--where experience,
empirical knowledge and common sense are important--circulation may be both
possible and highly useful.
(d) That more systematic attempts are made among the researchers of
the Bank to try to understand what applied people need to know in their work.
It is not easy to say what is the most efficient way of achieving this. One
possibility would be to undertake joint ventures between researchers and oper-
ational staff - certainly in operational activities, but sometimes perhaps
also in the design and to some extent in the execution of research. It is
likely that dissemination of methodological knowledge is more efficiently
achieved if researchers and operational staff jointly apply suggested method-
ologies to concrete issues in the operational deparLments--in studies of proj-
ects, sectors, markets or countries. Perhaps it would also be possible to
induce operational staff to make more research suggestions. More informal--
i.e. less bureaucratic--procedures when drafting and planning new research
projects might increase the possibilities of operational staff to contribute
to the initiation of and participation in research.
Apart from individual research projects, cooperation between
researchers and operational staff can contribute to the development of a
research program in the area of industry and trade. A beginning in this
direction has been made through the establishment of the internal panel on
research on industry and trade.
The suggested research units within the operational offices
(see the section on application of research) probably also could help the
dissemination and assimilation of research results among the operational
Staff.
Moreover, the earlier suggested employment of people with research
background in operational positions would not only facilitate the import of
research knowledge to the Bank; it would also be a way of disseminating and
assimilating research knowledge within the Bank from researchers to opera-
tional staff. This is potentially important, as imported research may be
more difficult to disseminate and assimilate than in-house research. Both


- 14 -
the suggestion to let researchers circulate between research and operational
positions, and the suggestion to recruit (more permanently) people with a
research background to operational positions means that some bridges would
be built between research and operations. Thus, we suggest in fact that the
Bank tries more systematically to build up a staff of "bridge people" in the
operational departments.
It is important to realize that the limits to using more research
knowledge within the Bank are probably determined more by the absorptive
capacity of research among the operational staff--limited time as well as
limited ability and interest to absorb research knowledge--than by the
capacity of researchers within the Bank to produce and summarize research.
This means that a larger voluie of research within the Bank should perhaps
not be expected to have much effect on the operational side of the Bank, as
long as the deficiencies of the systems of dissemination and assimilation of
research within the Bank have not been removed.
The role of researchers
Bank researchers have separate roles to play in the import, produc-
tion, application, and dissemination of research. It is very easy, in an
operational institution, for the production of research to be sacrificed to
the objective of better utilizing existing research. This is particularly
true in an institution like the Bank, where significant efforts to apply
research are just beginning to be made. Continued production of high quality,
innovative research within the Bank will require a strong commitment to
protecting the time of research staff for the production of research. This
should of course not be interpreted as an endorsement on our part of an
"ivory-tower" attitude among researchers. This said, it must be considered
whether the time spent by researchers in their non-research roles is being put
to its potentially most effective use. Direct involvement in operational
missions by research staff consumes much of their non-research time, with the
result that they are not as available for consultation in regard to application
as might be desirable. We equally suspect that dissemination could be much
better organized were more resources devoted to it.
C. Research implications of alternative development strategies
A major task of our report is to discuss future research priorities
of the Bank in the field of industry and trade. An important background to
such a discussion is (i) a specification of what types of countries we are
talking about and (ii) some kind of "vision" of what the mechanisms and driv-
ing forces of economic development are supposed to be in these countries.
For instance, if we talk about countries with a strong emphasis
on central planning, nationwide planning models and empirical studies of
the process of central planning in various countries would probably be top
priorities. It is important to remember that there are substantial ele-
ments of decentralization of information, decision making and initiative


- 15 -
also in centrally planned economies. Thus, it is of great interest to study
incentives and constraints on behavior at various levels in centrally planned
systems.
The reason why the Bank has not put any great amount of research
resources into the acquisition of research knowledge in this field is most
likely that very few member countries of the Bank are centrally planned
economies. Elements of planning exist everywhere, however, because of the
great role of government decision making in economic matters in all countries
today. This means, of course, that the process of government decision making
is an important area of research for all countries.
(The macro models that have been developed by the Bank for entire
economies, or even for the whole world, should probably not be regarded as
tools of central planning, but rather as descriptive models or forecasting
models, and to some extent also tools of analyzing the effects of alternative
policies.)
A more modes: version of planning would be sectoral planning or
programming of investment decisions, for instance, in sectors where there
are huge returns to scale, externalities, or (direct) intersectoral linkages.
Then it may also be possible to consider aspects that are usually not well
caught, if at all, in conventional static microeconomic investment calcula-
tions. With this approach, studies of investment planning in some sectors
would be of rather high priority. It has been the case for the Bank in recent
years (see Appendix I.1), in the sense of investment programming - mainly
in cases where the optimum size of a firm is of about the same magnitude (or
larger) than the entire national market. Another reason why sector planning
models, or at least investment programming models, may be of interest is that
in many countries, perhaps particularly in less developed countries, a number
of infrastructure and processing industries, for which investment programming
models may be particularly useful, are in fact under rather detailed central
government control.
It is important to emphasize that microeconomic investment planning
models of this type, which are really tools of management decision making, are
quite consistent with either a market-oriented or a centrally-planned macro-
economic system. In reality the Bank's research in this field has not been
framed in the context of nationwide central planning models, but rather as a
means of exploring the range of choice of firms operating on markets. The
same holds for Bank research on capacity utilization, capital--labor substi-
tution and technological change (see Appendix I.5)--a research field of great
interest in the case of both centralized and decentralized versions of the
development process.
Research knowledge of patterns of growth of production and trade
(see Appendix 1.2) is also of considerable importance--especially perhaps for
acquiring a broad understanding of the development process--in the case of
both rather centralized and more decentralized strategies of economic devel-
Opment. In particular, Bank research in this field has helped provide norms
of growth patterns of industry and trade, against which developments in in-
dividual countries can be judged.


- 16 -
In most of the member countries of the Bank, the bulk of the devel-
opment process is no doubt guided by decisions of decentralized units moti-
vated by profits and stimulated and constrained by markets. However, it is
important to realize that the adherence to a pronounced decentralized strat-
egy of economic development, in the context of a market economy, does not
imply the absence of central policies and planning, but rather more reliance
on different types of policy and planning operations from those in centrally
:lanned economies. Obvious examples, besides general monetary and budgetary
policies, are institutional reforms and improvements in the systems of incen-
tives. In countries of this type it would therefore be of interest to find
out how conducive alternative institutional arrangements and incentive systems
are to releasing efficient decentralized initiatives.
While research on incentives in product markets, in particular trade
policy incentives for firms, has been given a high priority in Bank research
(see Appendix I.1), it is only recently that research has been launched on
incentives and imperfections in factor markets and incentives for employees
(households). By this we mean for instance the structure of interest rates,
the performance of credit rationing and the mobility, and flexibility in
aeneral of the credit and capital markets, but also the structure of wage
rates, the incentives and possibilities for labor to move, acquire skills,
and advance.
Nor has there been much emphasis in Bank research on the importance
for economic development of institutional arrangements (see Appendix 1.4),
though an increased interest in institutional factors can be detected in
various research efforts of the Bank in recent years.
The only aspect (dimension) of development strategies mentioned
so far has been types and degrees of centralization of economic decision
making, which concerns the mode of economic organization. Development strat-
egies also differ with respect to resource allocation strategies, such as
the choice between export-oriented (outward-looking) and import subsitution-
oriented (inward-looking) strategies of economic development--a topic high-
lighted by Bank research on trade policy incentives and patterns of growth.
Another important dichotomy is between strategies that rely on
the assumption that incomes and employment opportunities will rather auto-
matically "trickle down" to poor groups of the population, and strategies
that more actively promote employment opportunities and income redistribu-
tion at an early stage of economic development, which according to experi-
ence (for instance, South Korea and Taiwan) is not inconsistent with an
export-oriented and market-oriented strategy of economic development.
A third dichotomy concerning resource allocation, finally, is
between countries that rely on a rather passive attitude towards what par-
ticular consumer goods are supplied to the domestic population, and strat-
egies that rely more actively on the provision of some basic needs of food,
shelter, health, etc.


- 17 -
In reality, the economic systems of the LDCs that are members of
,he Torld Bank are characterized by various combinations of centralized and
eacentralized decision making, and with different strategies for employment
creation, redistribution and the provision of basic needs. The development
Strategies are also heavily dependent on a number of other features of in-
,;vidual countries, such as the initial condition of the stocks of physical
a hLnan capital, natural resources and historical traditions. Bank research
ould reflect this diversity concerning possibilities and strategies of eco-
comic development. So, of course, should our recommendations.
D. Final Comment
The purpose of this chapter has been to formulate certain general
rinciples when pursuing and using research within the Bank, in particular
as it relates to industrialization and trade. Two of the principles dis-
zussed--the ones concerning comparative advantage and residual supply of
research--suggest strong emphasis on studies of the consequences for eco-
:omic development of alternative institutional arrangements, technologies,
government policies and incentive regimes. The reason is that both approaches
are based on the assumption that research should be related to the fields of
3ank operations, either because the Bank acquires special competence in such
issues through its operations, or because knowledge is needed in these fields
or successful Bank operations (and is not easy to import). Though there are
good reasons for emphasizing empirical and applied research for an organiza-
tion like the Bank, there is also a need for theoretical and methodological
studies. One reason is that the competence of the Bank as a research unit is
:hereby enhanced, another, that the Bank can play a major role in advancing
the frontier in these fields. The third principle considered--to create
externalities within the Bank in the sense of high general competence--sug-
gests a much more diversified research portfolio, with various types of re-
search that contribute to broad knowledge about the development process.
Finally, the fourth general principle--to increase the research competence
in the LDCs--suggests that the fields of research chosen are those of in-
terest to scholars and research institutions in the LDCs, and that research
procedures are adopted that facilitate the participation of such scholars
and institutions in the research program of the Bank.
We have also discussed various ways of improving the ability of the
Bank to make import, production, application and dissemination of research
Zore efficient. In particular, we have emphasized the potential usefulness
of increasing the awareness of both researchers and operational staff in the
Bank of each other's interests, needs, and competence. We have also suggested
a number of arrangements to help achieve this.
In general, both researchers and operational staff are quite aware
not only of the potential importance of Bank research but also of the diffi-
culties involved in choosing efficient research procedures and useful research
fields. We would in fact be inclined to argue that the Bank worries too much


-18-
about its research, in the sense that perhaps too much time and energy are
devoted to planning and reviewing the research activities of the Bank. A
slightly sanguine, but perhaps useful, recommendation would be: "Worry less
about your research, do it instead!"   Or more concretely: discuss carefully
the general direction of research efforts, i.e. the research fields, employ
the best senior and junior scholars you can get in these fields, and then
give them considerable autonomy to do their work--in contact with operational
staff but without too frequent demands as to the reporting and evaluation of
their work.


- 19 -
CHAPTER II
PAST AND FUTURE RESEARCH ACTIVITIES
ON INDUSTRY AND TRADE: A SUTMARY
A. Review of Past Research
Over the past years the World Bank has dedicated a considerable
volume of resources to research on industry and trade in economic development.
This research has explored a wide range of topics, a number of these in con-
siderable depth. A list of projects and papers is given in Appendix II. For
convenience, the Panel divided up past research into six broad clusters.
These are briefly described below. The six sections of Appendix I provide
more detailed descriptions of the research in each of the areas. Research
reviewed included projects approved by the Bank's Research Committee (so-
called RPO projects) as well as other research output financed by the Bank.
Incentive Policies and Economic Integration
One cluster of Bank projects has been concerned with incentive
regimes and development strategies of LDCs; we also placed in this group
studies on economic integration among developing countries. Included here
are RPOs 670-01 (Development Strategies in Semi-Industrial Countries), 670-22
(Economies of Scale and Tariff Levels), 670-87 (Industrial Policies and Eco-
nomic Integration in West Africa), 671-10 (Promotion of Non-Traditional
Exports) and, now under way, 671-75 (International Trade Policy for the
Development of Bangladesh), and several other (non-RPO) projects as well.
Most of these projects deal with the role of incentive systems in
economic development, from both a theoretical and an empirical standpoint,
and explore the resource-allocational, growth, employment and balance-of-
payments effects of various government policies, particularly of those aim-
ing at import protection, export promotion, and economic integration. Con-
siderable use has been made of the concepts of effective protection and
domestic resource costs, appropriately improved in theoretical and compu-
tational terms. While the incentive structures have been analyzed in a
number of different countries on a comparative basis, there have also been
attempts to evaluate incentives on a firm-by-firm basis and to appraise the
impact of protection at the level of the individual investment project.
With the exception of the small study on economies of scale and
tariff levels, the projects in this cluster were completed successfully or
are in good progress. A definitive evaluation of the Bangladesh study is
not possible at this stage; completion is scheduled for end-1979. Its rele-
vance, however, is beyond any doubt, particularly so as it is a case of appli-
cation of the more basic research on incentive regimes.


- 20 -
The major findings of the research referred to in this cluster,
especially those of RPOs 670-01 and 670-87, are of considerable interest for
policy-making purposes as they support the efficacy of promoting rather than
protecting industries, and of avoiding discrimination against exports rather
than overemphasizing import substitution, if efficient and rapid economic
growth in LDCs is to be achieved. The conclusions and policy recommenda-
tions rest upon firm theoretical foundations and a sound factual basis.
Moreover, they are timely, because reliance on and use of import controls
by developing countries, combined with an array of additional government
interventions in the domestic markets, still persist and guidelines for pol-
icy reforms are therefore needful. Finally, this research has been found
quite helpful by the Bank's operational staff, who are applying both the
findings and the methodology in country economic reports, policy analyses
and in-house estimates of incentives, let alone its initiatives for addi-
tional research in this field.
Comparative Advantage, Trade Patterns and Economic Growth
This cluster of projects includes RPOs 670-07 (International Model),
670-19 (Expansion in Manufacturing for Exports in Developing Countries), 670-79
(Economic Development of East and Southeast Asia), 671-05 (Patterns of Indus-
trial Development), 671-32 (A Comparative Study of the Sources of Industrial
Growth and Structural Change), and 671-79 (Sources of Growth and Productivity
Change), and two non-RPO studies as well.
The unifying theme is the objective of explaining the pattern of
resource allocation within and between countries, economic growth, and changes
in industrial and trade structures as a function of various country character-
istics, including policies employed. While two projects (670-19 and 670-79)
were clear failures, the other ones met, by and large, the research objectives.
The failures are regrettable, indeed, since the projects addressed the ques-
tions of how to shift manufacturing activities from developed to developing
countries and of how to plan competitive export industries in the developing
countries; information on these matters would have been an extremely important
ingredient of any effort to shape rational industrialization strategies.
The projects included in this cluster differ greatly in their
methodologies. Some of them involve modeling that is very simple or indeed
primitive; others involve attempts to empirically implement a very complex
general equilibrium methodology. In the view of the Panel the most successful
of the projects is RPO 671-32 that employed and elaborated an accounting frame-
work based on sector-specific supply-demand identities for analyzing the nature
of modern industrial growth. The research guided by the more ambitious general
equilibrium conception (RPO 670-07) has not yet added much to existing knowl-
edge about the development process. Moreover, skepticism about the useful-
ness of this type of research predominates within the operational staff. It
still remains to be seen whether the computable general equilibrium model
applied in the "Sources II" project, RPO 671-79, which has only just started,
will produce significant empirical results to an extent which could not have
been obtained at lower cost from simpler macroeconomic formulations.


- 21 -
Export Promotion Policies in the less developed countries and Access
to Markets in the developed countries
Included here are RPO projects 670-20 (Industrialization and Trade
policies for the 1970's), 670-21 (Export Promotion and Preferences: A Case
Study of India), 671-35 (Export Incentives in Developing Countries), 671-56
(Marketing Manufactured Exports), 671-66 and 671-67 (Effects of Increased
Imports of Manufactured Goods from Developing Countries in Western Europe
and in the United States, respectively), and 671-68 (Key Institutions and
the Expansion of Manufacturing Exports), as well as a number of non-RPO
analyses and reports. All but the first two RPO projects are still under
way.
The major thrust of the research included in this cluster is the
hypothesis, prominent at the Bank, that successful expansion and diversifi-
cation of exports is a key characteristic of many recent development experi-
ences. The work done so far includes data compilations, surveys of selected
industries, analyses of commodity markets, studies of problems of import
restrictions by the developed countries on the LDC manufactured exports,
consideration of national policies and institutions for trade promotion in
the developing countries, and analyses of the overall environment for the
exports of these countries. Some of the reports on these subjects can be
characterized as reports aimed at guiding Bank operating departments and
serving as background material for Bank projections and policy work in this
area.
As was the case with the projects discussed under the second clus-
ter, the research methodologies and styles of the projects considered here
have differed widely. The more descriptive studies on selected industries
provide a great deal of empirical information which, however, will be useful
only if they are kept up to date. Among the completed RPO projects the one
on India, while a priori important in itself as a case study of a less suc-
cessful country, was not well conceived and has not achieved publishable
standards. But some of its conclusions fed discussions in India on the Gov-
ermnent's export policies. By contrast, the project on industrialization
and trade policies deserves high praise. It made an important contribution
in documenting the impact on industrial countries of the manufactured export
expansion by developing countries and it has stimulated further research on
adjustment problems in developed countries.
Of great relevance to the operational staff and policy makers
might be the ongoing project on export incentives in developing countries,
Which is designed to yield practical information for countries that contem-
plate the effective promotion of export activities. The ongoing research
on key institutions holds also good promise and might fill a large gap in
existing information on the marketing of exportables. The two ongoing proj-
ects on import market penetration in (twelve) developed countries reflect a
serious attempt at understanding the political economy of trade protectionism.
The topic is important from the export-oriented developing countries' point
of view and the Bank plays the role of a residual supplier of research in
this field.


- 22-
Small Enterprises, Credit Markets and Public Enterprises
The projects here include RPOs 670-77 (Financing of Small-Scale
Industries), 671-59 (Small-Scale Enterprise Development), 671-65 (Small
Enterprise Financing: Role of Informal Credit), 671-69 (Capital Market
Imperfections and Economic Development), and 671-11 (Managerial Structure
and Practices in Public Manufacturing Enterprises), in addition to a number
of non-RPO studies relating to social security institutions. The importance
of these topics for the Bank's lending operations and advisory role in devel-
oping countries is obvious.
Many of the projects have not yet been finished and, in some cases,
they are still at the conceptual stage. It is therefore difficult to provide
for a definite evaluation of the research in this field. In contrast with the
work on incentive regimes and export promotion policies which proceeded under
the presumption that the objective was clear enough and that the task was to
find the appropriate instruments, the projects included in this cluster have
faced much more uncertainty regarding what ought to be achieved. This is an
explanation for the explicitly exploratory character of projects in this
field. The difficulties for the researchers were compounded by the fact
that the existing literature is mainly descriptive rather than analytical.
While applauding the willingness to enter this important area, the
Panel recognizes that research on these topics is still groping for appro-
priate methodologies. We also notice that research on capital market imper-
fections and public enterprises put too much emphasis on the experience of
one country, namely India. Mcreover, the available paners on capital market
imperfections suggest that analytical improvement is still possible in this
field. And while the quality of the research done so far is reasonably good
by international standards, particularly in the case of small scale enter-
prises, it is still below the quality levels achieved in the other areas
reviewed in this report. It may be some time before the Bank develops the
capability to do first rate studies of small-scale industries, credit markets
and public enterprises and their influence on development.
Capital Utilization, Capital-Labor Substitution, and Technological Change
Like the work on incentive regimes, trade patterns, and export pro-
motion, and unlike the research on small enterprises, credit markets, and
public enterprises, Bank research on capital utilization, capital-labor sub-
stitution, and technological change has proceeded within standardized method-
ologies and has attempted to test concrete hypotheses. Included here are
the RPO projects 670-23 (Scope for Capital-Labor Substitution in the Mechani-
cal Engineering Industry), 670-25 (Industrial Capacity Utilization in Selected
Latin American Countries), 670-54 (Employment and Capital-Labor Substitution),
670-95 (Industrial Capacity Utilization), and 671-51 (Appropriate Industrial
Technology). Most of them are completed, the RPO 671-51 being behind sched-
ule. Three non-RPO studies are completed as well.


-23-
The policy thrust of the research has been provided by the observa-
tion that in developing countries labor is cheap and capital expensive rela-
tive to developed ones, and that this ought to be reflected in the use of more
labor-intensive techniques. But while this has occurred to some degree it
is still possible and desirable that the techniques employed be more frugal
in the use of capital. At the same time, it is shown that despite the rela-
tive scarcity of capital in developing countries, productive capacity is not
used very intensively in many cases.
Research has been concerned with market and other forces that explain
the prevailing situation, and with policies that could improve the environment
so that the choice of technique could be made more appropriately, techniques
used more efficiently, and appropriate adaptation and learning proceed more
effectively. On the whole, the research has been of varying quality. The
most serious effort was RPO 670-23, where much care was taken in translating
a well-designed theoretical framework into an empirical analysis. At the
other extreme was RPO 670-54, which had to be abandoned as hopeless. RPOs
670-25 and 670-95 were in between and they did not add much to existing knowl-
edge. The research on the design capabilities of domestic capital goods pro-
ducers (RPO 671-51) is promising, and may lead to important further research.
In general, however, the Panel finds the recent work exploring in great detail
the scope of capital labor substitution in particular technologies important,
but running into diminishing returns as a research endeavor.
Programming in the Manufacturing Sector
The sixth cluster of Bank research is cn investment programming
and has many connections with research on the topic considered in the section
above, although the emphasis so far has been placed on optimization. The
major project included here is RPO 670-24 (Programming in the Manufacturing
Sector). The analytical work has been concerned with optimal programming
of investment where there are significant economies of scale, or strong
interdependence among manufacturing activities as for example the sharing
of machinery. Empirical studies have been done of several process industries
and one non-process industry (mechanical engineering).
The research has estimated the size of scale economies in certain
process industries such as fertilizers, cement, pulp and paper, forest prod-
uCts, energy, etc. (such scale economies are shown to be important in some
cases). Moreover, it has tested the utility and feasibility of using formal
Programming models in guiding investment decisions under economies of scale
(with encouraging results). The research also has considered some of the
implications of economies of scale and strong inter-activity interdependence
for regional cooperation.
So far, the research has been successfully carried out. It has made
a significant methodological contribution, it has the potential to improve the
rationality of government decisions in developing large process industries, and
it may prove useful in guiding Bank lending operations (provided the research
results are available in time and can be understood by the operational staff).


