37333
WORLD BANK
* RESEARCH
NEWS
Volume 1 Number 3
Fall 1980
CONTENTS
The World Bank's Research Program: How Operationally
Relevant Is It?                                           1
Gobind Nankani
Completed Research
Distributive Impact of Public Expenditures                 12
Commercial Bank Behavior                                   14
Country Case Studies of Agricultural Prices, Taxes,
and Subsidies                                            15
Consequences of Risk for Agricultural Policy               16
Migration in Western Africa                                18
New Research
Brazilian Income Distribution and Growth                   19
Poverty, Fertility, and Human Resources in Indonesia       20
Case Studies of Determinants of Recent Fertility
Decline in Sri Lanka and South India                     21
Improved Technology for Animal-Powered Agriculture         21
Reduced Information Methods of Real Income Comparisons     22
Pricing of Indigenous Energy Resources                     22
New and Forthcoming Publications                           23



--k



It'orld Bank Research News will be issued three times a year. It supplements the descriptions of
socioeconomic research projects in progress given in the annual World Bank Research Program: Abstracts of
Current Studies and the annual Catalog of World Bank Publications. Research News is available free of
charge to institutions and individuals with a professional interest in development. To be placed on the
mailing list or to rieceive additional copies, please send a complete address, including your title, to the
Publications Unit,World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A.
Enquiries on particular research projects should be addressed to the individuals or departments cited.
Other enquiries, comments, and suggestions for future issues will be welcomed and should be addressed to
the Editor, Office of the Vice President, Development Policy, World Bank, 1818 H Street, N.W.,
Washington, D.C. 20433, U.S.A.



I



THE WORLD BANK'S                                          this task compels us to adopt the simple ap-
RESEARCH              PROG RAi           :                 proach of using ex ante research objectives as our
RC   PM                      criteria for judging relevance. In the final analy-
HOW        OPERATIONALLY                                   sis, what is being offered is an interpretation,
RELEVrAN               TIS  IT?                           with the logical underpinnings clearly laid out, so
RELEVA                   I5  IT? 1 1.that readers may judge for themselves.
Goebind Nankani, Economist, Office of the Vice             Section A prepares the ground for this assess-
President Development Policy'                              ment by discussing common usage of the phrase
"operational relevance" and choosing a definition
based on the objectives of the World Bank's
In recent years, it has become fashionable even            research program. In Section B, a model of the
for academic economists, at their annual meet-             research process in the Bank is developed. These
ings, to bemoan the lack of "relevance" of much            two sections help to fashion a measuring rod
that passes for economic theory. With university           with which to assess the Bank's research pro-
departments under such attacks from within, it is          gram in Section C. Section D offers examples of
hardly surprising that the "operational rele-              the different kinds of research projects identified
vance" of the World Bank's research has been              in the preceding section, while Section E
questioned. The aim of this article is to assess           attempts to reconcile the findings of the preced-
how "relevant" the Bank's research program is.2           ing sections with the not uncommon notion,
within the Bank at least, that the Bank's research
As noted in the first issue of Research News, the          is, for the most part, not operationally relevant.
bulk of the Bank's research is done by the Devel-          Section F concludes.
opment Policy Staff (DPS) and Central Policy
Staff (CPS). For purposes of clarity, it is impor-         In a nutshell, the arguments of this article are
tant to draw a distinction between the research            that: (1) the phrase "operationally relevant" is
activity and the nonresearch activity of DPS and           commonly interpreted too narrowly, given the
CPS. This article refers to the former activity.3          objectives of the Bank's research program; (2) the
research process in the social sciences differs
Given the controversial nature of the subject and          from that in the physical sciences in that their fi-
the difficulties involved in arriving at an assess-        nal products are of different degrees of applica-
ment, no claim of definitive judgment is proffered         bility to situations in the real world; and (3)
here. Also, while it is clear that an attempt to           while the distribution of the Bank's research
judge the operational relevance of research must           resources has not been unduly biased against
ideally rest in part on ex post evaluations of             "operationally relevant" research, there is scope
individual projects, the monumental nature of              for a greater effort toward research applications
and design/development studies than heretofore.