- 24 -
The work complies with both the comparative advantage and the residual suppl,
of research approaches. It is our impression, however, that a stage has been
reached where the methodology developed for process industries has to be dis-
seminated effectively for application in both the Bank and LDCs. The Manuals
currently under preparation constitute an important contribution to these
extension attempts. A conduct of a case study for a sector, which is carried
out with an active participation of a LDC team or institution, would support
the extension work. As far as research on nonprocess industries (namely the
study of the Korean Mechanical Engineering Sector) is concerned, its generali-
zability has not yet been proved, and the majority of the Panel is skeptical
whether an expansion of the project is worthwhile.
B. Overall Evaluation
The Panel attempted to evaluate Bank research on industry and trade
along a range of dimensions, reflecting the multiple purposes of research at
the Bank. Some of our criteria related to the Bank as a research producer and
as a member of the scholarly research community. Here we attempted to assess
the contribution of Bank research to the understanding of the economic devel-
opment process and policy issues relating thereto. What was the absolute
quality (in some sense) of the research output of the Bank? To what extent
did Bank research reflect its comparative advantage? To what extent did
Bank research proceed in conscious awareness of past and ongoing research
elsewhere? Other criteria related to Bank research as it contributes to the
applied objectives of the Bank. How useful has the research been in guiding
Bank decision making, either regarding lending operations or regarding policy
advice? How useful has the research of the Bank been to policy makers in
LDCs? What contribution has the Bank research program made to the building
up of indigenous research capabilities within the developing countries?
Finally, we attempted to identify the factors that seemed to explain
why certain areas or styles of Bank research were more valuable or important
than others. Were there certain styles of research that the Bank did well?
Could one identify certain confluences of factors associated with particu-
larly good and useful research, or poor and not-so-useful research? Were
there certain distinguishing administrative arrangements associated with
good and poor research?
Appendix I treats these questions field by field. The Panel noted
significant differences in the overall quality and relevance of Bank research
in the various fields, and the more fine-grained evaluations also differ from
field to field. However, there were certain general and common judgments that
we made. These we recount below.
By and large, we are impressed by the overall high quality of Bank
research on industry and trade in economic development. Viewed solely in
terms of its research output (much of which has been published), the Bank
clearly ranks as one of the most distinguished development research centers


- 25 -
in the world and certainly the leading one among international organizations.
In many cases, the researchers have made a remarkable effort to improve the
methodology for policy analysis and investment appraisal. The work has been
to a large extent creative rather than imitative and, in its applied aspects,
generallY complementary to the research in the field undertaken elsewhere.
Bank research on industry and trade, being mainly empirically
oriented, has made outstanding contributions to knowledge about the struc-
ture of incentives bearing on business firm decision making about import
substitution and export expansion in developing countries, particularly
regarding the effects of tariff and non-tariff devices. Bank research has
been in the forefront of scholarship positing and supporting the view that
outward looking development policies are feasible and have been highly effec-
tive. More recently, research at the Bank has contributed importantly to
the understanding of changing patterns of LDC exports. Work at the Bank has
shown how resource allocation patterns within a country relate to country
characteristics such as its income level, market size, and policy orienta-
tion. Research at the Bank on intensity and efficiency of use of capital
and labor has documented the wide range of choice of techniques available,
and also the informational and institutional aspects of an economy that bear
on choice of technique. Bank research on programming methods, while not yet
bearing much operational fruit, has explored and pushed forth the state of
the art. Bank research on small enterprises, credit markets and public enter-
prises, while just beginning and still floundering somewhat, has a chance of
providing leadership for a kind of research that has been sadly neglected by
the academic research community, provided high capacity resources are mace
available.
By and large, Bank research on industrialization and trade has
reflected its comparative advantage; but in a number of cases the residual
supply of research approach was pursued. As the research in this field places
high demands on data, much of what was done could not have been done at all,
or would have been very difficult to do, in a university setting. This is
the more so as Bank research in this area has been concentrated on compara-
tive studies, which allow for generalizable policy prescriptions, rather
than on specific cases, which would be of limited value only. With very
few exceptions, Bank research has been undertaken in good awareness of the
state of the art and of what was being done elsewhere.
It has proved much harder for the Panel to evaluate the influence
of Bank research on Bank decision making, or on policy making in the devel-
Oping countries, or upon the strength of the research communities in these
countries, than it has been for us to judge the scholarship on its own terms.
Our discussions with operating personnel within the Bank have helped us to
understand these issues to some extent. The basic problem we had in those
discussions was the tendency for operating people at the Bank (this we believe
is a tendency of operating people everywhere) not to talk about the influence
on their own thinking of the basic ideas and understanding that emanate from
a research tradition regarding the applied problems they face, but to discuss


-26-
the contribution of research in terms of detailed pieces of analysis, or data
that are of use concretely and specifically in decision-making. In our ju
ment, the influence of ideas and concepts on policy-making usually is much
more important than the influence ;f particular facts that might come from
research.
With these caveats in mind, it is our impression that a number of
Jifferent strands of Bank research have influenced, directly and indirectly,
Bank operations. The influence probably has been stronger on operations aime.
at influencing overall policy within countries, than with respect to specific
lending decisions, although there are a number of instances of the latter whe7,
Bank research clearly has had an impact.
The concept and the quantification of effective protection rates
together with the arguments, as well as the evidence, that protection often
leads to uneconomic use of resources clearly was in the heads of the Bank
officials with whom we talked. Similarly, there appeared to be widespread
adherence to the proposition that an exort-oriented development strategy
was an attractive alternative to excessive import-substitution policies for
countries to consider. Both of these notions seemed to be mentally connected
with the view that decision makers did face a choice of techniques, that the
highly capital-intensive techniques of U.S. manufacturing were often uneco-
nomic in the context of less developed countries, but that uneconomically
rigged factor mar".ets and import protection regimes often encouraged and
supported unnecessarily capital-intensive investments.
In their statements about the kind of research that they found
useful, and not so useful, Bank personnel tended to laud studies which pro-
vided data, or examined particular institutions, let alone the whole field
of incentive regimes. It is our conjecture that this very policy-oriented
research may in fact have been more influential regarding decisions on par-
ticular loans than the more general analyses done by Bank researchers. How-
ever, it is the provision of the more sweeping ideas that has influenced
the way Bank officials view appropriate economic development policies and
set their positions in bargaining with LDC officials.
We feel in an even weaker position to judge the impact of Bank
research on policy making in the developing countries. A real impact could
be recognized with regard to the studies on incentives and domestic resource
costs in industrial and agricultural activities. For the other projects, we
would conjecture t1at all of our remarks above obtain. Where the research
done at the Bank has had influence, we suspect this has been largely through
affecting the general climate of thinking, and through its effects on dia-
logue between the Bank and government officials of developing countries. But
we were unable to acquire much direct confirmation of these conjectures. On
the other hand, wc noticed that some shifts in Bank policy thinking (as the
growing interest in the "basic needs" approach) have not (yet) influenced
research either.


- 27 -
Research projects at the Bank have differed significantly in the
extent to which they have contributed to the building up of research capabil-
ities in the less developed countries. There has been very little effort
to work with research institutions in these countries specifically with the
purpose of helping these to develop. Our conversations with researchers at
the Bank indicate a considerable reluctance to do this, on the grounds that
it is very difficult, and would tend to interfere with the task of getting
on with the research.  Some of the Bank's projects have been done almost
exclusively in-house, and have not involved LDC researchers at all. But
a number of the projects, particularly those involving primary data collec-
tion in developing countries, or case studies of particular industries or
policies, have involved researchers in the countries concerned. These proj-
ects, therefore, have helped to bring these researchers into the mainstream
of development research, and to establish or reinforce contacts with the
scholars at the Bank.
Though we have no way of assessing the overall importance of the
contributions to the growth of research capabilities in developing countries
due to the participation of local scholars and research institutions in com-
pleted or ongoing Bank projects, we found some cases in which further research
was stimulated in the countries concerned. Generally speaking, the Bank pol-
icy of working with researchers and institutions of developing countries
should also be recognized as a means of enhancing the research capabilities
in this part of the world.
Our relative assessments of the research projects that have been
undertaken by the Bank in the industry and trade field suggest two strong
correlates of research quality. One is strong interest and leadership by
a senior researcher on the Bank staff; by and large, Bank research has not
been particularly successful when it has been farmed out to consultants. The
second is a confluence of strong conceptual or methodological elements in the
project and a set of broadly but clearly defined questions. In general, we
have not been impressed with the success of Bank projects that have been
motivated largely by either "pure" interests devoid of clear-cut connections
with important policy questions, or a particular policy interest involving
little analytical structuring.
We recognize that the Bank's research portfolio should contain a
diverse mix of projects, involving different degrees of farming out. We
would point out, however, to the fact that quite detailed attention and
involvement of first-rate senior Bank researchers in a project has in the
past been almost a pre-requisite for research success. We also recognize
that in the pulling and tugging between the intellectual interests of the
research staff and the more applied interests of Bank operating officials
the outcome should be a spectrum of projects ranging from relatively basic
to quite applied.  But we propose that the Bank's research successes in the
past have not been at the extremes of that spectrum, but rather on projects
Where intellectual interests and policy concerns in terms of issues and usable
methodologies have come together. As research in the industry and trade field


- 28 -
was mainly applied rather than "pure", policy recommendations made by the Bank
to governments in developing countries were consistent with the most recent
body of knowledge generated in this area.
C. Future Research Priorities
We have argued in Chapter I that the choice of future research prior-
ities has to be based both on the development strategies actually pursued by
member countries and on some assumptions ("vision") about what are the most
important forces and mechanisms of economic development in these countries.
The diversity of the economic institutions and policies of member countries,
and of the views about the development process, suggests a rather "pluralis-
tic" research program for the Bank.
We have suggested four general principles, or motives, for Bank
research:
(a) To contribute to the fund of research knowledge in the
world about the development process; a "comparative
advantage approach" is then appropriate.
(b) To improve upon the research knowledge that is needed
for Bank operations and policies; a "residual supply
of research approach" is then appropriate.
(c) To create externalities within the Bank for its opera-
tional and policy formulating staff in generating a
more analytical view of the problems and an increased
level of "sophistication."
(d) To help generate research knowledge and research
capacity in the LDCs.
This means that recommendations regarding future research priori-
ties must rest on subjective judgments regarding a number of matters, in-
cluding the importance of different kinds of research in enhancing general
understanding of development processes, the comparative advantage of the
Bank in different kinds of research, Bank needs and LDC needs for certain
kinds of studies to enhance their decision-making ability, the kind of re-
search that is likely to attract and hold excellent scholars at the Bank,
and the kind of research most amenable to cooperative endeavors between the
Bank and LDC institutions.
As we have seen in Chapter I, the different principles suggest
somewhat different priorities, strategies and procedures of research within
the Bank. However, in reality, it is not advisable to choose one of them but
rather to make compromises between them. If we would emphasize some aspects
of such a compromise more than others, we would suggest that research is con-
centrated in fields where:


29 -
(a) knowledge is particularly strongly needed for
Bank lending and policy advising;
(b) the Bank in its operations acquires research com-
petence that is unique;
(c) a strong research organization and a system of
follow-up research, mainly in the case of large
projects, are required.
A basic requirement in all three cases is that the Bank have, or be able to
hire, highly competent researchers.
Appendix I presents rather detailed views about the kinds of research
that, according to our view, ought to be cut back and the kinds that ought to
be augmented, for each of the six broad fields of evaluation. Here we attempt
only a general statement of research priorities. It may be useful to consider
first the possibilities of freeing research resources from previous research
areas, and thereafter to consider areas into which we recommend the Bank to
put more resources.
We think that there are certain lines of research at the Bank which
in the past have been forceful and productive, but which now are running into
diminishing returns. These include such traditional and successful Bank
research fields as research on rates of effective protection or subsidy, and
on patterns of growth and development. In both of these fields Bank research
has broken new ground, but the ground is now well broken.
In the case of research on trade policy incentives, it is reasonable
to argue that the research phase is now largely over, and that what remains
to be done are further applications - by including more countries, and by
updating previous calculations. However, the resources for these activities
should, in our judgment, not be taken from the research budget, but either
from the budget for operations and policy formation of the Bank, or from a
special (separate) budget to be allocated to the regions and the other oper-
ating units for application of research after the "pure" research phase is
over. Otherwise the suggested research units for application would perhaps
not be able to shield their resources from the demand of operational work.
Similarly, while Bank research on patterns and sources of growth,
based on regression and input-output analysis, has been useful and illumi-
nating, it is unlikely that much new will be learned from doing more of
these studies, or from doing them in a slightly different and more sophis-
ticated way. Thus, the studies of patterns and sources of growth are also
zainly completed, or near completion. However, it is not clear whether these
studies lend themselves to application for the use by operational staff; the
studies have perhaps mainly served to improve rather general knowledge about
the development process.


- 30 -
We also propose that Bank research exploring the range of technica
choice and opportunities for capital-labor substitution has run into diminish-
ing returns. The basic points have been well documented. It is unlikely that
doing more studies would add much to the ability to persuade people that in
fact tl- range of choice is quite wide, and that it matters what choices are
made. The Bank lending departments need to be able to do these kinds of
studies themselves in the context of exploration of the range of choices
available for particular investment programs they are contemplating, and
to educate and persuade borrowing governments or governmental agencies about
the range of choice. We propose that this body of work, like the work on
effective protection rates, should be moved out of research and moved into
applications.
We have the same judgment regarding Bank research on industry in-
vestment progr;mming for process industries, though some "software" develop-
ment is necessary to make applications more routinized. What is needed now
is for the operating departments to develop the capability to work with the
models; moreover, extension services for application to practitioners outside
the Bank are also required, since this type of study will generally be under-
taken in the LDCs.
In the case of both labor-capital substitution and process indus-
try programming, the work on applications should probably not be done in the
regions but rather in some more centrally located unit in the Bank - consid-
ering the size and complexity of analysis of this type.
The Panel is somewhat divided regarding whether or not the Bank
should cut back on its research on programming models for non-process indus-
tries, and economy-wide models based on a computable general equilibrium
framework. We recognize that professional research on quantitative general
equilibrium analysis and modelling is proceeding at many institutions around
the world and that it is considered a challenging and potentially rewarding
field of activity. Most of us doubt that these bodies of research will con-
tribute much directly to understanding relevant to policy-making. We believe
that an understanding of the development process requires mainly other types
of knowledge than is likely to be acquired by these projects. Thus, the
majority of us are rather skeptical about the fruitfulness of this type of
research, relative to some other fields.
It may be useful if we spelled out the grounds for our skepticism.
General equilibrium models have obvious and important attractions as instru-
ments of economy-wide economic analysis. They have the advantage of making
explicit the interconnections among economic activities, the complex and not
always immediately obvious relations between prices and quantities, and the
economy-wide impact of policies that work through the price system. But asso-
ciated with these advantages, the large-scale general equilibrium models have
two disadvantages, which, in the view of this Panel, weigh heavily against
them.


- 31 -
One is that these models tend to be so complex that the mechanisms
which they describe are difficult to evaluate and modify in the light of
knowledge about what the model leaves out or oversimplifies while the addi-
tional insight gained by more complexity is often not worth the effort.
Second, the equilibrium character of these models makes it very difficult
to deal with the fact and the consequences of technical and institutional
changes which, we believe, are the prime driving forces in economic devel-
opment. These changes can be built into the model, but only in a rather
mechanical way. Our concern is that economic development is an inherently
dynamic process with constantly changing structures and this is basically
what these models tend to repress.
We recognize that not all economists share our views on these
matters. We also believe it important that the Bank keep up with new trends
in development economics, and the use of computable general equilibrium models
may be such a trend. Also some of the Bank's most highly trained economists
are interested in this work. These arguments make some of the Panel reluc-
tant to urge that the Bank withdraw from the field, and willing to urge con-
tinuing support of a low cost effort. But the Panel is unanimous that the
most important research topics lie elsewhere.
We think there are several broad fields of research to which the
Bank should allocate more resources. These include, first, a set of topics
related to exports of the LDCs, evolving patterns of international trade, and
policies of LDCs that reflect trade opportunities and constraints. Second,
a set of topics concerned with economic institutions and domestic economic
policies that facilitates or deters development. Third, research relatimg
to adoption of technology, innovation, and technical change in developing
countries.
Topics Related to Trade
Over the years the Bank has built up a comparative advantage in
research relating to LDC trade patterns and policies. Research in these
fields has contributed importantly to policy thinking within the Bank, and
has been sensitive to policy questions, thus also meeting the "residual
supplier" criterion. The research that we recommend below represents con-
tinuation and strengthening of work now under way at the Bank and we hope
that present staff constraints in this field can be relieved soon.
We believe that Bank research on export promotion policies and mar-
ket access should be continued and expanded. We think particular emphasis
should be placed on studying evolving trade patterns, with special focus on
inter-LDC trade. Productive areas of research include cost-benefit analyses
of inter-regional trade in the framework of preferential agreements, an eval-
uation of different avenues of economic integration as well as an assessment
of the prospects for such an integration, and the effects of common financial
institutions on investment patterns. Studies in this field should not be con-
fined to regional arrangements, but should also analyze in depth inter-LDC
trade that takes place independently of such agreements.


- 32 -
We also feel that there is a need for a more systematic differen-
tiation among LDCs with regard to industrial growth paths, trade patterns
and policies to support them. Developing countries can be grouped in those
,.at are rich in natural resources, those that are just beginning their
industrialization, and those that are still extremely poor and have not
vet sta.rted to develop manufacturing activities. An alternative distinc-
tion could be made between countries rich in mineral resources, those with
substantial agricultural resources and low population density, and mineral-
poor countries that are densely populated. These groups of countries are
characterizoad by significant differences in terms of domestic saving poten-
tial, labor skills, entrepreneurship, economic structure, export orientation,
the role of direct foreign investment, etc. The development model of semi-
indu,ztrial economics in Latin America and East Asia, extensively studied in
the past, does not cover the wide range of issues that the "late starting"
LDCs must face. By analyzing them in depth, and evaluating their trade
opportunities, country and sector work in the Bank will obtain firmer oper-
ational guidance. In this connection, the research should also provide
insight into the prospects of these "late starters" for self-sustained eco-
aomic growth and successful integration in the world economy. It should
also assess the (potential) advantages and disadvantages of pursuing the
"baFic needs" approach combined with a strategy of (total or selective)
"delinking" from the international economic system, as forcefully advocated
for in some respectable quarters.
We have also seriously considered the idea that the Bank should
launch a major research effort to the question of adjustment mechanisms
to reallocate :-sources in the developed countries in response to changes
in technologies, preferences and comparative advantages in the world economy.
:ain reason for such a research effort would be that one of the main things
::e developed countries could do for the LDCs is to adjust their own economies
to the export efforts of the LDCs, i.e., to provide access to markets for these
exports. However, we believe that the developed countries should do this type
of research themselves. Our recommendation on this issue is therefore that
the World Bank strongly advise the developed countries to give high priority
to research on reallocation of resources and adjustment policies in the devel-
oped countries, rather than the World Bank moving heavily into that area in
its own research.
Tooics Related to Internal Policies and Institutions
As with research on LDC trade, the World Bank over the years has
been a leading institution doing research on the distortions to economic
allocation and deterrence to economic efficiency associated with protection
of domestic industry from external competition. We believe the time is now
ripe for the Bank to shift the focus from the economic effects of tariff pol-
icies to more general consideration of how policies and institutions influence
resource allocation and efficiency within a country. We propose that such
research fruitfully can exploit the comparative advantage of the Bank in
doing comparative country studies.


- 33 -
Earlier work on capital utilization and capital-labor substitution
led to a recognition that factor market conditions played an important role
in influencing choices. In turn, labor and capital markets are strongly in-
fluenced by a variety of government policies. These policies, for example,
labor legislation, and policies imbedded in financial institutions, warrant
considerable study on a comparative basis.
We think the Bank should venture into the study of industrial
organization in LDCs and exploration of the effect of different kinds of
policies on organizations and on economic performance. Many LDCs engage in
price control activities, and other forms of price and input regulation. It
would be fruitful, we believe, to examine the effects of regulation in devel-
oped countries. Bank research on small-scale enterprise is evolving to con-
sider the effects of controlled markets and imperfect markets more generally
on the viability of small firms. We endorse research in this field and sug-
gest considerable efforts be made to apply best available methodologies and
to try to improve these. We think it would be fruitful to study more gener-
ally whether it is economies of scale or market imperfections, that support
the monopolistic or tight oligopolistic structures that mark many LDC indus-
tries.
Moreover, in many LDCs, public enterprises usually provide trans-
port, power, and a variety of other public services. Many countries are also
employing public enterprises for the production of manufactured goods, par-
ticularly when significant economies of scale are involved. The question of
the relationship of public enterprise with other industries and with govern-
ment pricing and incentive policies, and more general issues relating to man-
agement and investment planning in public enterprises, strikes us as important
to study, probably in a country- or industry-specific context. The World Bank
has initiated some research in this field. We urge that the field be given
quite high priority.
Topics Related to Innovation, Entrepreneurship, and Technological Change
Earlier we expressed our belief that economic development must be
understood as a process involving technological advance in an essential way.
Bank research on capital-labor substitution and appropriate technology in-
creasingly is recognizing this. We recommend that the Bank explicitly and
consciously do research on mechanisms of technology transfer, adaptation of
technology to better fit local economic conditions, innovation in industry
in less developed countries, and the policies and institutions that support
and stimulate technological progressivity.
Bank research in several different areas increasingly has come to
recognize that choice and implementation of technologies is a much more active
and creative process than sometimes presumed. A considerable amount of re-
design, adaptation and learning often is involved in "technology transfer."
Several recent studies have shown domestically adapted or invented tech-
nologies to be playing a significant role in growth of productivity in manu-
facturing industries in certain less developed countries, and to be occurring
in exports. We think that the Bank should join more actively and provide


-34-
greater support for research trying to understand and better characterize
the nature of the processes involved.
A number of important policy questions are at stake. For example,
it would seem to be important to know the extent to which having a number of
well-trained engineers in a company facilitates their choice of techniques,
adaotation, and innovation. One can go on to probe regarding the kind of
training that effective engineers have had, and to ask whether this is the
kind of training that is offered by the country's engineering schools.
It is important to gain a better understanding of what kinds of
firms are adapting and innovating most successfully. Do they tend to be
small, medium size or large? Do small innovative firms tend to grow larger?
Are there differences between domestically owned firms and subsidiaries of
foreign corporations? Between private and public firms? We think it of
high priority that the Bank begin to study these questions.
Arong the important policy and institutional topics for study,
examination of a set of issues relating to entrepreneurship strikes the
Panel as particularly important. This is not only a field of industrial
organization - including issues such as market structure, types of competi-
tion, and the supply of equity capital - but also a sociological problem
concerning attitudes to entrepreneurship in society.
To summarize our recommendations about future research priori-
ties, we think that the three broad areas described above - international
trade patterns and inter-LDC trade; studies of factor market distortions,
Dolicies and institutions (comparative studies); and study of entrepreneur-
ship and processes of adaptation and innovation - delineate the broad areas
to which the Bank should be allocating more of its research resources.
If the Bank does decide to increase significantly its,research
efforts in certain new fields, our observations about the kinds of research
that the Bank has done well or poorly in the past might be kept in mind. The
projects chosen should involve a blend of analytical and policy questions.
And there must be a senior researcher at the Bank knowledgeable about and
interested in the research. Some of the new departures we suggest represent
natural evolution of the research and interests of researchers currently at
the Bank. But we believe that to design and carry out the research well, the
Bank is going to need some new research talent with skills currently not well
represented at the Bank. We recommend strongly that the Bank hire some first-
rate researchers with experience in analyzing questions of industrial organi-
zation and technical advance. Where senior Bank researchers are moving into
these newer fields, additional appointments could be of more junior research-
ers. But we suspect some new senior appointments would be very helpful.
To facilitate the design of some of the new projects, the Bank might
consider establishing groups of consultants to discuss with Bank researchers
the existing state of research in fields that the Bank is entering, to help
identify promising research opportunities and methodologies. But, while such
consultative groups can help the Bank get into a field, over the long run
there is no substitute for strong in-house talent.


- 35 -
To avoid that research in the new fields that are recommended here
ends up with descriptions of institutions and policies that do not lead to
generalizations, we would recommend new research departures with a wide rele-
vance, promising reproducibility of the results.
Even though we have suggested that some research areas are now
mature for application, that others should perhaps be phased out (though
first brought to completion in the most promising cases), and finally that
other types of research should not be "moved into", it is obvious that our
suggestions would require a somewhat larger research budget in the field of
industry and trade. However, we believe that this would worthwhile for the
Bank, considering how important it is that the Bank has the highest possible
competence in the field of its activities, among which operations related to
industrialization and trade are prominent. It is, we believe, the competence
of the Bank, rather than its lending volume, that will count for its contribu-
tion to the economies of the less developed countries.
Against this background, it is not unreasonable to increase the
number of scholars of the Bank in this field with at least a handful (approx-
imately five), highly competent persons. This is, in fact, a prerequisite
for shifting research to the areas that in our opinion should be given higher
priority in the future than in the past. It will, of course, be the number
of new researchers and the scale of additional resources made available to
them which determines the number of new research areas that should be entered.


APPENDICES I AND Il


- 36 -
Appendix I
Page 1
INDIVIDUAL FIELD EVALUATIONS AND PROPOSALS
1.1 Incentive Policies and Economic Integration
Juergen Donges
The Subject
The following evaluation is based on a sample of five RPO projects:
(i) RPO 670-01, Development Strategies in Semi-Industrial Countries (directed
by Bela Balassa); (ii) RPO 670-22, Economies of Scale and Tariff Levels (Garry
pursell); (iii) RPO 670-87, Industrial Policies and Economic Integration in
west Africa (directed by Bela Balassa); (iv) RPO 671-10, Promotion of Nontradi-
tional Exports (staff member responsible: David Greene); (v) 671-75, Inter-
national Trade Policy for the Development of Bangladesh (staff member respon-
sible: C. Jayarajah). Four additional non-RPO projects (three by Balassa,
one by Stephen Guisinger) will be referred to where appropriate. Seven out
of the nine projects under review have been completed - successfully as will
be shown below. RPO 671-75 is to be completed by November 1979, whereas
RPO 670-22 has never produced a final report.
Most projects deal with the role of incentive systems in economic
development, from both a theoretical and an empirical standpoint. They
examine primarily alternative forms of import protection, export promotion
and economic integration, and they analyze the effects of these measures
on the allocation of resources, the balance of payments, the generation of
employment, and the overall growth in developing countries. Moreover, there
have been attempts to evaluate incentives on a firm-by-firm basis and to
appraise the impact of protection at the level of individual investment proj-
ects.
In view of the widespread belief among policy makers in many LDCs
that their economies cannot develop (i.e. industrialize) without direct gov-
ernment interference in the market mechanism, the focus of these projects
is essentially applied. Not only could the findings, if accurately substan-
tiated, provide the government officials of LDCs with guidelines for apprais-
ing the allocative, growth, distributional, and balance-of-payments impact
of the incentive regimes, and for reforming them where necessary, but the
studies could also strengthen the ability of the Bank's operational depart-
ments to assess both the economic feasibility of individual investment proj-
ects and the overall effects of the economic policies pursued in LDCs. Most
of the studies under review meet these objectives.
Project Review
From both an operational and a policy-making point of view, the
main contribution in the field under review is the project on Development
Strategies in Semi-Industrial Countries (henceforth "Strategies"), by which
the World Bank has continued and deepened its research on (trade) incentives;
Work in this area has been closely connected with the Bank's main policy
interests since 1967. The "Strategies" project started in mid-1971 and was


- 37 -
Appendix I
Page 2
completed, with a formidable volume of over 1,200 (double-spaced) pages an,
tables, in mid-1978. Its purpose was to quantify and analyze relative incen-
tives provi;ed -o import substitution and export expansion in countries that
have already established a relatively diversified manufacturing sector and
that are pursuing different development strategies.
For the sake of comparability, the oroject concentrated on six
countries at a siiar level of industrial development: Argentina, Colombia,
Israel, Korea, Singapore, and Taiwan. Argentina and Singapore are relatively
new to the group of countries studied in this field, which is an important
aspect in itself, because they represent the two extremes of economic policy
shaping: highly Interventionistic in one case, and strongly market-oriented
in the other. The six country studies followed up prior methodological and
empirical research done in and outside the Bank, and so did the comparative
analysis of the experience of the six. On the whole, the policy corrclusions
and recommendations rest upon firm theoretical foundations and a sound factual
basis.
The methodological framework underlying this project is the effec-
tive protection concept. It has been extended to include export subsidies
as well as credit and tax preferences, in addition to import tariffs and
quotas. By d-in  this, fortuncL+ly at = reasonable level of sector disagre-
gation, the "Strategies" project has increased the knowledge inside and out-
side the Bank about the effects of incentive regimes. This study is an
example of how a-alytical tools, which have proved useful in earlier analyses,
can be improved in theoretical and computational terms. It is particularly
noteworthy that, unlike many other studies in this field, a great effort has
been made to collect comparative price data. Only price comparisons between
domestic production and import values make it possible to calculate overall
effective protection (i.e. including non-tariff measures) and to identify
the degree of tariff redundancy, which often exists.
That the analysis in the "Strategies" project has been carried out
in a partial equilibrium framework does not reduce its practical relevance,
provided the estimates are taken as orders of magnitude and their economic
implications are interpreted in a comparative way rather than in isolation.
General equilibrium models tend to have a rather limited explanatory power,
let alone their high sensitivity to the parameter specifications.
What could have been wished from the user's point of view, however,
is either that the country studies had been completed earlier or that the
quantitative estimates had been more up to date. Earlier completion of the
project - originally planned for the second half of FY73 - was outside the
realm of possibility as various researchers were unexpectedly absorbed again
and again by other duties in their own countries. Thus, the delay has to be
interpreted as the price to be paid for embarking on large cooperative ven-
tures amongst Bank's researchers and external economists; the "price" has
been kept as low as pos5ible, however, by circulating draft versions of var-
ious chapters well ahead of the project's completion.


- 38
Appendix I
Page 3
With regard to an updating of the results, one should remember that
the quantification of the incentive measures refers to a single year in the
end-sixties. While this helps to understand the past economic performance
of the countries under study and to assess cross-section indicators of com-
parative advantage, most sample countries have changed these measures, espe-
cially those related to the foreign sector, significantly since then. Had
the "Strategies" project included these changes in the analysis, its results
would have become even more significant than they are in any case. Of course,
we are aware of the practical difficulties in keeping current a comprehensive
analysis such as this one; it probably would have required a full-scale re-
estimation of the effective rates of incentives. We nevertheless think that
there is a need for the Bank, when applying this research, to bring the results
up-to-date. The Bank is, as compared to academic institutions, in a favorable
position to make such an effort; it has the staff and it can try to collect at
least part of the required information through the economic missions that regu-
larly review the LDCs. We were told, however, that many such missions have
attempted to collect these data, but that this effort has not contributed sig-
nificantly to regular updating. In such circumstances the Bank should encour-
age competent organizations within the countries concerned to do this work.
The message of the research on incentive schemes is that, whenever
LDCs want to industrialize and think that this objective requires government
assistance, they should promote rather than protect industries. This implies,
for instance, that import substitution should and could be achieved without
discriminating against exports, particularly nontraditional manufactured
exports, as is so often the case in reality. Neutrality between production
for domestic sales and for exports is in the interest of LDCs because, as the
"Strategies" research shows, it will result in faster growth, more employment
creation and higher levels of efficiency than when industrialization is aided
only in relation to the home market.
These are conclusions that can be used immediately by the opera-
tional staff. We were told that this work has been found quite helpful,
particularly so by the country economists in the Regional Departments and
by the IDF and IFC staff; skepticism with regard to the usefulness of this
type of research prevails within the IPD staff, which has little to do with
incentive systems, however.
It should also be noted that researchers from the countries con-
cerned have been involved in the "Strategies" project to a large extent.
This has in itself created some problems, in terms of time schedules as
well as diversity of contents. But it has also contributed to stimulating
policy-oriented research in the countries themselves, which otherwise would
not have been undertaken and, more important for Bank purpose, it has influ-
enced policy decisions in these countries. Argentina is one case in point.
The quantification, for the first time, of effective rates of protection
(taking 1969 as the base year) has influenced significantly the current
thinking of the government in the direction of decreasing the high average
level of protection, of narrowing the dispersion of the effective rates of


A;nendix I
protection and of encouraging. K' meanr of financial incentive", those indus-
tries that havO, or could :-v: 1lp, in1>rnational compietitiveness.  Furthermo,r
one of the authors of the choer on Argentina of the "S*rategies"' project has
recently followed ip chE earlier 4indings, taking 1977 as the year for calcu-
lating effective ratel of protec ion.  We have also found for Colombia and
Israel a growing amc1nt o- r3search resources devoted to the empirical analy-
sis of Crade policy issues, as a follow-up to the "Strateoies"r project. In
Korea, Singacr&, and T _wan, where the authors have moved into proninent
governmental positions, _-onomic policy making takes into account the recom-
mendations that were for aaed in the "Strategies" study. Thus, this work
has contributed, direct-' . indir::l1.i, to promoting applied research in
the Bank's member count'ri: , which w have suggested in Chapter 1 as being
one of the major principles for guiding Bank research.
Studies of this toe are not only worthwhile because they may assis7
in shaping more razic-ial incentive policies in the countries concerned. They
are also iportant for the lessons they can provide for many other LDCs a.
lower zazes of econmic development. As one can never take for granted,
however, that chese lessons will be absorbed by those other countries, it is
always helpful to exte'-  the regional coverage of the research on incentives.
The project on litdustri I Holicieo and Economic Integration in West Africa
was a useful extension of Bank research in this field. Its purpose was to
appraise the potential _f, and obstacles to, industrial development and eco-
nomic growth in small, primary-producing countries under alternative policies,
including expansion of intraregional trade.
Four countries were chosen for analysis: Ghana, the Ivory Coast,
Mali, and Senegal. 1hey differ from those studied in the "Strategies" proj-
ect in that they are less developed and have not yet established a diversi-
fied industrial base. The incentive systems of these countries have been
analyzed with regard to the'r impact on import-substituting industrializa-
tion within each country, ircreased inrraregional trade, and expansion of
exports to third countries; furthermore, the structure of (long-term) com-
parative advantages of these countries, in both agricultural and industrial
activities, has been evaluated.
In measuring the incentives, basically the same methodology was
applied as in the "Strategies" project. However, the structure of effective
rates of protection and subsidy does not reflect accurately the comparative
advantage or disadvantage of particular activities since it is calculated
ex post and on -he basis o' market prices rather than shadow prices. For
this reason, the project has rightly applied the domestic-resource-cost con-
cept as well as estimates of the cost-benefit ratio for foreign capital in
measuring comparative advantages--using, as in the calculation of incentives,
firm level data.
While the findings provide a clear indication as to substantial in-
centive-generated distortionr and a great diversity in the domestic resource
costs of foreign exchange. Li,y -hould be taken as reflecting orders of magni-
tude rather than exact qua- 1ification. As the authors themselves admit, the


- 40 -
Appendix I
Page 5
data are not completely reliable in all cases and the estimation of shadow
prices, particularly of the shadow exchange rate, always involves some uncer-
tainty. The sensitivity of the ranking of activities in domestic resource
costs to the shadow price assumptions has been tested in the studies on
Ghana and the Ivory Coast; such alternative estimates are still to be done
for Mali and Senegal.
Research undertaken in this project, as with the "Strategies" proj-
ect, met the Bank's immediate operational needs. Indicative of this is the
fact that the project was carried out in close cooperation with the Western
Africa Regional Office and with support from the Industrial Projects Depart-
mient (of CPS). The operational staff found the methodology, the empirical
results, and the policy recommendation(s) useful in evaluating government
policies and programs in that area. Recently, the regional office has even
undertaken the initiative to follow up the research of this project by study-
ing the effects of trade and other incentive policies on resource allocation
in Cameroon (which is the dominant member of the Union Douaniere et Economique
de 1'Afrique Centrale). And the Agricultural and Rural Development Department
(of CPS) is now sponsoring a comparative study on the economic effects of price
intervention in agriculture. The West Africa research project has also influ-
enced policy making in the countries concerned. For instance, the analysis of
Senegal has led to the preparation of a tariff and tax reform; and the Nigerian
Government has invited and financed a special World Bank mission to obtain
guidance on the costs and benefits of a revised system of incentives, using
the same methodology as in the West Africa project.
Operational guidance with regard to an Asian nonindustrial country
is expected from the ongoing project on International Trade Policy for the
Development of Bangladesh, which has been initiated by the South Asia Regional
Office. This project may be considered as a country application of the effec-
tive protection-domestic-resource-cost methodology.
It aims at three objectives, according to the project proposal:
(i) to assess the structure and functioning of factor and product markets
ith a view to identifying relative resource scarcities at present and in
five years from now; (ii) to determine the structure of comparative advan-
tages for existing agricultural and industrial activities as well as for new
industries that might be established in the future; (iii) to design a trade
Policy, properly synchronized with domestic incentives measures, conducive
to a removal of the balance-of-payments constraint to Bangladesh's economic
development by encouraging investment in activities in which the country has,
or can develop, a comparative advantage.
Doubtlessly, these are important subjects. Whether or not the analy-
Sis can be conducted at a high rigor will crucially depend on data availabil-
ity. The project proposed displays much optimism in this regard, but there is
a risk that difficulties in collecting the needed information can become enor-
mous. That the execution of the project involves researchers from Bangladeshi
institutions might facilitate the access to existing, though unpublished, data.


- 41 -
Appendix I
Page 6
But the real problem is that the existing stock of statistical information in
Bangladesh is far from being sufficient to measure shadow prices, quantify
incentives, and calculate domestic resource costs. There will be great need
for additional data collection by the researchers themselves. One should
therefore not be surprised if the research leads to a narrower range of per-
mis3ible conclusions than is expected, or if its completion will be delayed.
The internal needs of the Bank relating to incentive policies do
not exhaust themselves with country reviews. As a lending institution, it
should also have a great interest in knowing the degree to which bo:h the
financial and economic profitability of individual investment pro fe:ts are
influenced by trade policies. In this respect, the non-RPO study on Indus-
trial Protection in IFC Projects, assigned to a research consultant, could
be useful--and has in fact been considered helpful by an internal evalua-
tion report.
Nine IFC past investment projects were selected. They are indus-
trial in character and located in Brazil, Colombia, Iran, Nigeria, and
Pakistan (one each) as well as in Ethiopia and India (two each). Basically
the same conceptual framework as in the "Strategies" project referred to
above was used to examine empirically the relationship between levels of
orotection afforded to these projects and their economic and market rates
of return (related to all assets). It is emphasized that the standard neas-
ures of effective rates of protection do not provide all the information
required for making decisions at the project level. Those measures are
inherently static and they are sensitive to the exchange rate used to con-
vert foreign into domestic values. Judgment about the overall financial
profitability of a new investment should also take into account the pattern
of protection (and other incentives) over its lifetime. This can be accom-
plished, in the opinion of the researcher, by means of supplementing the
current IFC technique for appraising the protection aspects of new invest-
ments (the 25-percent guideline on nominal protection) with a benchmark that
indicates the extent to which, in a protection environment, the market rate
of return may exceed the economic rate of return without reducing the eco-
nomic net benefits of the investment to an unacceptable level (a 50-percent
guideline is suggested).
This is an interesting suggestion, as has been recognized as such
in the corresponding offices (including the IFC itself). We have to mention,
however, that the application of the approach suggested may encounter serious
data problems. The "Strategies" project shows how difficult it is to obtain
adequate data for making price comparisons for past periods. It might prove
even more difficult to make such a comparison for a period of years in the
future, as.required ideally in appraising investments under systems of protec-
tion.
These caveats notwithstanding, this project is an example of research
that can probably be undertaken only in an institution like the Bank with a
reasonable chance of success. Researchers in academic institutions frequently


- 42 -
Appendix I
Page 7
are reluctant to embark on large data collection activities, which are regarded
as not adding much to professional reputation. And in a purely academic en-
vironment they may not work hard enough on the methodology of project appraisal
because they do not perceive this issue as crucial as researchers closely con-
nected with an operational staff might do. In fact, literature on project
appraisal shows that much of the pioneering work has been done by senior
researchers under the sponsorship of international organizations involved
in this field.
In estimating the efficiency effects of trade policies, the proj-
ects reviewed hitherto were mainly concerned with the allocation of resources
within the economy. But protection not only permits domestic production that
would not be able to compete with imports in a free-trade situation (allocative
inefficiency), but also permits poor performance in the protected industries,
which would improve if necessary for survival in a competitive environment
(X-inefficiency). Comparison of these types of protection-induced ineffi-
ciencies and an assessment of the relative importance of the two is proper
indeed. The small project on Economies of Scale and Tariff Levels was sup-
posed to study this subject explicitly, taking Colombia, Indonesia, and the
Ivory Coast as case studies.
The research has yielded a first conceptual draft containing a priori
propositions about the X-efficiency consequences of import protection. There
is also a first draft available for the case study on the Ivory Coast (includ-
ing five industry analyses) and, as a by-product, a case study on Iran (with
three industry analyses). The project was, however, never completed: the
access to data proved to be much more difficult than anticipated; in the case
of Indonesia there was no progress at all. The Colombian consultants to whom
the Colombian study was commissioned reportedly did not have the experience
and capability for doing good work in this field, while the responsible re-
searcher in the Bank was absorbed by other duties. In retrospect, the proj-
ect, although reporting some interesting empirical findings that are not
available elsewhere, must be considered as a failure.
The project on Promotion of Nontraditional Exports is, by comparison,
more descriptive. Carried out in cooperation with the UN Economic Commission
for Latin America, the project aimed at evaluating export expansion policies,
with particular reference to Argentina, Brazil, Colombia, and Mexico. For
greater comparability, the project was extended to include also the experience
of non-Latin American countries, namely India, Israel, Korea, and Yugoslavia.
The initiative was taken by the Latin America and the Caribbean Regional
Office.
In general, the country studies have surveyed the existing stock of
knowledge in this field, rather than increasing it through new empirical in-
vestigations. Their significance for policy making is, however, great. It
is shown that the economic policies of developing countries have a substantial
impact on the evolution of new manufactured exports, and in this regard this
Project has complemented the "Strategies" one. It is an example of how Bank
research findings relating to incentive policies can be diffused among both


- 43 -
Appendix I
Page 8
the Bank's operational staff and local researchers as well as government
officials in developing countries; dissemination could have been even wider
had the results of this work been published not only in Spanish but also ia
English.
General Evaluation
The studies on incentive policies and economic integration which ha.
been completed so far show five distinct features:
(a) One is the high professional quality of the work. This
has allowed various authors to have already published
part of the research results in highly reputable journals
(see Annex II), and thereby to contribute to the effec-
tive dissemination of the findings.
(b) Second, in most of the cases the authors have made a
substantial effort to improve the methodology for
policy analysis and investment appraisal. The Bank's
research so far has been to a large extent creative
rather than imitative.
(c) Third, the applied component of the research has been
generally complementary to the research in the field
undertaken elsewhere (UN organizations, OECD, academic
research institutions). There has been, however, some
overlapping with regard to the LDCs chosen for analysis,
which is perhaps a reflection of the uneven distribution
of useful statistical data among LDCs.
(d) Fourth, the studies have been designed with a view to
responding to important needs of the Bank's operational
departments. The research was empirically oriented,
which is of paramount importance if soundly based
policy prescriptions are to be achieved. In fact,
the operational staff is to a large extent convinced
of the relevance of this research, and has expressed
this by applying the findings of the research in
country economic reports and policy analyses, by
using the methodology for in-house quantifications
of trade incentives, and by initiating additional
research in this area.
(e) And, fifth, a great deal of the studies reviewed has
involved participation from developing countries--in
most cases relatively successfully. This is a signi-
ficant achievement since it could help strengthen the
role of rational choices in the political decision-
making process in the countries concerned--to the
benefit of their economic and social progress.