1. The author is grateful to Shankar Acharya, Stanley Please,  The note concludes by stressing the need to judge
and Martin Wolf for their extensive and very useful comments.  the Bank's research also by its contribution to the
Helpful comments were also received from K.Y. Amoako, Surjit  Bank's medium-term and long-term goals, rather
Bhalla, Nancy Birdsall, Hollis B. Chenery, Armeane Choksi,  an same         its  andbongt    o   als,erate
Susan Cochrane, John Duloy, Dean Jamison, Javad Khalilzadeh-  than largely by its contribution to immediate
Shirazi, Deepak Lal, Guy Pfeffermann, Robert Picciotto, Anan-  objectives.
darup Ray, Ardy Stoutjesdijk, Arvind Virmani, Bevan Waide,
and Shahid Yusuf. Pradeep Kotamraju provided indispensable
research assistance and Rachel Weaving's editing was, as usual,  A. Operational Relevance:
excellent.                                                A Pandora's Box?
Responsibility for the views and interpretations expressed here
rests entirely with the author; they do not in any way reflect  It is not uncommon to find research being under-
those of his colleagues or of the World Bank.             taken in organizations whose principal activity is
2. Further views on this topic will be included in future issues  not research. Nor is it uncommon to find that
of Research Nev',                                         research activity is viewed with much skepticism
3. Total annual resources spent on research (staff time, travel,  within such organizations. The charge, very
consultants, and computing costs) in the Bank are the equivalent  often, is that much of the organization's research
of approximately 80 professional man-years.                     relevant.
is not rlzat



As commonly used in the Bank, the phrase "op-        any given research project and the relevance of
erational relevance" means different things to dif-  the research program as a whole.
ferent persons, depending largely on their parti-    The Bank's research program, at its inception,
cular location in the organization. Based on         w   as   red  earch four    at  tS     oc -
casual empiricism, three interpretations of the
phrase seem quite common among operational           tives were recently reendorsed by the external
staff:                                               General Research Advisory Panel (GRAP), of
outside experts, chaired by Sir Arthur Lewis, in
1. Short Pay-off Period. Research is deemed       its evaluation of the program. They are:
more operationally relevant the closer it is to      I. To support all aspects of the World Bank's
application and development, the farther it is       operations, including the assessment of devel-
from basic research, and hence the shorter is        operon     s     in the     countries.
the period between its initiation and its frui-      opment progress in member countries.
tion.                                                2. To broaden understanding of the develop-
2. Congruence with Operational Departments'          ment process.
Planned Work Program. The greater the fit            3. To improve the Bank's capacity to provide
between the research project's objectives and        advice to member countries.
those of any operational unit's work program,
the more relevant is the project deemed (by          4. To assist in developing indigenous research
that unit) to be.                                    capacity in member countries.
3. Confirming "Hunches." The closer the            If we accept these objectives, three points become
hypotheses of the research project to rules of     clear. In the first place, "operational relevance"
thumb already in use by operational staff, the     as defined above, is directly addressed only to
less relevant is the project deemed to be (be-     two of these four objectives (that is, objectives 1.
cause operational staff see no need for confir-    and 3.). The charge that the Bank's research
mation of these).                                  program is not "operationally relevant" may very
well reflect the view that much less effort has
It is, of course, understandable for an organiza-    been devoted to the first objective than is consid-
tion whose main function is not research to value    ered appropriate. We may distinguish between
research applications and development more than      the narrow definition of "operational relevance"
basic research. And yet basic research is an input   (objective 1. only) and the wider definition (ob-
into research applications and development. The      jectives 1. and 2.). Secondly, the relevance of the
need for some measure of balance is clear. Simi-     Bank's research program is primarily a question
larly, it is understandable that any unit's evalua-  of whether the allocation of its research projects
tion of the relevance of a research project should   over these four objectives satisfies some precon-
be influenced by the latter's potential contribu-    ceived or subjective distribution. It is reasonable
tion to its own planned and current activities.      to ask that the research program exhibit some
And yet, to emphasize this criterion to the exclu-   degree of balance over the four objectives, but
sion of any broader view of the role of research     quantifying this balance on an a priori basis is
is unduly myopic. Finally, it is often forgotten     difficult. Few attempts have so far been made to
that the confirmation of hunches is as important     compare the actual allocation of the research
a research result as their refutation.               effort over these objectives to any notional alloca-
If it is true, as we have argued, that each of       tion. A third point to note is that these objectives
these considerations is of some value, and yet       make clear that research output is as important
none is by itself sufficient, what broader standard  as are project loans and economic reports.