44-
Appendix I
Page 9
As the research in this field makes high demands for data, the Bank
has proved to be an appropriate place for undertaking it, given its experience,
as well as the fact that researchers from academic institutions usually do not
have a comparable access to all information required. Furthermore, while it
is always difficult to sell a research result to a government, the Bank was--
because of its leverage--in a good position to induce governments in develop-
ing countries to draw as many benefits as possible from the flow of thinking
coming from its research units. That research on incentive policies and
economic integration was also undertaken by both academic research institu-
tions and other international organizations should not be considered as an
argument for not doing it at the Bank. Research inside and outside the Bank
was complementary to a significant extent; the Bank exploited its compara-
tive advantage even more by concentrating on comparative studies. And,
more fundamentally, advances in policy-oriented research, including its
methodological foundations, are normally greater, the greater the competition
among researchers--analogous to the productivity growth in an economy resulting
from active competition in product and factor markets.
With regard to the relationship between research and operational
activities, one important lesson can be drawn from the studies on incentive
policies and economic integration: while it may not always be possible to
provide the operational staff with ready-made solutions to pressing problems,
the overall research work in this field can contribute, and in fact has con-
tributed in the past, to sensitize the country economists in the regional
offices and to influence lending operations. Particularly, this work helped
them to fully understand the efficiency problem at both the macro- and micro-
level and to explain its implications to governments in developing countries.
It is noteworthy that, until the early seventies, the Bank's economic and
sector missions to developing countries used to adopt benevolent attitudes
towards (protectionist) import-substitution policies, while they now advo-
cate vigorously for fairly neutral incentive systems combined with reason-
ably liberalized trade regimes and realistic exchange rates. Moreover,
specific investment recommendations are based now more often on comparative
advantage criteria than was the case until recently.
As the effective protection-domestic-resource-cost methodology can
be considered mature, and understandable to the operational staff, the past
research under review has established the conditions for entering the applica-
tion phase according to concrete demands. The preparation of other (compara-
tive) studies on incentive schemes may involve, from a purely academic point
of view, diminishing returns--a point also emphasized by the DPS staff. Yet
many developing countries still favor policies of import substitution, main-
tain high levels of effective protection, and prefer quantitative interven-
tions over pricing incentives. And the Bank will presumably continue lend-
ing to them. A regular analysis of the incentive policies pursued by member
countries could increase the advisory capacity of the operational offices.
Moreover, if incentive schemes are to be changed in order to make develop-
ment policies in developing countries more effective in the medium and long


-45-
Appendix I
Page 10
run, the Bank has the authority to explain this forcefully to national gov-
ernments, whereas individual researchers of academic institutions, including
those of the countries concerned, frequently lack this influence. This is
particularly important with regard to the question of how a successful tran-
sition to a more efficient incentive scheme could be made in practice. The
answer to this question, while touched upon in past work, in any case still
requires a good amount of research.
Areas for Future Research
From the above it follows that Bank research for appraising incen-
tive policies should not be totally discontinued, though shifted in emphasis.
(a) To begin with, it would be desirable to investigate
in depth the existing incentive regimes with regard
to factor markets, the implications of these regimes
for both the functioning of product markets and the
changes in the structure of trade, and the avenues
of eventual reforms.
(b) Second, analyses of the existing or required overall
incentive systems should be oriented towards the fol-
lowing groups of countries: those rich in natural
resources, those that are just beginning their
industrialization, and those that have still very
low per-capita incomes. Such studies are of interest
for assessing these countries' prospects for self-
sustained economic growth and successful integration
into the world economy or, by contrast, the (potential)
advantages and disadvantages of pursuing the "basic
needs" approach combined with a strategy of (total
or selective) "delinking" from the international
economic system.
(c) Third, the continuation of research on promotion and
diversification of manufactured exports, with emphasis
on marketing aspects and on the contributions that
export processing zones can make, is of considerable
importance to developing countries and to Bank policy-
advisory work as well. In this connection, the rela-
tionship between the emergence of export industries
in developing countries and the implementation of
discriminatory protectionist measures by developed
countries should be analyzed further and then kept
under review. Furthermore, policy changes in devel-
oping countries with regard to export promotion and
import substitution as reaction to external shocks
should deserve some investigation.


-46
Appendix I
Page 11
(d) Fourth, and given its importance for Bank lending,
research on future world trade patterns--with par-
ticular reference to changes in export structures of
developing countries according to their stage of devel-
opment, the size of their domestic market, their endow-
ment with natural resources, and skill levels in their
active population--should be intensified.
(e) Fifth, in view of the data intensity of research in
these fields (including applications), it seems to
be worthwhile to consider the establishment of a
central data bank, particularly with information on
world prices.
As regards research on economic integration, there are at least
three subjects where further research is of importance.
(a) One would be to follow up the work on West Africa's
integration referred to above by investigating the
effects of the Fonds Communautaire de Developpement
on investment patterns, the impact of the monetary
union on the international competitiveness of indus-
tries in the countries concerned, and the conse-
quences of the Taxe de Cooperation Regionale for
the structure and intensity of intraregional trade.
(b) Second, the different avenues of economic integration
among developing countries (i.e. regional versus
interregional framework; mutual trade liberalization
versus product specialization agreements; customs
union versus free trade area) should systematically
be analyzed, taking into account the factors that
have contributed to the many failures of integration
schemes in developing countries in the past.
(c) Third, more insights into the costs and benefits of
economic integration among developing countries
should be provided for operational purposes.
These are some illustrative issues that, in our view, have impor-
tant applications in the Bank's advisory, financial and industrial sector
work. Obviously, the various topics are linked together at many points.
As staffing constraints might not allow the expansion of research into all
areas outlined above at the same time, priorities among them should be estab-
lished by the Bank's Research Steering Group according to actual and poten-
tial policy and operational needs with which this Group is more familiar
than we are. We strongly suggest, however, that once resources have been
allocated to specific lines of investigation, the completion of the work
should not be jeopardized by overloading the responsible researcher with
too many duties (as was frequently the case in the past). If the in-house
demand for policy work increases, and if this increase is not only temporary,
we recommend that the Bank enlarges its research staff accordingly.


- 47 -
Appendix I
Page 12
1.2 Comparative Advantage, Trade Patterns and Economic Growth
Edmar L. Bacha
This section addresses a set of projects on general equilibrium
trade models, comparative advantage structures, patterns of industrial devel-
opment, and sources of industrial growth. Six RPO projects plus two non-RPO
papers by Bela Balassa are reviewed, followed by a general evaluation of the
field and a brief discussion of possible future extensions of the research
under review.
Project Review
670-07: International Model. This project proposes to construct
an international general equilibrium model for the world economy, to study
problems of interest to developing countries, such as the consequences of
changes in tariff rates, varying growth rates of industrial countries, and
oil price increases. There are sub-models for each of the three main devel-
oping regions (Latin America, Asia and Africa), for the developed world,
and for the oil producers. The model is based on the principles of general
equilibrium theory, and makes extensive use of activity analysis procedures.
The model was contracted out in September 1972, with the work being
carried out by Victor Ginsburgh and Jean Waelbrock, both as outside consultants
(Waelbrock was at one time a Bank staff member). The project is supervised by
Bela Balassa, and it is practically completed, although the simulation results
are not yet available.
The reviewer had access to two papers of a methodological nature,
the first describing the computational algorithm and the second presenting
the general equilibrium model of world trade. These papers appear to be very
innovative. According to the authors, "this seems to be the first successful
computation of a general equilibrium for a large realistic model." The use
of a General Equilibrium model also "forces the model builder to incorporate
only theoretically significant constraints, avoiding the use of the crude
heuristics which play such a large role in development planning models."
But, once the theoreticians decide to come down to the real world,
they must accept its crude rule that the test of the pudding is in the eating.
From this perspective, a judgment on the project must be suspended until its
empirical findings are published and its results compared with those of less
sophisticated world models.
670-19: Expansion in Manufacturing for Export in Developing Coun-
tries. This project proposes to analyze which industries are moving out of
industrialized countries because of high wages and labor shortages, into
low wage developing countries. It intends to survey export industries that
have been transferred to Singapore, Taiwan, Korea (and possibly Me-xico), as


- 48 -
Appendix I
Page 13
vell as to sample international firms in Japan, the U.S., and Europe, on
their prospects for further shifts of export manufacturing operations to
LDCS.
The objective is to provide developing countries with information
for planning export industries. Knowledge should be added on (i) segments
of industry that have already moved to developing countries, (ii) manpower
and ancillary industrial requirements, (iii) dependence on foreign partners
for technology and marketing, and (iv) outlook for access to foreign markets
in Japan, the U.S., and Europe.
Jack Baranson was the staff member responsible for the project,
for which the first contract was drawn up late in 1971. Two papers were
written by outside consultants. One was on the migration of Japanese indus-
tries to Taiwan and South Korea, and the other on the offshore activities of
the Japanese electronics industry. The project was cut short by Baranson's
departure from the Bank.
An internal evaluation report is available on the two papers, con-
cluding that, while they produced some useful information, they are not of
good quality. Both papers are in fact very descriptive. One at least delivers
its message in a readable format, while the other paper is a poorly written
dossier, with interest only to research analysts in the electronics industry.
The project was a failure from its inception. In view of its poor design,
most probably it would have led nowhere even if Jack Baranson had stayed in
the Bank. This is a pity, as the subject matter appears extremely important
to inform the discussion on industrialization strategies in developing
countries.
670-79: Economic Development of East and Southeast Asia. As with
the previous project, the objective of this one is to identify and analyze the
industries most likely to find it advantageous to relocate their processing
activities from Japan to East and Southeast Asian countries.
The study was divided into two parts. The first was macroeconomic
in nature, involving projections of the Japanese economy and of the structure
of trade between Japan and other main regions through 1985. In the second
part, selected industries such as textiles, electrical machinery, steel and
miscellaneous manufactures, were studied in more detail to identify their
changing comparative advantage within the Japanese economy.
The project was supervised by Parvez Hasan, and carried out by the
staff of the International Development Center of Japan (I.D.C.J.). It lasted
from June 1973 to August 1974, when a final draft report was completed.
This draft was not made available to the reviewer. However, a com-
pletion report by Mr. Hasan says that the draft was disappointing because it
did not provide a comprehensive or cohesive view of Japan's economic rela-
tionship with East and Southeast Asia in the coming decade, which was the
main objective of the study. Apparently, the Bank requested a revised ver-
Sion of the paper, but the I.D.C.J. was unable to provide it.


- 49 -
Appendix I
Page 14
This project, although producing some useful information, can also
be considered a failure. iuch was promised but only a few projection exer-
cises without much interest were delivered.
671-05: Patterns of Industrial Development. The main purpose of
this project seems to be the organization of macroeconomic industry and trade
data banks at the World Bank. The economic analysis is limited to updating
the Chenery-Taylor regressions on patterns of industrial growth, with some
marginal methodological improvements.
The project was started in early 1976 under the responsibility of
Vinod Prakash of the Development Economics Department. Completion was expecte,
by October 1978.
We had access to two papers by Vinod Prakash, with the second, on
the measurement of industrial exports, being an improved version of part of
the first one, on statistical indicators of industrial development. Alterna-
tive definitions of industrial exports, commonly used by U.N. agencies and
economic researchers, are tested. The conclusion is that these definitions
are not interchangeable as commonly believed. Hence, there is an urgent need
for a standard definition of industrial exports, if the data base for country-
comparative studies is to be improved. Adoption by all countries of SITC
(Rev, 2) would permit such uniformization.
This project seems most useful for the World Bank's work on industry
and trade. Its conclusions should be discussed with the operational staff of
the Bank, since apparently some statistical pitfalls are related to specific
country procedures, which country specialists may clarify. For example, the
discrepancy pointed out by Prakash in industrial value added as between the
National Accounts and the Industrial Censuses in Brazil is only apparent.
The problem is that the Brazilian Census Bureau used the concept of "value
of industrial production", which is an intermediate concept between value of
production and value added.
Users' Guides to the data bank have been written, and it is hoped
that the operational staff of the Bank will be using these files in their
country and sector analyses.
671-32: A Comparative Study of the Sources of Industrial Growth
and Structural Change. This project is designed to contribute to an empiri-
cally based theory of industrialization which Hollis Chenery and associates
have been building for the last 19 years.
For each of eight countries, the sources of industrial growth and
structural change are determined using input-output data. Determination of
these sources starts from an accounting proposition asserting that, for any
sector of economic activity, domestic production is equal to the sum of domes-
tic intermediate and final demand plus exports minus imports.


- 50 -
Appendix I
Page 15
Observed changes in sectoral domestic production through time are
attributed to changes in domestic demand, exports, and import substitution,
and "growth contributions" are computed for each of these demand components.
Deviations of the sectoral production levels from a proportional
expansion path (where the factor of proportionality is the growth factor of
domestic income or aggregate value added) are also calculated. Such devia-
tions from homogeneous growth are designated as "structural change". Appli-
cation of formulae derived from the accounting proposition above permits the
calculation of the "contributions to structural change" of each of the demand
components.
The emphasis of the analysis is placed on exports and import sub-
stitution. The purpose is an evaluation of import substitution and export
promotion strategies from a long-term, sector-by-sector perspective, stressing
questions of sequencing as well as problems of transition to a flexible, viable
industrial structure.
Parallel to the country studies, simulation exercises are developed
to assess the relative importance of universal and country-specific influences
in industrial structure and growth. "Normal" patterns for the relative con-
tributions of each demand component to industrial growth and structural change
are estimated from cross-country regressions. These "normal" or universal pat-
terns are assumed to vary with per capita income, population size, and factor
endowment of the country. Deviations from "normality" are country specific
and hopefully may be explained by the development strategy and associated
government policies of the country under consideration.
The project started in early 1976 and completion was expected by
mid-1979. People responsible for the project include Sherman Robinson, Yuji
Kubo, Hollis Chenery, Larry Westphal and Moises Syrquin.
Only part of one country study (on Norway, by Bela Balassa), which
actually antedates the research project, was made available to the reviewer.
Three other papers were read, on the cross-country simulation study, authored
by Chenery and Syrquin.
This project may be considered as the end-product of a twenty-year
quest to establish an empirically based paradigm for modern industrial growth.
It demonstrates the importance that simple ideas may have for the development
of economic knowledge. Chenery's seminal 1960 paper consisted of little more
than the numerical estimation of an accounting identity. However, as his
Papers for the Nobel Symposium and for the Tokyo IEA Congress (the latter
cc-authored by Syrquin) demonstrate, within a 20-year period such simple
exercises made it possible to organize a considerable body of empirical
knowledge regarding the nature of modern industrial growth.


- 51 -
Appendix I
Page 16
It is unlikely that this project will achieve its ambitious initial
aim to develop an analytical framework to articulate the connections between
individual policy instruments, changes in industrial structure, and economic
performance. Ve must wait for the final report to see how close Chenery and
associates have come towards this goal. However, the papers reviewed are
indicative of the high quality of the work done, including the important
methodological contributions by Balassa and Syrquin to the measure of import
substitution.
671-79: Sources of Growth and Productivity Change. This project
is a follow-up to the previous one, with specific focus on Korea, Turkey and
(possibly-) Yugoslavia.
Three levels of study are contemplated: (i) linking the demand
oriented analysis of the previous project to the study of sources of growth
from the supply side; (ii) constructing general equilibrium models for the
countries in the study, aiming at a "consistent, comparable and rigorous"
analysis of different policy packages, and (iii) complementing the previous
analysis with a micro-focused study of two common industrial sectors in the
three countries.
The project started last summer and completion is expected by late
1980. Sherman Robinson and Kermal Dervis direct this research project, with
Larry Westphal being involved in planning, and Yuji Kubo in carrying out the
work.
The reviewer had access to the research proposal, which contains
two appendices, one on the Turkey model and the other on the decomposition
of sources of growth according to factor inputs and technical change.
The study of sources of growth from the supply side is of some
interest. It is a useful way of organizing relevant economic data, which
the demand-oriented view of the previous project leaves relatively untouched.
However, after the methodological and empirical criticisms of this type of
work by Frank Fisher and Griliches and Jorgenson, one perhaps should approach
the study of production function decomposition with more mbdesty than is appar-
ent in Appendix A to the research proposal.
The micro-focused analysis of two sub-sectors seems to be the most
promising part of the research. Unfortunately, the proposal is somewhat
unclear not only on the choice of the sectors but also on the methodological
approach to be adopted.
The general equilibrium models are presented in the proposal as a
superior way of looking at industrialization processes. They would overcome
the partiality of the "sources of growth" studies, and provide an integrated
approach to the problem of industrial expansion. It is our feeling that
experience says differently. Accumulated knowledge on the development of
large models seems to indicate that their main merit is as organizing devices


- 52 -
Appendix I
Page 17
for the collection of relevant data for economic policy making. The important
empirical results that they produce apparently can as well be obtained much
more inexpensively from simple macroeconomic formulations.
In view of this experience, perhaps it is wrong to say that multi-
sector models are a step ahead of demand-oriented sources of growth studies.
Theory simply is not firm enough to allow a reasonably complete specification
of the most relevant socio-economic interactions in the process of industrial
growth. It would seem more appropriate to accept with scientific modesty
the extent of our ignorance, use simple tools, and proceed bit by bit. The
answer apparently lies not in implementing programming models, but in devising
simple ways of approaching critical socio-economic relations, from the empiri-
cal exploration of which additional knowledge may be generated.
Non-RPO: Two Papers by Bela Balassa. The papers by Bela Balassa
contain empirical tests of important predictions of orthodox trade theory.
Trade barriers are presumed to reduce international trade flows and to affect
the international location of production. This topic is analyzed empirically
in Balassa's contribution to the Nobel Symposium.
Balassa's paper to the Tokyo IEA Congress tests the prediction
that comparative advantage in trade in manufactures, as revealed by relative
export shares, is affected by the capital-labor ratio of trading countries.
In devising his tests, Balassa demonstrates once again his justly
praised ingenuity in empirically implementing abstract economic theories.
His results are most comforting to orthodox trade theory. The Tokyo paper
also serves the purpose of dispelling some of the "export pessimism" that
has become fashionable since the oil crisis, when it stresses the importance
of supply side factors to the success of export promotion policies for manu-
factured products.
A minor complaint relates to Balassa's propensity to carry his
findings perhaps a bit too far. For example, on p. 24 of the IEA paper, he
asserts that, "The empirical estimates show that inter-country differences
in the structure of exports are in large part explained by differences in
physical and human capital endowments". First, alternative theories were
not tested; the empirical analysis refers to the statistical significance
of the relevant coefficients for the Hecksher-Ohlin hypothesis, investigated
by means of ordinary least squares. Second, the coefficients of determina-
tion of the regressions explaining export shares are not published (though
we have been told that in the best regressions the coefficient reaches .68).
Hence, the author can say that his results are statistically significant,
but should be cautious in claiming that they explain a large part of the
differences in export structures.
Balassa's "stages approach" to comparative advantage may turn out
to be a useful way of bridging the gap between "static" and "dynamic" views
on desirable patterns of trade specialization for LDCs. Hence, it would be


- 53 -
Appendix I
Page 18
interesting to see him building a model of comparative advantage as a con-
tinuation of his work on the "stages approach" to international trade pat-
terns.
General Evaluation
Seven projects were analyzed. One, consisting of two non-RPO papers
by Balassa, was completed successfully. Two others, on the migration of indus-
tries from Japan to South-East Asia, were clear failures. A fourth one, a
general equilibrium model of international trade, has not yet produced empiri-
cal results. The three others are on sources of industrial growth and struc-
tural change: one is designed to organize a data bank on relevant industry
and trade magnitudes; a second has only recently started; and a third is about
to conclude--but its eight country studies have not yet been made available.
A definite evaluation of the relevance and impact of the projects
is not possible at this stage, as most of the final products have not yet
been written. However, one may say with confidence that, excepting the two
failures, all the material that was reviewed is of the best quality that one
can find anywhere in the profession. One may argue with methodologies or
implicit visions of the world; competence is indisputable.
There is one important lesson to be learned from the failures. The
Bank research group has some difficulties in dealing with outside consultants.
Misguided choices of persons apart, the trouble seems to lie in that the con-
ceptualization of the project is done in detail within the Bank. Then, out-
side practitioners are expected to do high quality research on these topics,
without close monitoring by the Bank people. It cannot work. Either the
consultants are good enough researchers to share in the conceptualization
stage, or they should be handed clearly designed tasks, to be performed under
the supervision of the people responsible for the projects.
Expertise in promoting joint research ventures with outside con-
sultants seems to be lacking. Perhaps the Ford Foundation format of first
designing broad terms of reference and then submitting these to an open in-
ternational competition should be tried, as a supplement to the present pro-
cedures of hiring outside consultants.
Future Directions
The studies of sources of growth on the demand side are nearing
completion. Further accounting exercises of this type do not seem to be
necessary. But one still would want to see additional analytical work on
a framework of analysis that articulates "the connections between individual
policy instruments, changes in industrial structure, and economic perfor-
mance". The purpose would be to develop a kit of tools for development
economics similar to that already available for short-run macroeconomic
policy making. It is not clear that the Bank is better equipped than the
academia to do this type of research, but the policy need is evident in the
LDCs for such an analytical effort.


-54-
Appendix I
Page 19
For reasons that were mentioned here and in Chapter II, we do not
think that it would be advisable to expand the project on sources of growth
from the supply side beyond its present size. But perhaps general equilib-
rium models will be found more useful for general policy analysis outside
the scope of the panel's review.
In spite of the failures of the projects on the export prospects
of labor-intensive manufactured products from LDCs, we consider that this is
a promising line of research. Perhaps this is a good candidate for testing
the proposed system of open international competitions for research projects,
to be sponsored by the World Bank.


- 55 -
Appendix I
Page 20
1.3 Export Promotion Policies in the Less Developed Countries (LDCs)
and Access to Markets in the Developed Countries (DCs)
Jae-ik Kim
The Subject
One of the recent developments in international trade is that a
growing number of countries, mostly middle income countries that have been
recipients of Bank loans, have become exporters of manufactured goods to
developed countries. Typically, these countries have moved away from import
substitution policies and promoted exports of manufactures, particularly
labor-intensive light industrial products. Such development strategies,
however, have brought a reaction from many developed countries, which began
to restrict the new exporters' access to their markets.
Tne twelve studies reviewed signify an important step towards an
understanding of this problem, which could very well develop into a political
issue gravely undermining the Bank's role as a development institution. What
follows is a review of these projects. Most of them contain material that
falls within the following three categories: (i) surveys of selected indus-
tries, data compilations, analyses of co-ciodity markets or the overall environ-
ment of trade; (ii) analyses of problems concerning import restrictions by the
DCs on the LDCs' manufactured goods (the reverse problem, that of obstacles to
inter-LDC trade, is given little attention); and (iii) a discussion of national
policies and the role of institutions in promoting trade in the LDCs.
One immediate problem is that the Bank has a very narrow definition
of research, so that the projects solely in the first category above are not
always considered to be research. As a result this extremely valuable work
has to be conducted under exceptionally difficult conditions with a totally
inadequate allocation of manpower. It is only natural that there are more
problems in this area than any other.
It cannot be stressed too highly how valuable this sort of study
is to policy makers in developing countries. Even countries like Korea, with
relatively sophisticated research institutes, can easily learn of their rela-
tive position in world markets only through this sort of work. From the
point of view of developing countries it is strongly recommended that this
category of research be given a higher status in the Bank. 1/
1/   Some of the remarks made may seem harsh to the staff concerned who know
that, if more time or manpower had been available, the faults could have
been avoided. Where this is the case they should understand that all
this criticism is essentially a commentary on the management, which
deprives them of adequate resources to do their job with optimal effi-
ciency. If the remarks given below seem critical it is also because
of the very high standard of research one has come to expect from the
Bank.


- 56 -
Appendix I
Page 21
All but four of the projects are ongoing and only drafts and pro-
posals were available for review. The three finished works, by Balassa,
aughes and Sharpston, were of notably higher quality, and since this report
was originally drafted there have been further improvements in the other
projects.
Project Review
Plesch's study of Developing Countries' Exports of Electronics and
Electrical Engineering Products, which was compiled over a few months, pre-
sents a survey of the remarkable growth of electronics exports by eight newly
industrializing countries (NICs) to the United States. It will be useful if
it is kept up-to-date. At present production data are only given for 1973
and exports to mid-1977. In view of the recession of 1975 it is too early
to judge how far the trend has been interrupted. It does provide a useful
starting point for serious research, but it could have been improved as an
intermediate product by additions to the policy section and by incorporating
some of the observations of Sharpston.
Keesing, Plesch and Triner's unfinished survey entitled Developing
Countries' Exports of Textiles and Clothing: Perspective and Policy Choices
has an excellent summary, which also deals with the Multi Fiber Agreement
(MFA) complexities in a lucid fashion, and advocates the desirability of a
shift to LDCs of textile manufactures. In textiles and clothing, as in elec-
tronics, the dominance of a few countries is apparent. The authors raise
the interesting question of windfall gains resulting from quotas but, unfor-
tunately, because of restrictions on this type of research, can offer only
"speculative inferences from unsystematic observations" concerning this
aspect. As far as the policy choices section is concerned the recommended
policies could have been more clearly differentiated. Policies recommended
to the firms in textiles and clothing, which assume no changes in the poli-
cies of the importing countries regarding either quota or tariff, should be
distinguished from policies that take into account the impact on the importing
countries' policies of any expansion of exports. Also, policy choices for
government have to be comprehensive, and their successful implementation
requires close cooperation with the exporting countries.
Keesing and Plesch's work, Recent Trends in Manufactured and Total
Exports from Developing Countries, is essentially a commentary on a set of
tables and was compiled in less than a week. It is invaluable as a reference
work, but has a rather limited lifetime unless updated every year. It is
understood that it has subsequently been revised. Again what emerges is
that eight newly industrializing countries (NICs) enjoy 76% of manufactured
exports from LDCs. If more research time were available, could not more
resources be given to investigating unsuccessful countries in all the above
studies?
Morawetz's study (RPO 671-56) on Marketing Manufactured Exports
was intended as an analysis of the institutional and non-price aspects of the
growth in that trade from Colombia. In fact, in a recent seminar, Morawetz


- 5.7 -
Appendix I
Page 22
has shown that this growth was not as remarkable as was first thought. It
is not clear how useful this study will be for countries that hope to expand
their clothing exports to developed markets, and it may prove that the selec-
tion of another country would have been more appropriate.
Baldwin's proposal, The Effects of Increased Imports of Manufac-
tured Goods from Developing Countries in the United States (RPO 671-67), and
Waelbroeck and Others', The Effects of Increased Imports of Manufactured
Goods from Developing Countries in Europe (RPO 671-66), are investigating
the market share of manufactured products from developing countries and the
sources of protectionist pressures in industrialized countries. The European
group also proposes to study the role of the EEC in trade restrictions in the
future. These proposals hold good promise and are to be joined by a study on
a similar basis of Japan and other Pacific countries, which was noc cffered
.or review.
Although much has been written about the economic impact of .anu-
factured imports to developed countries, few studies so far have made any
serious attempt to analyze the political impact of these imports. This sort
of analysis should be very valuable, and should provide important insights
into the political processes that transform these impacts into protectionist
oressures and eventually into protectionist policy actions. Both studies
b?an with identical hypotheses on protectionism. Protectionism and i:s
successful reflection in government policies is believed to be a function
of a dozen or so broadly defined factors, some quantifiable and others not.
It is not clear from the proposals how weights will be assigned to each of
these multiple factors. In the case of the American study, Baldwin originally
proposed to measure the market penetration of some 400 items, classified on
a 4-digit SITC basis for a time span of eight to ten years. The history of
the protectionist trend and its politico-economic causes will be studied,
while both projects will provide empirical analyses of the structure of
protectionism.
It is not possible to review these projects satisfactorily since
they have so far produced no concrete results. A recent paper by Balassa
entitled "The Market Penetration Studies: Approach and Methodology" sets
out standards for methodology and comparability, which suggests that, after
the usual problems where a number of outside consultants are involved, the
Bank is pulling these projects into a clear and thoughtful shape. Recent
reports suggest that close cooperation with the EEC will in the end produce
some very useful results.
Export Incentives in Developing Countries (RPO 671-35), under the
supervision of Balassa, evaluates the export promotion efforts of four devel-
oping countries, with a comparative framework using a cross-section investi-
gation of major export products, and a time-series analysis of the effects
of export-promotion measures. The project is a major effort in data collec-
tion and is designed to yield practical information for other countries that
contemplate the introduction of a system of incentives for export promotion


- 58 -
Appendix I
Page 23
or a reform of the incentive system. This project is of the greatest rele-
vance to operational staff and to policy makers. It is precisely the sort
of work the Bank should be doing. Its grasp of methodology at the initial
stages is much clearer and appropriate than either .aelbroeck or Baldwin at
a comparable stage, and its potential pay-off for developing countries may
even be higher.
Seen from the same perspective, Keesing and Wortzel's proposal,
Key Institutions and the Expansion of Manufactured Exports (RPO 671-68), is
expected to make an equally important contribution. Its purpose is clear,
its frame of investigation concrete, and its theoretical and methodological
foundations sound. As any export promoter knows, marketing has been the most
serious bottleneck for export expansion. Particularly for the inexperienced
exporter, who usually takes low prices for granted and bases his export pro-
jections on them, a marketing bottleneck is a frustrating, and frequently
invisible, roadblock. Keesing's research will therefore fill a large gap
in existing information and be a valuable educational tool for exporters of
consumer goods to the US market.
This project explains the relevance of a number of the surveys
previously reviewed. It frankly admits many of the criticisms a reviewer
might make of those other projects (p. 2) in that "in predicting or pro-
jecting manufactured exports, areas of uncertainty (and plain ignorance)
are also evident in World Bank operational work on manufactured exports,
and in the work of the team proposing this research".
Sharpston's study of International Sub-contracting is an excellent
piece originally printed outside the Bank. Although presumably written in
1974 the quality is such that the analysis is still of great value. He
answers questions raised by other writers, or suggests lines of approach
that have not been taken up. The check list of tasks suitable for sub-
contracting is one of the most valuable short pieces of analysis in this
field. Perhaps the mixed background, part academic, part World Bank, pro-
vides the perfect balance.
On the other hand, Wall's Export Promotion and Preferences: A Case
Study of India (RPO 670-21) used an inadequate methodology to attain the pur-
pose of the project. It failed to identify the system of policy measures
that would be desirable and sufficient in India to enable entrepreneurs to
respond to new opportunities in foreign markets, and was never published.
Nevertheless, in terms of results it was not a total failure. It contributed
to a series of Bank reports on India that were influential, at the time, in
changing India's policy on exports and export incentives. The critical tone
of the report itself prevented it from being published in India, but although
opinion within the Bank is divided, it is understood that it has been influ-
ential in other ways, not least through training local researchers who have
subsequently gone on to assist in policy formulation. This suggests that
even where the end product is faulty, the conducting of extensive research
with the assistance of local scholars may still serve a useful purpose.


- 59 -
Appendix I
Page 2"
The editor, Helen Hughes, and authors who contributed to Indus-
trialization and Trade Policies for the 1970's (RPO 670-20) made an impor-
tant input to documenting recent growth in nanufactured exports from the LDCs
and its impact on industrialized countries. Systematic and comprehensive
studies of the effects of increased manufactured exports from low-income
countries on employment, social policies, and investment in the developed
countries were long overdue. Without effective policy measures on adiust-
ment assistance in developed countries, economic growth in many low-incone
countries will have to slow down, causing a chain reaction. This work, which
contains important statistical information based on a thorough analysis of
primary data, has also stimulated other studies.
Balassa's recent paper, World Trade and the International Econ-
omy: Trends, Prospects and Policies (May 1978), is a timely study, valuable
to the entire development economics community. It compares trade liberali-
zation and economic growth prior to the oil crisis with trade barriers adopted
during the post-crisis recession. Recent non-tariff restrictions, government
aids to industry, international cartels, and market sharing in major developed
countries are documented and their effects evaluated. Concluding that the
risks of new protectionism are high, Balassa proposes policies for long-term
growth, structural adjustment assistance, and an international code of good
conduct. Semi-industrial developing countries were also advised to r.duc?
existing protection, to upgrade and diversify their exports, and to abandon
gradually the export of simple, unskilled-labor-intensive manufactures for
the benefit of countries at lower levels of development, a course of action
that some developing countries in Asia have already set out to follow.
Overall Evaluation
In the past, Bank-financed research has been well respected for
the quality of information on specific problems or situations. The country
reports and industry surveys have filled large gaps in existing knowledge
that other research was not able to fill because of its distance from the
scene. A continuation of this "division of labor" is still a good arrange-
ment. In research, the Bank's comparative advantage lies in its proximity
to ongoing development problems. It has a unique global network of infor-
mation, and Bank researchers, whether in-house or commissioned, are well
received by most member governments.
One would therefore have expected the Bank to have evolved a satis-
factory system of rapid data collection and assimilation. But if we consider
the three categories of work--data compilations and analysis; problems concern-
ing import restrictions; and national incentive policies and the role of insti-
tutions--it is with regard to papers solely falling within the first category
that the Research Department is severely constrained through lack of staff,
and the Bank's narrow definition of research. Apparently the lack of staff
has led to the virtual abandonment of the practice of providing information
and assistance to the operational staff, thus creating an unnecessary and
undesirable gap between research and operational staff. Even more important


60 -
Appendix I
Page 25
than this, there is a vital need throughout the developing world for an in-
depth analysis of what is happening in trade on a global scale, and if the
Bank does not provide the necessary resources to tackle this area, it will
be doing a disservice to its member countries.
Turning to category two, the question of restrictions on trade by
developed countries, it is not clear whether this lies within the Bank's
comparative advantage. However, it has to be recognized that these restric-
tions could do more to upset the evolution of LDC's economies than any other
action. The effect of these restrictions is to tend to undermine the rele-
vance of both the brilliant general theory of comparative advantage stages
of development and specific studies on the income elasticity of manufactured
exports from developing countries. The fundamental causes of protectionism
need to be better understood and developing countries and various lending
agencies that must predict the expected return on investment in manufacturing
for export need to tackle this problem. The next stage would be to analyze
which forms of protectionism have the most serious and detrimental effects
on LDCs.
Turning to category three, it is here that the Bank has a clear com-
parative advantage. The relevance and application of the results of such re-
search to the Bank and policy makers are clear and there is little doubt that
only the Bank could pursue this line. It is noticeable that the methodology
and general approach are much clearer in the proposal of Balassa (RPO 671-35)
and that of Keesing and Wortzel (RPO 671-68) than in the others submitted
for review. One hopes that the general surveys in category one will be kept
updated in the interest of these projects, which they are clearly intended to
service.
Possible Directions of Future Research
Judging by the project descriptions, more research could be com-
missioned on several new areas that so far seem to have escaped professional
scrutiny.  Some of these proposals technically fall outside the Bank's defi-
nition of research. The Bank should probably reconsider its rigid boundaries
and whether they really lead to optimal efficiency. Restrictions on market
access for manufactured exports from developing countries, which are expected
to intensify, will increase the need to monitor protectionism in industrial-
ized countries and their adjustment policies. The same trend will encourage
further research on manufactured trade between developing countries. In
connection with this, the Bank might initiate a series of research projects
on the feasibility of industrial cooperation between developed countries and
NICs with an eye on positive adjustment through intra-industry specializa-
tion. At the same time, as suggested above, an analysis of which forms of
protectionism have the most serious and detrimental effects on LDCs would
be most useful.


- 61 -
Appendix I
Page 26
With regard to export-promotion policies, considering the impor-
tance of socio-cultural factors that affect a country's entrepreneurship
and that vary a great deal from one country to another, it is essential for
each country to identify in advance major inhibiting factors that might
otherwise render the transplanting of promotion policies ineffective, and
then devise methods that will circumvent them. Furthermore, LDCs, whether
NICs or not, may have to adopt a new perspective on access to markets. The
Bank can assist LDCs by-launching projects on trade restrictions by NICs
against each other. The trade potential of developing countries, in general,
needs to be assessed, not on the basis of past performance records, but on
their future possibilities, before the trade barriers start to arise between
developing countries.


- 62 -
Appendix I
Page 27
I.4 Small Enterprises, Credit Markets, and Public Enterprises
Gerardo Bueno
The Subject
The following discussion treats each of the above subjects in
a separate manner, although it could be argued that they are, in some sense,
interrelaced. It rests on the analysis of five RPO projects and two non-RPO
projects, which are mentioned throughout Appendix I, as well as in other
relevant documents.
The themes are important for the Bank from an operational angle,
while the research will be valuable for both the Bank and other potential
users provided it is of good quality and that it provides the relevant in-
sights. This appears, in general, to be the case. Small- and medium-scale
enterprises, public enterprises and credit markets are subjects that appear
in the fora of discussion, particularly within the LDCs, where it seems impor-
tant to have a better understanding of the problems involved for policy for-
mulation. Progress in research in these areas by the Bank will allow it to
increase its capacity to give meaningful and relevant policy recommendations.
Still, in evaluating the research done in these areas, at least
two points should be kept in mind. First, that they are relatively "new"
areas of Bank research. Therefore the Bank, as yet, has neither the expe-
rience nor the influence on policy recommendations that it has gained in
other research. Second, many of the projects have not been finished and,
indeed, in some cases, they are still at the conceptual stage.
They are, however, very promising areas for relevant research on
which, we think, it will be necessary for the Bank to consider an intensifi-
cation of its efforts.
Small Enterprises
Research on small enterprises has been to a large extent the result
of a growing interest in the Bank in these questions at the policy formulation
and the operative levels. Not only does one find a sectoral policy paper
(Employment and Development of Small Enterprises); the subject is also fre-
quently referred to in the latest Bank annual reports and in the President's
addresses to the Board of Governors. 1/ Accordingly, and although the subject
1/   The Bank's position in relation to these questions is summarized in its
latest annual report: "The Bank has found that the benefits of growth
cannot be assumed to'trickle down automatically; to ensure that devel-
opment benefits the poorest it must be deliberately directed to the
poorest--(thus)--financing for industrial development (should) shift
towards small scale enterprises."