may we use? It would seem only fair and reason-      B. Choosing a Model for the Bank's
able that the relevance of the Bank's research       Socioeconomic Research Activity
program be judged against the objectives set for
it, at its initiation in 1971. This has the effect of  We begin with a few observations on research.
drawing a distinction between the relevance of       Research is an inherently risky affair; one must,



therefore, be constantly aware of the possibility    complete freedom to choose its areas and topics
of failure in any given research project, and be     for research, the Bank's research effort must sat-
prepared to judge the success rate of a portfolio    isfy the objectives laid down for it. On the other
of research projects. Also, there are long and       hand, unlike the research department of a private
uncertain lags in research; as such, it is difficult  enterprise, which in the final analysis is justified
to predict when applications will flow from some     by its contribution to the output and profits of
piece of basic research, and whether or not addi-    the enterprise, the Bank's research program is
tional basic research will be required before        hinged not only to Bank operations, but also to
applications can be expected. And third, many        other objectives, including a broadened under-
areas of concern that are of great relevance may     standing of the development process. In a sense,
not be susceptible to research because of the pau-   the Bank's research objectives explicitly recognize
city of tools possessed by the social sciences in    the Bank's "publicness," while requiring that
relation to the complexity of socioeconomic          "operational relevance" receives adequate
phenomena.                                           attention.
Two sets of comparisons help to place the Bank's     A second comparison that throws some light on
socioeconomic research in context. First, the ob-    the Bank's research program is that between the
jectives of the Bank's research place the Bank, in   research process in the physical sciences and that
the institutional spectrum, somewhere between a      in the social sciences (including economics). In
university department on the one hand and the        the former, the research process is traditionally
research department of a private enterprise on       classified into three stages (with appropriate
the other hand. This is because, unlike a univer-    feedback mechanisms connecting the different
sity department which by design is meant to have     stages):
Basic ReNearch -.4pplhteI Research -  Degn and D)ev elnpmnenf
Traditionally, the research department of a             are to be useful, are likely to be greater in the
private enterprise undertakes work in each of           social than in the physical sciences, primarily
these areas, while universities tend to concentrate     because of the greater variance in social and
on the first two stages.                                economic institutions than in the properties of
matter.
Prima facie, there seems to be no reason why
research in the social sciences cannot be similarly     2. For some socioeconomic research, the design
described. Indeed, some aspects of socioeconomic        and development stage is essentially redundant
research-such as industrial investment program-         and applied research activities have the same
ming techniques-are best approximated by what           kind of research value as the "lessons of histo-
is essentially a unilinear research process. The        ry"-namely, they provide insights which of-
difference arises, however, in that the social          fer some basis for policy advice, but not defi-
sciences have to rest on assumptions about social       nite policy prescriptions; in other words, not
and economic institutions which can vary over           all socioeconomic research projects can be ush-
time and place, while the properties of matter          ered along to the design and development
underlying the physical sciences are invariant.         stage.
The overriding importance of history and of
varying institutions is what gives rise to the es-      3. Finally, applied research in the social
sential complexity of socioeconomic phenomena.          sciences can proceed on the basis of some un-
The implications of this for one's view of the          derlying general paradigm (such as "neoclassi-
research process in the social sciences are as          cal economics" or "institutional economics")
follows:                                                and need not be preceded by a specific element
of basic research.
1. For any given successful basic research
effort, the subsequent applied research and        Diagrammatically, these points imply the follow-
development efforts that are needed if results    ing scheme for socioeconomic research:
Basc Rear~ h-     pplit-d Re-search .-[ f-), in and I)rve/,opmen
A.4pplied Research -     'Leivnns ni Ihstirn



The preceding discussion has implications both        The Allocation of Research Effort
for the organization of the Bank's research, and
for the allocation of the effort over the different   The need for coupling, as well as the recognition
research stages.                                      of a higher ratio of nonbasic research to basic
research in the social sciences than in the physi-
The Organization of Bank Research                     cal sciences, both imply the need for a careful
o te                        assessment of the distribution of the Bank's
To the extent that the Bank's research is ex-         research effort over basic and nonbasic research.
pected to partake of the nature of university
research (objective 2.) as well as of the nature of   In particular, this suggests the need for some
private industrial research (objectives 1. and 3.)    measure of balance, as judged by the research
presearc indutheiBank musearch beobo otives d from    program's ability to contribute to all four objec-
research in the Bank must be both protected           tiesorBnorsarh
and linked to the Bank's operations. One              hves of Bank research.
research manager has used the words "insulated"       C. The Model Applied to Bank
and "coupled" to describe these relationships.        Research
There are many ways of achieving such insula-
tion and coupling, but this is not the place to       One of the conclusions arrived at in the previous
pursue that discussion. What is important is the      section was that the division of the Bank's
rationale underlying these relationships.             research effort over the program's four objectives
and, in particular, over basic research and non-
Insulation in the conduct of research is important    basic research, is of the essence. This section will
for two reasons. In the first place, it is necessary  assess the Bank's portfolio of completed and on-
for researchers, given the nature of their activity,  going research projects from the latter two points
not to be subject to the day-to-day pressures in-     of view.
herent in operational work. Secondly, and this
is particularly true of those involved in basic       It is well to note, at this point, that the General
research, the world of science is a very important    Research Advisory Panel endorsed the Bank's
segment of the good researcher's audience, and        research effort in the 1970s as being of a high
part of his/her principal output is publication.      caliber, but suggested, inter alia, that a greater
While this view applies less forcefully to those      effort be made in the area of research applica-
engaged in applied research, it should not be dif-    tions (design and development research activities
ficult to agree that unless the Bank's researchers    is the terminology used here) and the dissemina-
continue to be subjected to the judgments of their    tion of research results. In the past year, in pur-
professional peers, they will not remain good         suit of a greater degree of coupling, four Steering
researchers. In what is one of the largest centers    Groups (one each in Industry, International
in the world for research on development issues,      Economy, Agriculture, and Urban/Regional
this concern for quality must not be lightly          Economics) have been established, in which
treated.                                              research producers and consumers are brought
together to advise the Research Committee,
On the other hand, coupling research with oper-       which administers the Bank's "external" research
ations is also crucial because the Bank's research    budget, on research priorities.4 In addition, a
is meant to be responsive to goals other than just    Research Applications Adviser has recently been
an improvement in the understanding of the            appointed to aid in promoting research applica-
development process. Insulation is very different     tions (in the terminology used here, design and
from isolation. Few research departments outside      development proposals).
academia undertake to replicate the working con-
ditions of a university department. Although the       4. Centrally funded or "external" research refers to all
Bank's research must be cushioned, it must not        research which is partially funded by the central research budget
be isolated; it has to be responsive to the Bank's    and is approved by the Research Committee; it accounts for ap-
There is little rationale for having aproximately 65 percent of the total nonstaff resources devoted to
operations. dere     within thea                      research. Tbe other 35 percent is accounted for by departmental
university department within the Bank.                studies, including special studies and research preparation.



RiES ..s L 7W t Y COPY
F 4.LL 1980                                                                               Ri   i
The following chart classifies the Bank's portfolio       design and development projects (this third being
of completed and ongoing centrally funded re-             analogous to the design and development phase
search projects (excluding in-house departmental          of industrial research).
studies) into basic research, applied research, and
Table I summarizes the evidence.