- 63 -
Appendix I
Page 28
does not fall squarely within the Bank's traditional lines of research, the
decision was taken to support it. Thus, it appears to be one instance where
the need for operative and policy formulations determined to a large extent
the opening of new lines of research.
In other places, too, the subject is very much in vogue on account
of the relationships that would appear to exist between small-scale enter-
prises' (SSEs) development and the growth of employment. SSEs have become
fashionable not only in international and regional organizations but also
on the national level in both developed and developing countries. However,
it can validly be affirmed that up to now this interest has manifested itself
mainly through the formulation of programs to support SSEs and in the adoption
of several specific measures to that end; but little has so far been done in
terms of policy formulation in the context of a global strategy. In short,
the emphasis has been placed more on action than on reflection. Research, on
the other hand, has tended to be descriptive and of a general character as
well as repetitive.
The approach taken by the Bank is more original and appears to be
more promising. To a large extent it is in accordance with the view that
regards development as a dynamic process involving innovation, imitation,
learning and competition in a Schumpeterian sense rather than as a movement
towards competitive equilibrium. The possibility of falling into some pic-
falls in following this approach is by no means negligible; but nevertheless
it is the right approach, and what may be suggested here is that the Bank
should be more explicit about it.
Research by the Bank in this area started more intensively with
the RPO project 670-77 (Financing of Small-Scale Industries); in addition,
there were two other non-RPO projects on Brazil and one on Malaysia. The
two non-RPO projects on Brazil are econometric research of good quality, the
main conclusions of which were that the SSEs sector had lost in relative
importance and that new SSEs had tended to concentrate on the growth poles
of Southeast Brazil. Therefore, they recommended (something that the Brazil-
ians were already doing) the formulation of support programs for SSEs giving
special attention to geographical distri'ution. The conclusions, as can rbe
seen, were not particularly different from those reached in similar studies.
The follow-up was RPO project 671-59 (Small-Scale Enterprise Devel-
opment). The formulation of the project owes much to lan Little, Dennis
Anderson and Larry Westphal, as well as, at a later stage, to Mariluz Cortes
and Dipak Mazumdar, and to their dissatisfaction with the type of research
that was being done, particularly with the absence of ways and means that
would allow the evaluation of different policy alternatives. The project
has two main objectives: on one hand, a review of the existing information
and literature on small enterprises in developing countries and, through sur-
veys, an attempt to define ways in which the information could be improved
and certain policy recommendations be formulated. On the other, the second
objective, which should be the most interesting in the medium term, is to


-64 -
Appendix I
Page 29
develop a basis for assessing the impact on income and employment of various
policy options and an attempt to estimate the costs and benefits of various
policies for small enterprises.
The first phase of the project includes a review of the literature
as well as case studies and enterprise surveys. The studies will cover India,
Colombia, the Philippines, Japan, Korea, Taiwan and Nigeria, and the enter-
prise surveys India and Colombia. The review stage of the project is not
confined to a useful revision of the literature and programs of support for
small and medium-sized enterprises; in fact, the main aim will be to examine
the role they play within the overall development strategies of the countries.
This, it is hoped, will permit the development of a series of policy recommen-
dations.
The Bank's approach departs in a useful and constructive manner
from the more general, descriptive approach followed in most countries and
by international organizations. This being so, it would have been beneficial
to extend the work to include more countries, especially given the fact that
in many countries there is an increasing interest in the subject. However,
this has not been possible because of staff and budget restraints. The work
carried out so far in spite of these limitations has been, indeed, quite sub-
stantial.
RPO 671-59 was started in November 1977 and the first phase is
expected to be completed by September 1979. Three papers have already been
completed, two of which touch mainly upon methodological questions, and the
third one presents in draft form the preliminary results of the small- and
medium-scale enterprises survey of Colombia.
On the methodological side, the first two papers cover, respectively,
the problems that SSEs face in the credit and financial markets and in regard
to technological development. The first one has made use, though to a limited
extent, of research carried out elsewhere in the Bank on credit market imper-
fections, and the second, of some of the work done in the area of capital-
labor substitution. Both of them present suggested approaches, more particu-
larly in relation to the distortions in LDC economies, which SSEs must face.
A lot of work remains to be done, and it is difficult at this stage
to provide an evaluation of the whole project. Nonetheless, some insights
into the possible outcome are provided by the draft paper on SSEs in Colombia
for three sectors (producers of pumps and compressors, producers of agricul-
tural implements and producers of cookers and ovens), although the authors
emphasize again and again the partial and preliminary character of the results
therein presented. But, however preliminary they may be, they show that it
is not only a valid subject of research for the Bank, but also a worthwhile
one on which to intensify effort. The survey presents a good discussion of
the main characteristics of the SSEs, the origin of the firms and the entre-
preneurs, their technological adaptations, their problems vis-a-vis the labor
and financial markets, etc. But this said, it also shows very clearly that


- 65 -
Appendix I
Page 30
there are at least three areas where it appears that further thought is re-
quired. The first one is factor productivity, especially in regard to the
different combinations of capital and labor; the second is the overall ques-
tion of technology selection, utilization and adaptation, as well as the
learning process; and finally, the overall effect of economic policies on
the development of SSEs. The first two are frequently mentioned in the
Colombian paper and indeed, it is expected that they will be dealt with in
the definitive version of the study. The basis for the first point will be
a technology model for which data about factor productivity were collected
at the shopflcor level. The third point is an important one, and it is to
be expected that at a later stage its significance will be more explicitly
presented once research in this third area has progressed. It must also be
taken into consideration, just as is mentioned in other Bank papers, that
many of these distortions that affect SSEs have their origin in the economic
policies followed by many of these countries. (And it is clear that there
is a role for the Bank to play in ascribing the responsibility for many of
these distortions to where it +r rightly belongs.)
Lastly, in this section we will mention some of the problems that
have surfaced during the actual carrying out of the research. Here, how-
ever, only general impressions will be put forward, since it is difficult
to be more precise given the lack of more detailed information. Amongst
these impressions, perhaps the most important is the absence of some basic
methodology, such as was provided, for examole, in the case of the incentive
studies and others. Although it can be argued that the development of just
such a methodology is the result of an iterative process, it is important
that this is done in order to attain the fullest possible comparability of
the results of the different studies. Another and perhaps more relevant
observation at this stage, is that the number of participating researchers
seems to be relatively small, especially in comparison with the aims of the
project.
Credit Markets
In this area we find two RPOs , one of them (671-65) on Small Enter-
prise Financing, and the other (671-69) on Capital Market Imperfections and
Economic Development, both of them under the responsibility of the Public
and Private Finance Division. The discussion that follows also took into
consideration a paper prepared by this Division (Domestic Finance Studies
No. 43) presenting its research program. It must be pointed out that some
of the papers provided do not attain the quality levels that prevail in other
areas of Bank research.
RPO 671-65 has not, as yet, any relevant paper. We have, therefore,
only a research proposal to study the role of informal credit markets. The
project is justified on the basis of the importance of those informal credit
mechanisms for financing small enterprises and trade in various Asian and
African countries. It is suggested that the studies could serve to ascer-
tain whether the real cost of lending in this market is significantly lower
than in the formal financial market and if it is possible to establish some
sort of link between the two.


- 66 -
Appendix I
Page 31
A pilot s:ud. has been initiated that focuses on "a significant
element of the in-or=a credit markets in India, namely the Shroffis". (The
Shroffis are indigerous style bankers who extend and create credit through
the issue of hundis as well as other credit instruments). With due regard
to the focus of the st:dy, certain questions arise in connection with this
project for which adez:ate answers are so far missing, perhaps because of
lack of time to delve -ore thoroughly into these matters. Nevertheless, it
is useful to state a- least a few of the questions. One of them has to do
with the fact that we are left in the dark as to the significance of these
mechanisms with respe:t to financing and allocation of resources; the other
would concern the si:=iicance of these questions for the normative aspects
of the Bank's operations. Without meaning to cast doubts on the validity of
this research proposal - which in fact did not receive overwhelming support
in the Bank - the poi_: that we would like to make here is that special care
should be.exercised in formulating these "new" fields of research in the Bank;
otherwise they may be rapidly abandoned in favor of more "proved" lines.
RPO 671-69 on capital market imperfections and economy developments
owes much to the formulations of V.V. Bhatt and to previous work done in the
Division itself. The nroject was initiated in January 1978 and it is expected
that the final report will be finished in 1979. Various papers have already
been produced and published (see Appendix II).
In the development of the project, three types of interests are
clearly discernible. The first one is to examine the impact of transaction
costs on lending as well as on borrowing in different markets. This is a
subject in which there are obvious gaps in the literature and particularly
in relation to the financial systems of LDCs. The second relates to the
possible effects of banking regulations in LDCs on the banking structure,
and to the long-run financial viability of the banks. The third one con-
cerns the need to have a model that would express the possible interaction
between the development of capital markets and general economic development.
These will be briefly examined in reverse order.
The third subject of research is covered by two papers, one of which
presents a general model and the other attempts an empirical verification of
some of the issues raised in the first one; but the latter, in fact, is more
than just an extension of the former. The aim of both papers is to better
understand the role that financial intermediation plays in the development
process as well as to identify the factors that limit its effectiveness and
the benefits that may accrue from improving it. This was justified by the
valid observation that economic analysis in the past has generated "neither a
body of reasonably rigorous theory nor a wealth of credible economic results".
The central argument of the model, which does not as yet seem
adequately proven, is that for a given production function and state of
non-financial technology there is a role to be played by financial inter-
nediaries channeling resources from low to high return sectors, thereby
raising the aggregate marginal productivity of capital. The limits on this


- 67 -
Appendix I
Page 32
role have been defined at one end by the overall rate of return that would
prevail in the absence of financial institutions and, at the other, by the
rate that would prevail when capital markets are perfect.
Together with the presentation of the model, the papers include a
relatively thorough review of the existing literature and, as mentioned abo.e
an empirical verification of a limited version of the model. However, it "Z
obvious that further work will have to be done. One obvious need concerns :e
nature of the data and also further examination of the results provided b-,-
model. A possible approach could be to choose countries not only on the     i
of geographical considerations but also on account of their level of deve!::-
ment and the structure of their economies.
Another paper examines the relationship between earnings and port-
folio of the commercial banks of three Asian countries: The Philippines,
Thailand and Singapore. The technique used was that of statistical cost
accounting and followed, with some variations, the methodology previously used
by Hester and Zoellner. As in other cases the research was also justified on
the grounds that analyses of this type are scant in developing countries.
The investigation is well conducted and the conclusions are interest-
ing in that both the rate of return on loans and administrative costs see- to
be positively correlated with the degree of regulation in banking operations.
The lowest figures relate to Singapore, which, of the countries examined,  s
the one that presents the most competitive financial market and is the least
regulated. This is an analysis that undoubtedly can be further refined and
extended to a larger number of countries, although we should mention that data
problems in several cases will prove to be quite significant.
The paper on the relations between interest rate and transaction
costs is an interesting one and developed with technical competence. It
raises a valid point in calling attention to the importance of transaction
costs for the borrower and the lender and thus, in a certain sense, to the
need to innovate in the process of financial intermediation. However, the
significance of this factor alone is not clear yet. The study previously
referred to on banking in The Philippines, Thailand and Singapore estimated
the administrative costs of lending on the average at 0.1-0.5 percent, 1.4-
2.0 percent and 0.77 percent, respectively. It seems, therefore, that there
are other more important factors to be taken into consideration than only
the size of the borrower.
Public Enterprises
RPO project 671-11 (1anagerial Structures and Practices: Public
Manufacturing Enterprises) is up to now the only research project by the Bank
in this area. It was started in March 1978 and it is expected that the final
report will be submitted in January 1980. In a sense, it is being considered
as an extension of some previous work and in particular of reconnaissance
missions that were previously made to Egypt, India and Yugoslavia. As yet,
the only report available is a case study of a public enterprise: India's
Swaraj Tractor.


68
Appendix I
Page 33
The research proposes to examine a set of hypotheses of which it is
said, "are best described by the general proposition that the performance of
public manufacturing enterprises is dependent upon their organizational and
managerial structures and practices as well as the policy environment within
which they operate". It is justified on the basis that a large part of the
lending operations of the Bank concerns public enterprises and that they are a
growing sector in LDCs industries. In addition, they are a type of enterprise
about which not much is known and where in general, little research has been
done.
As mentioned before, the only report available up to now is on the
Swaraj tractor plant in India. The report is interesting, particularly where
it shows the significance of indigenous technical developments and adaptations
in the shaping of and the improvements in the competitive position of this
plant. Nevertheless, it does not allow one either to evaluate the overall
approach taken by the Bank on this question or to arrive at definite conclu-
sions.
In view of this it might appear convenient for the management of
the Bank to give more reflection to the research work done in this area. The
area itself is of prime importance for the Bank from the points of view of
(i) allocation of resources, since a large part of its loans goes to public
enterprises and (ii) its advisory role to LDCs. In the absence of a coherent
body of thought on these questions there is a great risk for dispersion of the
research efforts. This would be unfortunate since it appears to us :hat cn
account of its experience and the insights that could be provided by research
executed in other areas, it is difficult to think, worldwide, of any institu-
tion other than the Bank that could do it.
General Evaluation
It is quite clear that the subjects pertaining to small enterprises,
credit markets, and public enterprise, have great relevance for Bank operations
and for LDCs. However, it is also clear that here we have a case where
relevance is not matched by the research effort actually carried out.
There are several factors that could account for this situation.
One of them, perhaps, is the fact that in the past the Bank thought that the
best way to approach these problems was on an ad hoc basis through the opera-
tional units and, thus, that there was no need to support their actions with a
research basis. This is what would account for the additional fact that all
these areas of research are relatively "new" within the Bank. Indeed, one
could go even further and argue that there was some sort of resistance towards
including these and other types of institutional research within the body of
research carried out in the Bank.
Even now, one can discern the prevalence of two types of attitudes.
On one hand, there is the attitude of considering research on institutional
questions as "interesting" but devoid of academic rigorousness and, on the
Other, the attitude towards institutional matters of some researchers who
tend to consider other research by the Bank as frequently too academic and of


- 69 -
Apendix I
Page 34
little relevance either for the Bank or, worse, for the real world. This
is by no means a unique situation in research institutions, but there is
no doubt that, as has been proven by the Bank itself, there is a lot to be
gained through a better coordination of efforts.
All of this, obviously, is not something that can be attained
overnight. But it is probable that a large par: of the problem could be
solved if the Bank could attract high caliber researchers with solid re,u-
tations. It has done successfully in the case of small- and medium-scale
enterprises and there is no reason why in credit markets and public enter-
prises it should not also have the kind of intellectual leadership that
prevails in other areas of research. Such leadership would provide the
Bank with a better analytical framework to enhance the coherence and sense
of direction of research as well as to establish a better working relation-
ship with researchers in other parts of the Bank and elsewhere. In the
absence of this, we think that there may indeed be a great risk of either
a dispersion of efforts, or of conducting research that could be of little
relevance for the Bank and for policy fornulation.
One other aspect that we would also like to touch upon is the ques-
tion of how normative this type of research should be. In our opinion it
ought to be high. Through its research the Bank should be able to discern
which types of policies and programs for  evelopnt are successful an-d under
what conditions, and likewise which are unsuccessful and why. This is a fre-
quently mentioned objective in the research proposals, but does not come
across clearly enough in many of the papers under review.   Lany of the RPO
projects are still at a preliminary stage and therefore it may be difficult,
at this point, to extract this type of conclusion. Nevertheless, this is an
area where the Bank would be utilizing, to the fullest, one of its compara-
tive advantages in research vis-a-vis other institutions.
To conclude this section, we will touch briefly upon the question of
basic research and on the degree to which research by the Bank has duplicated
that of other institutions. With regard to the first part of the question,
the answer is that there has been some basic research, especially in the area
of small enterprises and on the role of financial development in the develop-
ment process. But our recommendation here would be in favor of more, rather
than less, basic research. There are a number of aspects on which the existing
knowledge is imprecise, especially when dynamic factors are taken into consid-
eration. As to the second part of the question, it simply does not appear
that the Bank has duplicated work by other institutions. In practically all
RPOs a point has been made to undertake overall reviews of the literature on
the different subjects. This has assured, to a certain extent, that research
duplication is avoided. To be more precise, in many cases the research done
by the Bank on these topics has been original and pathbreaking. This is the
case, for example, in research on SSEs, and there are obvious possibilities
of doing likewise in connection with public enterprises.


- 70 -
Appendix I
Page 35
As to the impact of research on LDCs, the questions are somewhat
more difficult to answer. We would like to start by pointing out that in all
RPOs under review there is participation by researchers from LDCs, though in
some cases more than in others. As to the influence that this may have upon
LDCs' policy formulation and relations with the Bank, it is difficult at this
time to define what their precise character will be in the future. But the
Bank has alredy quite substantial experience in this regard and we think that
in this case it will not differ from former experience; i.e., that if there is
substantial interest in the problem under consideration, many of the recommen-
dations filter through to the policy decision-making bodies. There is in many
countries a real need to know more and to receive better advice on most of the
subjects covered by this chapter. And this holds true for the Bank itself.
We were impressed in our conversations with Bank staff and in reading Bank
memoranda by the difficulties that many Bank missions have in answering
questions that are frequently posed to them by government officials, such as
what would be a sensible policy vis-a-vis public enterprises in SSE-develop-
ment programs, and for which, up to now, only very vague or general answers
could be provided.
We come now to the question of allocation of research resources.
Here we can make only two suggestions. First, within the area itself, our
impression is that it would have been preferable to concentrate more resources
on SSEs and public enterprises than on SSEs and credit markets, as has been
the case. To us the former relationship appears to be more significant from
the point of view of the development process, policy formulation and Bank
operations. In studies of SSEs one cannot afford to ignore the fact that
small enterprises often face discrimination in the capital markets. Second,
and perhaps more important, there is the problematic issue of the financial
resources that have been allocated for research in these areas. There is a
need to increase them and, in particular, to obtain the concourse and col-
laboration of highly qualified people in all cases, but more particularly
in that of credit markets and public enterprises.
Recommendations Regarding Further Research
The work done by the Bank on capital utilization, capital labor
substitution, SSEs and on technical change itself has provided evidence of
the need to undertake more thorough studies of the role that technical change
plays. in economic development. This, by the way, coincides very much with a
growing preoccupation in many LDCs with the relation between technological
development and overall economic development. This preoccupation has to do
with a number of factors. In the past, for example, it had much to do with
the question of negotiating international transfers of technology, but now-
adays it is clear that this is only one of the aspects involved and that
perhaps more important are the aspects concerning the selection, adaptation
and absorption of technology.
It appears that the area of public enterprise has not been suffic-
iently studied in the Bank or, for that matter, anywhere else. The recommen-
dation was made above for the Bank to attempt what would amount to a fresh


- 71 -
Appendix I
Page 36
approach to research in this area. There are many questions here for which
answers are, at best, only fragmentary. Amongst them, we should like to point
out the relationships between public enterprises and the public sector, or
with the private sector of the economy, or their role in technological devel-
opment, etc. It is, to a certain extent, an area not fully explored, though
its importance for economic policy formulations and for the Bank's operations
is well recognized.


- 72 -
Appendix I
Page 37
1:5 Capital Utilization, Capital Labor Substitution, and Technological Change
Richard Nelson
The Subject
The body of bank research reviewed here is concerned with the
connected topics of the range of choice of technique, factors influencing the
technique chosen, efficiency in the use of a technique, adaptation of tech-
niques, and technical progress in industry.
The policy thrust of the research has been provided by the belief
that in less developed countries labor is cheap and capital expensive relative
to developed ones, that this ought to be reflected in the use of more labor
intensive techniques, but that while this was occurring to some degree it
was desirable that the techniques employed be yet more frugal on the use of
capital. Throughout, the research has been concerned with market and other
forces that explain the prevailing situation, and with policies that could
improve the environment so that choice of technique would be made more appro-
priately, techniques used more efficiently, and adaptation and learning pro-
ceed more effectively. The analysis has been concerned both with contexts
in which decision making with respect to technique is made largely by private
profit-oriented entrepreneurs, and with situations in which the government is
the decision maker or has a great deal of direct control over private deci-
sions. The specific topics studied have evolved over the years as work done
under one project generated insights and puzzles that influenced work under a
subsequent project. The overall impression is that of a cumulative research
program that not only makes a significant contribution to thinking in the Bank
about industrialization problems, but also influences thinking within the
development economics community.
Project Review
The specific RPO projects reviewed here are 670-23, 670-25, 670-54,
670-95, and 671-51. Three of these projects now are completed, and have re-
ceived earlier reviews. The Bank research on industrial capacity utilization
(670-25 and 670-95), both projects now completed, was concerned with assessing
empirically casual observations that despite the scarcity of capital in less
developed countries, the existing capital was not being used very intensively.
If the proposition held up under such scrutiny, the project aimed to contribute
to understanding the phenomena and to finding ways to improve the situation.
The basic ideas originated outside of the Bank, and scholars had been working
on the problem before the Bank got seriously into the study of capacity util-
ization. But the Bank work was of considerable scope, and contributed to
the growing appreciation that indeed in many cases capital was not being
worked very intensively in LDC industry. Bank research identified a number
of factors, associated with imperfect input markets and various legal and
institutional restrictions, which seemed to lie behind the non-economical use
of plant and equipment. On the whole, however, not much was added to existing
knowledge in this field. Project 670-95, which was completed largely by or


- 73 -
Appendix I
Page 38
under the close supervision of bank employees, was more successful than
;70-25, which was farmed out to a consultant and which suffers from some
methodological weaknesses. Neither of these two Drojects seemed to have
influenced Bank operations.
The project on employment and capital labor substitution (670-54),
now also completed, was related to the work on capacity utilization in that
this project also was concerned with examining ways in which capital could be
used more frugally. The research under the project did not meet the stated
obJectives and treated important questions improperly. Hcwever, the cuestions
considered under the project still are of: orme importance to the Bank, and
:heir exploration continues under another roject, 670-23.
The two major RPO projects currently under way at the Bank are the
above mentioned 670-23 "Scope for Caoital-Labor Substitution in t1e Mec-anical
Engineering Industry", and 671-51, Appro7riate Industrial Technology. 'The
:ormer project has been under way for much longer than the latter, more re-
sources have been invested in it, and the output is much more impressive.
The research task taken by the current project on capital labor sub-
stitution is to explore in microeconomic detail the nature of the production
processes involved in particular areas o: nanufacturing, and to examine the
range of capital-labor substitution availa'be in each. The project has had
:o face methodological, empirical, and theoretical questions of conSiderable
difficulty. The struggle has been valianz, and the projeut has providcedi
analysis of production processes and the range of choice that is more detailed
and more sophisticated than other work probing similar questions. Other stud-
ies under the project have explored the choices of technique actually made by
firms, and the market and other institutional factors that have influenced
those choices (which in a number of cases have been much more capital inten-
sive than would make sense from an economist's point of view). The project
appears to be having considerable impact within the Bank, and while academics
tend to lag in the development of their aporeciation of work of this sort,
the importance of the project is beginning to be recognized by the academic
development economics community.
The project on Appropriate Industrial Technology is behind schedule.
The project has two thrusts: measuring the gains to less developed countries
from adopting more appropriate technologies, and examining the capital goods
sector in less developed countries as a possible major determinant of the
availability of appropriate technologies. A few illustrative numbers have
been put together on the first issue, and some literature review, casual
empiricism, and thought have been directed at the second. The work, particu-
larly on the domestic capital goods industries, shows promise, but it is too
early to tell how successful the endeavor will be.
Projects 671-51 and 670-23 complement and support each other. The
relevance to good policy making of better understanding of choice, or (more
sharply) understanding why more capital-using techniques are chosen when less
capital-using ones are available, is obvious. The more recent reports on both
Projects reveal the authors' growing concern about innovation, as contrasted