1, :'I'C                          Irl  ,r  P_' ir
--II I dehn          NF-FII.Cli tlVe :luId,e;  Nec vr3llk .,ricneme
',ud e
De%elo.p.nv Itihnique~  Hq_pmlhci,t IeMIFnI  Rei ..tl spttn;:rrd
1.,r lejrl% delned    l -dmple vur%e% hjied rl   u1d,dr
pro:'blem!             md publi hed d.iaJ
PHI-I St tdtie.         tnnit r,n snd ,im-  'i ae-. I-ihe-.:ri
uloi-n 1.1 rrw.dek   re%iei  rudie,
lndu;it- riujd,e;
Table 1. The World Bank's Allocation of Research Authorizations: Basic
Research, Applied Research, Design/Development Projects
Fiscal Years 1972-80
External
Number of           research           Staff time
research            funds             (number of
projects     %      ($000)     %     man-years)     %
Basic research
Total                      65       30.1   8,922.2   40.9      159.7       41.4
Average per project                          137.3               2.4
Nonbasic research
Total                     151       69.9  12,903.9   59.1      226.4       58.6
Average per project                           85.5               1.5
of which:
Applied research
Total                   121       56.0   9,665.6   44.3      189.3       49.0
Average per project                         79.9               1.6
Design/development
projects
Total                    30       13.9   3,238.3   14.8       37.1        9.6
Average per project                        107.9               1.2
TOTAL                       216      100.0  21,826.1             386.1      100.0
0_     _   _    _  _  _  _  _  _  _   _  _  _  ~   ~  ~  ~  ~   . . . - -    -   _  -



Most of the Bank's research projects, in one way       2. All case studies (pilot and applied)
or another, further all three main objectives of
the Bank's research. As such, any attempt to al-       3. Techniques for clearly defined problems
locate research projects over these objectives is
somewhat arbitrary. However, for heuristic pur-        Research contributing to Bank's capacity to
poses, it may be worthwhile risking some arbi-         provide advice to member countries; and re-
trariness in order to obtain some idea of this allo-   search contributing to a broadened under-
cation, even if only a crude one. Such an attempt      standing of the development process
is also made in what follows.                          1. Hypothesis testing
Using the categories noted above, and adopting         2. Estimation and simulation of models
the narrow definition of "operational relevance,"
the following classifications are suggested:           3. Industry studies
Research supporting Bank's operations                4. Modeling
1. Design/development projects (sectorally ori-
ented research, regionally sponsored studies
and state-of-the-art reviews)                       Table 2 summarizes the evidence.
Table 2. The World Bank's Allocation of Research Resources
over Research Objectives
Fiscal Years 1972-80
External
Number of          research          Staff time
research           funds           (number of
projects    %     ($000)     %     man-years)    %
Research
supporting
Bank operations
Total                  140.0      64.8  14.775.5  67.7      242.5      62.8
Average per project                       105.5               1.7
Research supporting
policy advice and
research supporting
understanding of the
development process
Total                   76.0      35.2   7,050.   32.3      143.6      37.2
Average per project                        92.8               1.9
TOTAL                  216.0     100.0  21,826.1  100.0     386.1     100.0



[  XL I. 1 Q19811                                                                    i<! _ _  _i.__ _ _ _ _  j _  _
On the basis of Tables I and 2, the following ob-       Basic Research
servations can be made. First, the distribution of
the Bank's research resources over basic and non-       This category was stated above to include
basic research is approximately in the ratio of         modeling, techniques for clearly defined problems
1:1.5 (see Table 1). Given the earlier observation      and pilot case studies. Many of the efforts in this
that the ratio of basic research to nonbasic            area are designed to set the stage for work of a
research tends to be higher in social-science           more applied nature. In other cases, the initial
research than in the physical sciences, it is not       attempt is not successful, and there is no subse-
surprising that the GRAP Report advocated a             quent work. The studies described below exem-
greater effort in research applications (that is, in    plify both kinds.
the terminology adopted here, design/develop-
ment activities).                                      The "Highway Design Study-Phase I," com-
pleted in 1971, was a pilot study that sought to
Second, the distribution of the Bank's research        analyze the trade-offs among major cost compo-
over the three main objectives of the research         nents in alternative highway design and mainte-
program (see Table 2) suggests that "operation-        nance standards. It was considered of interest to
ally relevant" research accounts for approxi-          assess the trade-offs between initial construction
mately 65 percent of the Bank's research re-           costs and future maintenance costs and road-user
sources, where this category includes all research     costs, and their implications for highway design
that may be described as supportive of Bank            in the developing countries. Phase I developed a
operations (using the narrow definition of "oper-      framework for analyzing relationships among
ational relevance"). The rest of the research ef-      design standards, maintenance standards, and ve-
fort (35 percent) is designed to enhance the           hicle operating costs. It set the stage for Phase II
Bank's capacity to provide advice to member            of the research (see under "Applied Research").
countries and to broaden its understanding of the
development process. The preceding two observa-        Another example of this kind is the research
tions are not contradictory, since basic research      project on "Social Accounts and Development
has included many research projects that have          Models" which has aimed at providing a basis
gone on to become, in their later phases, "opera-      for macroeconomic models that permit analysis of
tionally relevant."                                    the trade-offs between alternative policy goals,
allowing, in particular, the study of employment
In short, the evidence suggests that the Bank's        generation, income distribution, and eradication
research has exhibited a high degree of opera-         of poverty. A pilot case study constructed a Social
tional relevance, but has also reached a stage in      Accounting Matrix (SAM) for Malaysia. Exten-
which there is a need for a greater effort in the      sion to other countries is likely in the near
field of nonbasic research, in particular in           future.
design/development studies.
A third example of pilot studies that are designed
D. Some Examples                                       for wider applicability is the "International
Comparison Project," which aims at establishing
The preceding discussion has attempted to assess       a reliable system of comparisons of real product
the operational relevance of the Bank's research       and the purchasing power of currencies. In Phase
program by (1) clarifying the nature of the            I,s the essential features of the methodology were
research process in an organization such as the        developed and applied to ten countries. The
Bank, and in the social sciences; and (2) studying     Phase vl   report6 added six countries while Phase
the allocation of research resources over basic         _
research, applied research, and design/develop-         5. Irving B. Kravis, Zoltan Kenessey, Alan Heston, and Robert
ment activities, and over the objectives of the        Summers, Phase 1: A System of International Comparisons of
Bank's research program. This section offers ex-       Gross Product and Purchasing Power (Baltimore and London:
amples of each kind of research.                       The Johns Hopkins University Press, 1975).
6. Irving B. Kravis, Alan Heston, and Robert Summers, Phase
I: International Comparisons of Real Product and Purchasing
Power (Baltimore and London: The Johns Hopkins University
Press, 1978).
_____                                                                              K  L--



III' extends the coverage to thirty-four countries,  models, and industry studies. Unlike the basic
In order to permit the extension of this metho-      research projects, these studies tend to have a
dology to a large number of countries, the Bank      higher success rate in that their results yield in-
has recently initiated a project to explore short-   formation and knowledge that is useful, although
cut methods, namely the "Reduced Information          to varying degrees. Some of these studies can be
Methods of Real Income Comparisons" study.            extended to the design/development stage to yield
results that are directly relevant for operations.
Examples of research projects that have at-          Others are destined to end at the research appli-
tempted to develop techniques for clearly defined    cation stage and to provide, in a general way,
problems include "Programming in the Manu-           isgt o      oiymkn        n    rjc   ein
facturing Sector," which deals with the problem      insights ofo polic mkings and pojfectd design.