-74
Appendix I
Page 39
with choice among "obvious and available" alternatives, and a complementary
interest in institutional structure. These intellectual developments will be
discussed later.
General Evaluation
The work reviewed above is of variable quality and merits a high
rating only in the case of RPO 670-23. The worst project was RPO 670-54,
whereas the others occupy a middle category. The better pieces of work have
no doubt contributed, and are contributing, to knowledge. In this regard, the
Bank has exploited its comparative advantage, has avoided duplicating work
that was being done elsewhere, and has proceeded in good awareness of that
work. This research certainly has influenced the thinking within at least
some parts of the Bank regarding the importance of the capital-labor substitu-
tion problem and the scope of possible substitution. In some instances,
researchers have been involved in operational support missions to countries
such as Argentina and Korea, evaluating technology issues. At the present
time efforts are being made to implement this appreciation in project decision
making. The work seems to have influenced thinking regarding both the impor-
tance of having the right prices, and the importance of having institutions
that channel funds and information effectively. Particularly the empirical
work has involved researchers in less developed countries. By and large,
the work has been led by people full-time at the Bank. Where outsiders or
part-time consultants led, the results were much less satisfactory.
Recommendations Regarding Further Research
We think that Bank research exploring and documenting the range of
possible capital-labor substitutions in particular technologies is now running
into diminishing returns. The general point that there is a wide range of
choice is now well substantiated. The techniques involved in identifying the
relevant substitution possibilities in particular technologies may now be
ready for practical application. We note that practical application almost
certainly will involve less detailed breaking down of processes than was
relevant when the endeavor was viewed as a research study. We think this
research endeavor should be phased down and effort placed in helping the
operative divisions and departments to be able to do the relevant analyses
themselves.
The project on appropriate technology is being conducted at a modest
level, and in an exploratory manner. We think that particularly the part of
the project exploring the design capabilities of domestic capital goods pro-
ducers has considerable promise. Research within the Bank, and elsewhere,
is revealing that there does indeed exist this design capability within the
capital goods industry of at least some countries. This project is a serious
candidate for more resources.
The Bank's research under the current two projects increasingly is
highlighting that effective capital-labor substitution, the adoption of appro-
Priate technology more generally, and over the long run the development of an
efficient industrial sector, are not adequately described merely in terms of


- 75 -
Appendix I
Page 40
"picking things off the shelf". The research has documented the wide range of
catalog items available. But it also has documented that it is no trivial
matter for a firm to be aware of more than a small portion of the range of
choice. Increasingly the research is showing that at the least adaptation and
in many cases innovation is involved in effective choice of technique.
Recognition of the importance of adaptation and innovation calls attention to
the entrepreneurial dimension in business leadership, and to the institutional
structures that encourage, suDport, constrain, and deter entrepreneurship.
We detect an ambivalence on the part of the Bank with regard to its
research comitments where their own research findings indicate they ought to
be put. The tradition of economic research at the Bank has stressed neat,
quantitative, formally specified models; there has been a reluctance to delve
into areas of economic research where precise models and econometric technique
cannot serve as the primary tools. And in fact the batting average in proj-
ects where there was little analytic structure has not been high. Issues of
innovation, and institutional structure tend to be viewed as "unresearchable"
or at least not amenable to rigorous research. But we propose that the logic
of the Bank's own past research endeavors has led it inexorably to a require-
.ent to engage in this kind of research. It already is engaging in such work,
with studies on the capital goods sector in selected LDCs, on small and medium
scale enterprises in Colombia, and on organizational and institutional factors
in Colombia's garment industry exports. But there is still apparenz m:nod-
ological uncertainty. However, the methodological situation is not as bad
as some people in the Bank may think. Over the last decade both the interest
and the rigor of research on industrial organization has increased greatly. A
considerable body of good and rigorous research on the economics of technical
change has evolved. However, for the most part, research in these fields has
been focused on issues and phenomena in advanced countries, not developing
ones. We think the Bank should accept the obvious challenge.
It may be useful to give several examples of the kind of work on
industrial organization and technical change that might serve as models for
research at the Bank. Over the past half dozen years a considerable amount of
research has been done, principally by American economists, aimed at exploring
the costs of government regulation. A good reasonable overview of much of
that work is provided in A. Phillips (ed) Promoting Competition in Regulated
Markets. While we are nervous that much of the research by American economists
on regulation has stressed the costs and downplayed the benefits, the identi-
fication and attempted quantification of the costs has been illuminating, and
such work has had a significant effect on policy.
The work of Edwin Mansfield provides good examples of quantitative
studies on R & D, technical advance, and productivity growth. A good collec-
tion of Mansfield's earlier work is contained in his Industrial Research and
Technological Innovation. A recent study of his (and his students) in which
he attempted to measure both private and social rates of return on industrial
innovation was published as "Social and Private Rates of Return on Industrial
Innovation", OJE, 1977. There also has been a very good work on the economics
of technical change in industry undertaken by Christopher Freeman at the


- 76 -
Appendix I
Page 41
University of Sussex, often in collaboration with other members of Science
policy Research Unit there. A portion of that work is presented, in capsule
form, in Freeman's book The Economics of Industrial Innovation.
There has been some work on technical change in less developed
countries, but not much. There are several interesting studies focused on
agriculture. Hayami and Ruttan's book, Agricultural Innovation, is a useful
reference. At present, Jorge Katz is directing a study on innovation in Latin
American industry under the joint auspices of the IDB and ECLA. These studies
represent a start. We think the time is ripe for the Bank to get into the
field in a big way.


- 77 -
Appendix I
Page 42
1:6 Programming in the Manufacturing Sector
Kirit S. Parikh
1. The Nature and Importance of the Research
A number of research reports and monographs are expected from the
research program carried out under the heading "Programming in the Manufactur-
ing Sector" (RPO 670-24). The list of reports/draft reports which were
studied for this review is given in Appendix II. These constitute only a part
of the voluminous output of the research program. The program has dealt with
the problems of investment planning in industries characterized by increasing
returns to scale and in industries where interdependence in the production of
different products is important. Interdependence may be important when
different products share capital equipment or when they use the same interme-
diate inputs, the manufacture of which may exhibit economies of scale. The
Bank's research in this area has focused on development of improved methods for
selecting investment projects from among the many alternatives in size,
timing, location, technology and output mix. In addition, it has investigated
the extent to which such interdependence affects project selection and planning
for the develo;pment of a sector and offers scope for cooperation among the
countries of a region.
The importance of externalities resulting from economies of scale
and interdependence of various kinds has been widely recognized in the lit-
erature on development economics. In fact, the "big push" theory of devel-
oDment is based on the recognition of such interdependence and indivisibil-
ities. Where economies of scale are predominant, a developing country's
market may be inadequate for the economic scale of the plant. The make/buy
choice would always seem to go against domestic production when products are
considered in isolation as they would be in a cost- benefit analysis within
a project-appraisal framework. The process of development could then hardly
begin. On the other hand when a large number of interdependent projects are
examined together, the market size for the products of some of the projects
could increase sufficiently to justify domestic production. Thus the research
program under review deals with issues of great importance for development
policy and planning and particularly for sector planning.
It is also a part of the mythology of development planning, that
there exists a shelf full of project reports and that the planners' task is to
select a subset of these projects. In fact, it is hardly the case that such
a shelf full of project reports is available. Detailed project reports are
expensive to prepare, particularly in terms of the skilled manpower which is
almost always scarce. Moreover, when the project-executing authority (or
Hinistry) is different from the project-sanctioning body (such as a Finance
Ministry of Planning Commission), a lot of vested interest gets created in a
project by the time a detailed project report is prepared. With the pressures


- 78 -
Appendix I
Page 43
from such vested interests it becomes difficult to have an objective evalua-
tion of projects. Such difficulties can be minimized if projects are identi-
fied at an early stage. Thus the development of a methodology that helps in
pre-selection of projects for further detailed investigation is of great
practical significance.
Moreover, before a detailed project report is prepared the techniques
have to be selected. To the extent that the choice of techniques itself is
affected by interdependences, the choice should be made in a wider context.
This would be the case when different products share the same capital equip-
ment. Besides its possible impact on the choice of technique, such capacity
sharing would also affect the economic scale of production. Though the
qualitative effects of economies of scale, indivisibilities and interdepen-
dences have been theoretically well recognized, systematic, empirical and
quantitative evaluations have been rare. The set of studies carried out under
the RPO 670-24 has explored these issues with commendable thoroughness.
The research has been carried out within the context of specific
investment planning problems in two sets of empirical studies, one set deal-
ing with what are termed "process industries" and the other with "non-process
industries". Those industries which are characterized by a manufacturing
process stream which is more or less continuous, which have a limited number
of processes and where the cost of carrying midstream intermediate products
is large are termed process industries. Examples of such industries are gas
transmission, fertilizers, cement, etc. These industries also have a limited
number of products which are more or less uniform.
The "mechanical-engineering sector" on the other hand has a variety
of products and processes and the same processing equipment can be used for
manufacturing many different products. Such industries are termed "non-
process industries".
Both these sets of studies have used mixed integer programming
models in a fixed charge formulation to account for economies of scale. The
major problem that has been faced in previous studies, apart from the consider-
able time and efforts that usually go into data collection and organization,
has been the problem of obtaining solutions of the mixed integer programming
models with a large number of integer variables. Solutions to such problems
require a large amount of computer time. The research program under review
has developed a number of procedures to eliminate, through simple but sophisti-
cated analysis, a number of integer variables which represented uneconomical
choices, to reduce the size of the programming model. This makes obtaining
solutions to such problems practicable.
Important Results
Apart from the specific sector development plans that emerge from
these sector studies, they have also provided some insights into the nature
of technology and its consequences.
U


- 79 -
Appendix I
Page 44
For the process industries that have been studied:
a) Significant economies of scale are present in production
activities and there is a good deal of potential interde-
pendence within the system as a whole.
b) The use of programming models help in evaluating the conse-
quences of alternative policies. The cost or benefits of
particular policies may be significant.
c) Programming models provide a tool to estimate the benefits of
regional cooperation to individual countries and help in
designing schemes for sharing the benefits.
For mechanical engineering, the only non-process industry studied:
a) The relative cost of complete neglect of interdependence in
choosing between production and imports is sometimes not signi-
ficant at the sector level. For the part of the mechanical-
engineering sector of Korea that was studied (120 carefully
selected products), this would have led to an increase in sector-
wide total supply cost of no more than 3 percent of the total
value added for the products involved in the study.
b) Of the 120 products involved in the study, only a few were
"critical" in the sense that for the others the make/buy choices
were clear and unambiguous for the relevant ranges of parameters.
Simple programming exercises are adequate to identify the criti-
cal products. As a percentage of value added in these critical
products, the savings in cost that result when interdependence
is accounted for are much larger and are found to be more than
25 percent in one variant. For individual products the savings
over cost of imports in some cases exceed 40 percent.
c) Though the loss at the sector level is small for particular
products, the conventional cost-benefit analysis, which neg-
lects interdependence, may lead to wrong make/buy choices
when the products are part of a sector that exhibits inter-
dependence.
d) The absolute cost of neglect of interdependence is "by no
means trivial" and is "far more" than the cost of conducting
studies that take interdependence into account.
From a methodological point of view the most significant contribution
of the research is the demonstration of the use of large mixed-integer program-
ming models. In particular, the following results have been demonstrated:


80
Appendix I
Page 45
a) Even without obtaining a globally optimum solution, use of
programming models can provide a lot of insight into the nature
and costs of the various alternatives.
b) Problems with a fairly large number of integer variables
can be solved with reasonable costs.
c) With a systemtic exploration of break-even analysis a
number of useful decision rules can be employed to eliminate
a significant number of integer variables.
Review of Studies
Process Industries. The studies carried out for the planning
of the fertilizer sector in Egypt and in East Africa have explored the choices
of technology, size, location, transport, product and trade. The East African
study has in addition explored in quantitative terms the gains from cooperation
in fertilizer sector development for the three countries of the region (Uganda,
Kenya and Tanzania).
The formulation of the models for the fertilizer sector is straight-
forward in mixed-integer programming terms, though the inclusion of substitu-
tion among products marks useful innovation. However, because of the shortcuts,
the models have been considerably more detailed, and consequently, operation-
ally more meaningful.
An obvious and recognized limitation of the models as developed is
their neglect of uncertainty. Accounting explicitly for uncertainty would
make the computation problem even more formidable. Therefore, the researchers
stress the use of "pre-analysis" and sensitivity analysis to account for uncer-
tainty.
A number of other studies have also been carried out for different
sectors, countries and regions. These include among others, fertilizers for
ASEAN countries, the Andean pact countries, and for India; the forestry
sector for Turkey; paper and pulp for countries of the ASEAN Region and FAO's
World Program; clinker production for Brazil, and energy for Nigeria. Other
studies are under way.
An attempt to use this approach to develop a model to quantify the
benefits of regional integration based on a simultaneous analysis of a number
of industrial sectors, the pros and cons of which were succinctly presented in
a small paper, was abandoned as being too ambitious.
The project has obviously been successfully carried out and to
judge from the number and variety of applications that have been made and are
planned it has been found useful too. Some time ago, the project reached a
stage where research ended and applications began. But further effort is
required to successfully disseminate and transfer the methodology for use in
practical applications. A most important element in facilitating such transfer
would be the development of computer software, which makes it convenient to
specify the problem preferably in the language of the users as opposed to the
language of the computer specialists. In addition, it would also be essential


- 81 -
Appendix I
Page 46
to generalize and automatize the break-even analysis, as otherwise applications
would need not only trained but also clever people, and would be severely lim-
ited in scope. The recently initiated GAAS project (RPO 671-58) should go far
towards providing the necessary software.
Non-process Industries - Study of the Korean Mechanical-Engineering
Sector. The study of the Korean Mechanical-Engineering Sector has explored
the gains from planning simultaneously the supply of a large number of prod-
ucCs. As already said, 120 carefully selected items were analyzed. The gains
in economic efficiency of domestic production are derived from the selection
of technique and scale of production, taking into account the possibilities of
sharing capital equipment for a variety of products, as well as the possibil-
ities of domestically producing on a large scale intermediate goods used in
a number of products. The import or domestic production decisions are taken
after considering the effects of such sector-wide interdependence.
The mechanical-engineering sector is not easy to model in the
conventional way. Problems of appropriate description and specification of
products and processes have to be faced, and such problems are not insignifi-
cant. In applying a micro-analytic approach to a sector-wide study, there
is a danger of getting lost in details and not seeing the wood for the trees.
This is avoided in the study, by describing certain standardized products
-hose production processes are described at the shop level rather than at the
machine level. Even then, the breadth and depth of the technical-engineering
decail that is incorporated in the study is to be found in hardly any other
programming study.
The model used here is a mixed-integer programming one, and break-
even analysis is developed and used to reduce computational difficulties.
In addition, the allocative consequences of the results are explored
in depth. An evaluation of alternative investment criteria is also made in
the context of the results obtained from the model. The conclusion is reached
that simple cost-benefit criteria, or the measures of comparative advantage
when substituted for the more thorough analysis of the type undertaken in the
study, lead to wrong make/buy decisions for some of the products with conse-
quent loss to the economy. It is shown that the estimated loss entailed in
using the best cost-benefit criteria is found to be large enough to pay for
the cost of more comprehensive analysis within one to two years. However, it
may be noted that in order to fully realize the benefits from a project which
becomes economically attractive only when interdependences are accounted for,
all the interdependent projects have to be executed together. A delay in one
of them would adversely affect not only its own economics but also that of the
others.
Moreover, if an integrated sectoral planning procedure leads to,
which it need not, additional delays in clearing projects, the benefits of
such planning to some extent become negated. Yet this is not to argue for the
exclusive use of simple social cost-benefit analysis. It is merely to under-
line the context in which sectoral programing can be fruitfully applied.


- 82 -
Appendix I
Page 47
Since such programming earns a handsome rate of return, there is no reason not
to pursue it.
The project also complements a number of other research studies
where attention to micro-level details may be crucial. The research programs
on The Scope for Capital-Labor Substitution in the Mechanical-Engineering
Sector (RPO 670-23) and on Appropriate Industrial Technology (RPO 671-51)
are such projects (see A.5).
Evaluation
Quality of Research. The research work is certainly of a high
caliber and at the frontier of work in this area. Moreover, such research
is hardly done outside the bank. The volume of research output is also high.
Usefulness for LDCs. Large programming models create an impression
that the vision behind the process of development that motivates such studies
is one in which an elite all-knowing planning authority attains economic
growth through effectively allocating resources to various sectors. Yet one
need not share this vision before considering such models to be useful. Pro-
cess industries such as fertilizers, cement, etc., characterized by economies
of scale and a relatively small number of plants, are the industries whose
development is usually guided and promoted by most governments of developing
countries. Starting a few large industrial projects is one of the easiest
things that governments do to promote development. The planning models
developed by Bank research have the potential to improve the rationality
of government decisions in developing these sectors. Some of this potential
is already realized in the numerous applications made for different sectors
and different countries.
However, the full potential usefulness can be realized only if
adequate "extension work" follows this Bank research. We shall return to
this later.
Usefulness for Bank. Clearly, the studies related to specific
areas must have been carried out in collaboration with the operational staff
of the Regional Department concerned. The results should have been useful in
guiding the Bank's lending operations, provided they were available in time.
But clearly a lot of potential is there for such work to be useful in Bank
activities.
Improvement in project identification methodology in the LDCs could
be of considerable significance for the Bank's operations. If more attractive
Projects are identified before the detailed project reports are prepared,
these reports would be made for comparatively better projects. This would
imply that better projects are put up to the Bank for financing.
Part of the work on the fertilizer sector plan for East Africa
Might have been made irrelevant by the subsequent political developments


- 83 -
Appendix I
Page 48
in East Africa and the breakup of the economic union. Even then, the non-
cooperation solutions could still have been useful to the policy makers in
the three countries.
The Bank may have a unique comparative advantage in carrying out
studies, such as the fertilizer study for East Africa, that identify areas
for regional cooperation and facilitate the process of realizing such coop-
eration. As an authority that lends money to the various countries of a
region, it may have access to data and policy makers in these countries.
Moreover, as a third party its analysis may be less suspect. On the other
hand, the Bank should also be interested in promoting such cooperation that
reduces the need for credit in the region.
Development of Research Capacity in LDCs: Creating research capa-
bility is a time-consuming task and learning by doing is an essential element
in the development of research skills. Significant participation of research-
ers from the LDCs is an absolute requisite for the successful transfer of
the research to applicaticn in the LDCs. However, time constraints hamper
research programs; communication problems across large distances and lack
of access to computers and xerox machines in many LDCs make such participa-
tion difficult. The country specific-sector studies undertaken to date have
involved participation of local persons. 'ut an effective program has ya:
to be designed to facilitate participation of researchers from the LDCs, to
ensure that there would be established within the country the capability to
either update and/or improve the particular sector study or to carry out a
similar study for another sector.
Suggestions for Research in Future
a) Extension: The research results and the methodology developed are
sufficiently important that they be brought to practitioners, planners and
policy makers in the LDCs. This would call for a considerable amount of
"extension" work. The manuals currently under preparation are vital to
facilitate extension, but their availability constitutes only one element
in the process.
Similarly, short training courses would also be inadequate by them-
selves. What would be required is a case study for a sector, which is carried
out with active participation of a local team or, better still, a local insti-
tution. It may even be desirable that thestudies be carried out by a local
team. Such studies may be coordinated by Bank staff who are familiar with
such research work, and should certainly be financed by the Bank. Even when
the success rate of such research is not high, it may be considered a neces-
sary investment in building up research capability in LDCs. Such support
should also include provision of computer hardware in the case of many LDCs.
A sector study based on mixed-integer models needs convenient and substan-
tial access to fairly large computers if the study is to be completed within
a reasonable time.


-84-
Appendix I
Page 49
Development of user-oriented software that permits convenient
specification of the problem by users, who may not be computer specialists,
could be of invaluable help in promoting such studies in the LDCs.
The emphasis on extension is motivated by a vital consideration.
An important benefit of any modelling accrues largely to the modeller in the
form of sharper insights and improved understanding of the processes being
modelled and their interdependence. The major benefit to be obtained from
using a programming approach to the planning of an industry is to be derived
from training that would be imparted, in following such an approach, to a set
of engineers and economists to think more systematically about the scope of
technical options in developing the sector. The engineers would be less
certain of their thumb rules and the economists would be more aware of the
technological constraints. The information and insights generated could
substantially improve the quality of decision making in the sector. Unless
these insights accrue to the people who will continue to take decisions in the
LDCs, the benefit would not be maintained.
The foregoing is not to deny the usefulness of attempts at extension
through previous case studies, but to recommend that more resources be devoted
to extension, and that the case studies have as their primary objective effec-
tive extension.
The decision, which is reported to have been taken to set up a
special group to promote case studies, is to be strongly endorsed. To begin
with, two to three full-time professions, capable of guiding projects in a
number of different areas and countries, should be provided for in this group.
b)   Consequences of Interdependence for Planning and Promoting Industrial
Development:  If economies of scale and interdependence are important, then the
development of that industry would benefit from some form of central planning .
or coordination. On the other hand, effective implementation and central plan-
ning of a sector such as the mechanical-engineering industry, characterized by
a large number of products and processes, is difficult, to say the least.
Thus the findings of the Korean Mechanical-Engineering Industry
study, that interdependence appears not to be of much quantitative consequence,
are very comforting as one can rely on implementable, decentralized procedures
for developing the industry. Yet, the authors of the Korean study emphasize
that the finding needs to be confirmed with further research. Moreover, as
they point out, there is reason to suspect that the aggregation approach used
may bias their findings. Would a different set of international prices, such
as may prevail at other times, lead to a different result? Would different
domestic prices, such as may be found in another country, give a different
result?  Would a different product-mix produce contrary results?  Is it pos-
sible that under such different circumstances the benefits of accounting for
interdependence would be highly significant, say 50 percent or more of the
value added in the sector?


- 85 -
Appendix I
Page 50
The quantitative dimension of the effects of interdependence is
sufficiently important for policy purposes that further explorations to test
the generalizability of the Xorean Mechanical-Engineering study are called
for. This has been the only study that has investigated these issues empiri-
cally, and after having made the investment in developing conceptual frame-
works and methodologies for such studies, it would be desirable for the Bank
to pursue this research to its logical conclusion. It would, according to
the reviewer of this section, be a grave loss if the Bank were to stop after
studying just one sector in one country.
c)   Programming models to evaluate Appropriate Technology for Rural
industrial Development: The models dealt with in this research can be used to
throw some light on issues of considerable interest in LDCs. To what extent
should one develop a decentralized industrial structure? What are the costs
of a rural-based industrial development? What is the implication for employ-
ment and income generation of development based on small-scale industries?
Are the benefits of dispersal sufficient to compensate for the extra cost of
setting up smaller industrial units? These issues need to be explored in a
systematic, technical and dispassionate way. The case study of pulp and paper
in Malaysia examines some of these issues, while the follow-up work under RPO
670-23 has a great deal of relevance to understanding questions concerned with
the organization of production.
Nonetheless, with the foregoing exceptions, the studies in industrial
programming under RPO 670-24 have tended to view the problems of development
in a purely technical way. The objective has been to find least-cost solu-
tions. Institutional issues in the organization of sectors, in the difficul-
ties of implementation, or in the realm of selection of policy instruments
should be brought to the fore in subsequent research.
In evaluating the appropriateness or otherwise of technology,
not only relative factor scarcities should be taken care of but also the
limitations of public policy in using certain instruments. Thus if income
redistribution policies are politically hard to pursue, one might put emphasis
on income generation in selecting "appropriate" techniques.
d)   Institutional Issues In Implementation and Capacity Utilization:
Though the bank research in this area of industrial programming has been
useful and has indicated benefit in excess of costs, one may still ask if
there are other issues that need to be researched for promoting industrial
development.
The problem of implementation of projects and the efficiency with
which even large industrial projects are operated in many LDCs are of great
importance. Mixed-integer programming models can be used to examine the
consequences of different levels of operational efficiency in existing plants,
as was done in the Egyptian fertilizer case study. Nonetheless, the issues
here go much deeper. To what extent are delays in installation of capacity
and inefficient use of installed capacity in large industrial projects due


- 86 -
Appendix I
Page 51
to improper organization, due to inappropriate objectives of the management,
or due to inexperience and inadequate skills? Are delays and inefficiencies
inherent in the organization of the public sector, which plays a large part in
the development of industrial sectors in some countries? Does the private
sector really perform better?  These issues need to be examined. The potential
gains from being able to design effective organizational and institutional
frameworks are enormous. Moreover, the Bank would have a comparative advan-
tage in carrying out research in this area as it would be able to pull together
experience from a number of countries with a wide range of characteristics.


- 87-
Appendix II
.ESEARCH ADVISORY PANEL DN INuDTRIAL - ,ELrPtfEi"1 AND 7?ADE
List nf Projeots and Parer =.-vewed -
cat_Title                                       Prolect Code                                                  -ercs
, Incentt-v PlicIIs and tcon=mic Intearation
aeiopneni Drategies in Seni-ndustrial Conties      670-01        3alassa, Bela, *Exports and Etanmic Groh:   --ther Evidence." Journal of Devel::cent Economics,
June 1978. Als. World Bark Reprint Sies, no. 58.
Ex"port Inaent:ves and Expcr: (erformance ln Deueloping Countris: A Conparative
Analysis."Wyeitwtschaftitches kArchi,   ia 1978.  a  'ld Bank staff Waraing Paper,
no. 248. Washi gtn, D.C: World Bank, ann- 1974.
ed. Develatnent Stranelges on (eni-lndustal Countries.    (nine,, 1978).
* "Reforming the Systen af Ient:~s In Develtping Countries." Worid Development 3
(Jane 197):   pp. 365-62.  ipantsh iaranslat:, in Caadernos de Econ~ta, Dee.r L974: pp. 33-46.
Alsa World Bank ?eprint .srtes No. 22.
ana.ies al Stale and Toif) Inoels                    670-22 *      Hansen, J.8. ,"cn-ies       t a nals and Tartff .tves i n the Ironian Maar Vehiile I-datry:
A Cano study." 'nine,. ':on--b 1931
Purso11, G. Ettnomies of (tals and TarofI Le-e    (inaonpleae) ninea, 1972),
.astrial Policies and Econo~2 !ntegration             670-87        Balassa, Dela. "Coqprai.e avange o.d the Frospects fan Econmi bntegration in Wesc Africa."
in West Africa                                                         Pae pnepared for the Calla~ue sar l'2ntsgrat;a   en Afrique de 1'an,s. heid in lakar, S,enega
in Manah-Apnil, 1978.
Monson, Tett' 3., and Portel¼, fany, "An fosloetio of Expatziate labar Replacemen: in the
Trony Coast." ::urnal of 'enelonn-en"'-- -a. 'Sal. 3, No. 4, tecenbe, 1975. lisnusston
Paper, nr. 49. Center for ?esearah or f::n:tc Developnent, Unoversnty of Micngan,
April 196). French tronslaton in L'An:slite Eneniaae, June 1977.)
Parsit1, Garry, "Crass-Se::n Cast-Beneft: Analysns af Manufactuing -ith FareiSn :pital:
Case S:ody fra Denn Afraca."   Sape, ,reaeoted to the Southern EcononLe AssocIaion
meetinga, Vaohington, D.C., Naverber 1P9.
Dhepherd, Ueoffrey and Bor-kon, D., "The Sy-:= of Lnentinno and Compara:ive Adventag. in
gralian Ag  -caultre and !ndu.stry",           19io,o :.oe 1975).
otion of Norraditional Exports                       670-10 f      ECLA. Politicas fe ?ronnitn de ExOortac:es.   dE/CEPAL/1046, Santaago, Chite, Catber 1977).
Irernatlerol lrade Policy for :h Developent           671-75        (No pa-er'
af langladesh
Ither Papers                                                          Balassa, Dela,   Policv ?efn in nevelrnine Countries. Oxford, Perga.on Press, 1977.
Varii, and -rade Polia:y  tn :e ndn Conon Maks-."   Journal of
Conmon Marker Studies,  December 1973, vv. 176-195. Spantoh translation in
Rertsta de la Incenratcin, May 1874, p. -o4.
, "Types of tconnic rntegrant." Economic Integracion. Worldwide,
Regtanal. Secoral. (Frits achl,p, ed.: Proceedings of the 4th Congress of
the Internaional Econic Assonci_an hel       M n Andapest, Hungary in August, 1974).
London,Macmillan, 1976, pp. 17-31.
Gulsinger, B .,"ndustrial Protection in :71 Projects." (mimeo, Decenher 30, 1975),
'iid It.  Comparative Advannae, Trade Patterns. and Econ=Lc Growth
.rnatinal Nodel                                      670-07        Gtnsburgh. Victor, "A General Equilbibru= Model of World Trade, Part I: Pall Format
Canpuation ad Naaonmia Equlbria."  Zowles Foundation Dscussion Paper No. 412.
New Rarea, cnn.: Yale Unierity, 19,
, "A teneral tquiliorin  Nodel of World Trade, Part II: the
Etpirial $pecificaion." CoSlos ioundatoon Daacussian Poper No. 413.
New Havn, Conn.: Yale University, 19-.
4Pan.on of maufacturing for Exports la Developing     670-19 e      chang, Y.S. "An At,alysis of the Offshore ac:vities of the Japanese Electronics
Countries                                                              Industry."  (nmeo, arch 1973)
oDawa, Teratom-, "Lab. Resorca Oriented (agration of Japanese Industries to
Tatwn, Singaporn and Bouth Korea." World Bank Staff Working Paper, no. 134,
Washington, DC.: World Bank, August 1977.
Z        -onmi  Developent of East and Southeast Asia  670-79 *     Okita, Saburo, et al., "The Interrelated E-aanic Devslopet of East and
Souheast daa in the C-ming Decade." nito., Augs 1974)
atterns of Industrial Derelapent                       671-05        Praaab. Vinod, "An Ovee Paper on RX' 671-05 'Datnens of Indsta
Development'."  (mineo, May 1977)
,"reaanring Industrial Eports: A C-mpatative Statistical
Study af Variations Arising fro Differencs af DefoIntan." Wtk Baak
Staff Working Paper, no. 225. Washington, D.C.: World Bank, Feksuary 1976.
, "Statistical Indicators of :ndustrial Development: A Critique
of the Bastc Data." World Bank Staff Wong Paper N. I9. Washington, DC.
World Bank, Septeber 1974,


- 88 -
Appendt. 1,
Paga 2
Pr--- -itle                                               Project Code                                           Renor, s
A9 Co  :.rorive dtudy af the S-ur:es of Industriao            671-32        Balasso, Bela, "Acouncing for Eonomic Growth: The Cas. of Norway." World ank
r    n nd Utnnorur.l Chanze                                               DevelopMenn Reseannh Center Dstuaon Paper Ko. 17 and nford Economicnpase
(forthcooing).
Chenery, Hollis   ., "Tronsrtional Growth and World Industrializaton."   Presentd
at the Nobel Synposium on the Interna:inal Allocation of Ecoo-tnic Activity,
Stockholm, inne 1976.
Chenery, ollis 3. and Moises Syrquin, -A C-poratio   Analysis of dindustrial  rnwth"
Presented at the <ifth World Congress of the Inrernational Eccncic Association 0
Economic Cro.th id Resources, -okvo, August/0eptember 1977.
Syrquin, Moises, 'Sources of Indusrial Growth and Change: An Alcrnciv.e Mesore.
Presented at the European Meeting of :he Econonetric Society, Hjelsinki, AugusC 1976.
Westphal, .arry   ,, "Progress Repor: on a Conparative Study of rhe Sources of
Irdnstrial Gro-th and Structuril Change."   Ags: - 1977.
Sr:;oos af Graob ond Proouot::ty Ch.n.ge                    671-79        (No paperg)
,t     ers                                                               .o Balas, 3ela, "Effecto of Commencial ?olicy on  nrernoatoal rde, nhe - lo i  af
?roduction, o-d actor Movementos.-he International Aflocation of E -'--i k.Nlvity
pp. 230-58, :-D-^: (Bertil Ohlin, ?er-Ove esselhorn, and Per "agnus    Wijrn ed.)
Proceedings of a Nobel Synpostun beld at StockLo    in june 197. Maa-illn, London, l9-
,"A 'sages' Approach ta :r.ratIre .-:tage,"        orld 3aok i-:ff Working
Paper NI. 256. Wasbingtn, D.C .Wrld 3ank.
F i   :::    Exoarz ?roo   Fritoes Ln Developing Countries and
Access to Ma.ets v   keveloped Couries
Indusr:aktation ond Trade Poce       for the 1970's         670-20 0     Hughes, Helen, ed.), Prioects or Par:nership: Mdustrializar:on and rdc Pclicies
in the 1970'3. altimore: Johns Uop-i- Ulversiry Press, 1?-3.
E,port :--tio   and Prefer-:ces:   India                    6C-21 -       Wall. D., "Dra;t Peport on industrial Export ?olic.'." nimo,Agot 19
"Export Deelopment nd Pro-o:in- Polices in India." nireo,March !9:.
, "EoPt ProPesin Zone-' -     u-ral af -2-ld Trade La, Auu-ept-er 1976,
. -489.
, "The mpot of the Generalioed Systec of Preferences an     ndia's Exporto."
roineo, Febroa 1973)
Ex - :,-rozntos En Deveo:     g l  :oreos                 671-35        'No ppers,
M-rke-z; Manufacured Expor:o                                 611-56        (No papers.
Effects of Increased Impor-s of Manufactured Goods
Oron Developing Countries to :he United States              671-67        (No poper,
rfiecr of Increased 'mports of Manufactured Goods             671-66        (No papers
Oran ev-eloping Countries to Western Europe
e' .ns:tutions and Expanson :f Manufacured Exporra        671-68        (No pope.
Other Piers                                                                 Balassa, Bela, "ro     Trade and the international Econony:  Trendg, Prospects ind Policy."
World look Stoff Working Popen   No. 2E2. Washingon, D.C.: World fack.
Keesing, D. and Plesch, P.. "Recent Trends in Manufactured and Total Esports fr=n
Developing Countries."   (mimeo, June 1977).
Keesing, D., Plesch, P., and Triner, G., "Developing Countries' Exports of TexCLI and
Clothing: Per3pective and Policy Choices."     (m,e, may 1978)
Plesch,  .," "Deveaping Countries' EXports of Electronics and Eleoric.l Enginerng
Products." Inne, Februaro 1971)
_Leld I":.  Small Eanrrrises; Credi Mokets, Public Enterprises
Fioncing of 3nall Scale Industra                             670-77 *      Koch.v, D., et al., "Financing the Dveelopment af Smalloale Indosries."
World Bank Staff Working Paper   No. 191. Washington, D.C.: World Bank.
Nonember 1974.
Snall-Scale Enterprise Developcent                            671-59        Anderson, Dennis, "Estieating the Economic Benefits of Small Scala Encerprise
Credit Projects." (mimeo May 12. 1978)
S,  ra:erprise Financing:  Role of Informal Credit         671-65        (N. popees)
Capital Market Imperfections and Econonic Development         671-69       Bharr, V.V., "lnterest Rate, Transaction Costs and Financial Innovations.' :-rld Oank
Publie and Private Fionnce Dvision. DomestiC Finnce Studies Ko. .7, Janurv i973.
Roe, Alan R., "Soe Theory of the Financial Internediation in Less Developed
Countries." World Bank: Public and Private Finace Division, Dornelc Finance
Studies No. 50, May 1978.
Safto, Katrine A., and Villanueva,Dan P., "Portfol1o Determinants of Commercl
Bank Earnings in Selecred Asian Countries,     World Bank   Poblie ond Prvte.
FL-nnce Diviston, Donestio Flna-ne Studies No. 49, ,arch 1978.


- 89 -
Page 3
TileProtect Code                                                                           Reports
public Manufacturing Enoerprise,                              671-71           (No paperb)
other Pacers                                                                   O'Mara. G., "An Econometric Analysis of the Role of Small-Scale Industry in
Brazilian Industrialization, 1949-1970." (mieo, June 1976)
______,"The Role of Small-Scale Industry in Recent Brattlian Industrialization."
(.tmeo, April 1978)
Field V.  Capital Utilization. Capital-Labr   Substitution, and Technological Chaae
Scope for Capital-Labor Substitution in the Mechanical        670-23           Lamyai, Thanong; Rhee, Yung W.;and Westphal, Larry   E., "Factor Substitution,
Engineering Industry                                                           Returns to Scale and the Zrganization of Production in the Mechanical
Engineering Industry."  (=meeo, July 1978)
Sam, Soan W.; RSe, Yung   ,; and Westphal. Larry  E.l, "Dact  evelopment for
a Study of the Scope for Capital-Labor Subsitution in the Mechanical
Engineering Indastries."  (niaeo, February 1973)
Rhee, Yung W., and Westphal, Larry E., "Institutional and Economic Criteria
for the Choice of Technology in Developing Countries."  (mieso, June 1978)
Lamyst, -hanong; Rhee, Yung, Wi, and Westphal, Larry E., "Factor Substitution,
Returns to Scale and the Crganization of Production in the Mechanical
7ngineering Industry."   =iono, July 1978)
ndustrial Capacity Utilization in Selected Latin             670-25           Schydlowsky, Daniel M., "Capital UtIlIzation, Growth, Employment and Balance
Aerican Countries                                                              of Pay-ents and Price Stabilization." Discussion Paper Series Po. 22
Boston, Mass.: Boston !niversity, December 1976.
-mplc-,-ent and Capital Labour Substitution                   670-54 *         Mu.dlak, lair, "Estiration of Production Functions; A Survey."
(nimee, June 1973)
Thorbecke, Erik, and Sengupta, Jati, "A Consistency Framework for Employmant,
utput and Income Distrizution Projections Applied to Colo,mbia.
(mime, June 1973)
ndustrial Capacity itilization                               670-95 u         3atista, Romeo, "on Excess Capacity in Philippine Manufacturing."
?hilippie Review of Business and Economics, December 1971.
='ughas, Helen et, al., Capital 1tilization in Manufacturing in Developin
Countries, World lank Staff Working Paper No. 242, September 1976.
Kin, Young C. and K-,on, Jte K., "The Uttilization of Capital and the Growth
of output in a Developing tconomy," Journal of Development "conomic,
April 1977.
ti=, Dav.d, "On the Measuretent of Capital Utilt.stion to LDC's".
Cxford Economic Papers, March 1976, pp. 149-159.
, "Capital Utilization of Local and Foreign Establishtes."
Review of Economics and .tattetics, May 1976, pp. 209-217.
Field V,. Progransnin in the Manufacturing Sector
Prograoning in the Manufacturing Sector                      670-24           3alassa, Bela, and Stoutjesdijk, Ardy, "Econosaic Integration among Developing
Countries." Journal of Conon Market Studies, 186 (September 1975). Also
World Bank Reprint Series No. 30.
Chokai, Armeane M; Meeraus, Alexander; and Stoutjesdijk, A., The Planning of
Investment Programs in the Fertilizer Industry, Volume 2 in The Planning
of Investment Programs.  (forthcoming, Johns Hopkins University Press.)
, "A Planning Model
of the Fertilizer Sector in Egypt." World Bank Staff Working Paper no. 269.
July 1977.
Kendrick, David and Stoutaesdik, Ardy, The Planning of Industrial Investment
Prozrams: A Methodology. Volume 1 in The Plannina of Investment Programs.
(edited by Alexander Meeraus and Ardy Stoutjeadljk). Johns Hopkins University
Press, Baltimore and London, 1978.
Stoutjesdijk, Ardy, and Westphal, Larry, (ads.), Industrial Investment Analysis
under Increasing Returns. (mimeo, September 1978)
hestphal, Larry E. , "Methodology of Investment Planning in the Noa-Process
Industries,"  (mimco, February 1976)
Westphal, Larry E., and Rhee, Y.W., "The Allocative Consequences of Econotes
of Scale."  Development Research Center, Discussion Paper gb. 18, World Bank,
January 1976.
" A Process Analysis Model of the Mechanical
Engineering Sector."  (:noea, January 1976)
Othar Papers                                                                   World Bank, "Fertilizer Requirements of Devaloping Countries." World bank
Report Ro. 446, May 1974.


- 90 -
Page 4
Proiect Title                                             Project Code                                          Reports
Industrial Capacity Utilization                              570-95*        Morwetz, savid, "Israei" in Caital Ut-lizatlon in Developng Countries
(Eorthcoming).
Thoumi, Francisco E., "The Utilization of FiXed Industrial C3piral in qpI,
Some l=pirical Findings," in Essays or Colombian Irdustrialization. lditef
R. Aliert Bery, Princeton, %.J.: Princeton Universitv ?ress, 1974.
'dinstor., Gordon, "The Theory of Capita' tilizatior and :deness,"   4arld as-,
Staff Gord    Paper No. 174. April 197-, -ournal of    conooic 'i:erator.
(Decceer 1974).
Appropriate Industrial Technology                            671-51         Pack, 'cward, "The Macroeconomic ^ffects of Appropriate Choice af s:i-re.
(ri=eo -ecember, 1977)
- "The Capital Goods Sector rc LDC's: 4 Surtev. 'i-co, Ya:-
Other Pavers                                                                Amselen, Michel, "Technological C'.oize in Developir.2 COntries      e   ooct
Differerces in Factor Costs."  Unpublished PhD Thesis, .arvard :,ivrsi:
Choi, .ar   and '.tphal , Larry E.. "Ct::i.e of the Mexicn "ltantrit  iprec:
Plannin; Study," Nove.ber 1976.
Creer. -., "Planning Interdependent Production Processes i,t :r easing
Returns to Scale: Some Theory ard an A'gorithm."    foole 7o         o-ozrl e,
Labotatoire dEconnoenrie, (m   C, Tece-ber 1976'.
Westpha. '.rry E., "INDP ?roposal - Meyico Plant Eauip'nern     t  :n  siz
March '4, 1978.
I/  Original research proposals and other documentation on the research projects -nre reviewed by the panel in all cases.
*   Completed research projects.