of investment planning in industries characterizedExmlsobthknsaeferdeow
by increasing returns to s,cale;- and "A General     A useful example of an applied case study is the
Algebraic Modeling System (GAMS)," which             "Highway Design Study-Phase II." As noted
explores a machine-intensive route to modeling,      above, Phase I developed an analytical frame-
and aims at a dramatic reduction in the time,        work for the study of trade-offs between design
technical skills, and costs associated with existing  standards, maintenance standards, and vehicle
modeling and data systems. In particular, the        operating costs. Phase 11 has focused on collect-
research project on "Programming in the Manu-        ing .empirical data on the underlying physical
facturing Sector" has adopted the GAMS repre-        and economic relationships necessary to calculate
sentation to guarantee consistency and replicabil-   cost trade-offs under various conditions. Field
ity of all models and their data as used in a        studies have been undertaken in Brazil, India,
series of applications.                              and Kenya. This phase has also further devel-
oped computerized investment models that are al-
An example of a basic research project that          ready being used at the prefeasibility and feasi--
proved to be a blind alley was that entitled "In-    bility stages of project design in the Bank. Under
direct Estimation of the Size Distribution of In-     Phase 1II, this project has embarked on the
come"; it aimed at obtaining indirect estimates of   design/development stage which will culminate
the size distribution of income across households    in several publications, including a revised
among various socioeconomic groups without           edition of Jan de Weille's Quantificati'on of Road
undertaking nationwide surveys. This was to be        User Savings,8 and a book evaluating the major
done by developing a methodology to use house-       options in road design and maintenance
hold surveys and national accounts for this pur-     strategies.
pose. Its completion report notes that "the results
of various attempts at estimating income distribu-    Examples of research projects aimed at testing
tion based upon estimated assumption shares          hypotheses regarding particular aspects of the de-
were all unacceptable in the sense that they did     velopment process include "Patterns of Industrial
not lead to nonnegative estimates of numbers of      Development" which aimed, inter alia, at testing
households in each income group." It should be       a series of hypotheses relating to systematic vari-
noted that 'Initially the project had been recog-    ations in the composition of manufacturing pro-
nized as a low-cost exploration of a potentially     duction and foreign trade. A research project
innovative idea.                                     based on sample survey data is the ongoing
"Case Studies of Determinants of Recent Fertil-
Appli'ed Research                                    ity Decline in Sri Lanka and South India,"
which seeks to uniderstand the process of fertility
This category includes applied case studies,         decline by testing a number of hypotheses, in-
hypothesis testing (sample survey based or pub-      cluding, in particular, those related to changes in
lished data based), estimation and simulation of     marriage patterns.
7. Irving B. Kravis, Alan Heston, and Robert Summers, Phase
III: Wo rid Product and Income: International Comparisons of
Real GD)P (Baltimore and London: The Johns Hopkins Univer-  8. Baltimore and London: The Johns Hopkins University
sitv Press, forthcoming, 1981).                       Press, 1966; 3rd printing 1970, out of print.



FALL 1980                                                                     R Si.s FR(I . Ni FN S9
Among the few research projects that estimate      criteria, now pays increasing attention to tariff
and simulate models, is the study of "Projections  structures, economic efficiency, and the use of
on the Extent of Food Deficits of Target Groups    pricing to achieve a range of social and economic
under Alternative Policy Programs," which          objectives.
aimed at projecting the supply and demand for
food of specific target groups in selected develop-  An example of a research project that was also
ing countries (Bangladesh, India, Indonesia, Mo-   sectorally oriented but not aimed at pricing prin-
rocco, Nigeria, Pakistan, Sri Lanka, and Tanza-    ciples is "Appropriate Technology for Water
nia). Also in this category is "Simulation of      Supply and Waste Disposal," which aimed at as-
Buffer Stocks," which applies the method of sto-   sessing various technological options available for
chastic simulation to estimating the costs and     water supply and waste disposal in developing
benefits of buffer stocks to farmers, consumers,   countries, in order to enhance the Bank's ability
the government budget, and general economic       to direct the benefits of its water and sanitation
welfare. Some generalizations about the stabiliza-  loans to the urban and rural poor.
tion effects, and the costs and estimates from     Regionally sponsored research studies include
stock and trade policies, are expected. In a sub-  "LaboraMigratonsand Manpower inthMde
seun phse spcii conr aplcain arLabor Migration and Manpower in the Middle
sle     p.    s      c       a                     East and North Africa," which is preparing mi-
planned.                                           gration projections for the subregion for 1980-85,
IDesign Development Research Projects             and based on these, proposes to examine the
prospects for future labor movements in the
These projects include sectorally oriented         region under alternative labor market policies.
research, research undertaken by the Bank's        Also in this category is the study of "Adoption of
Regional Offices to fulfill specific operational   Farm Technology in Northern Nigeria," which
needs, and state-of-the-art reviews. This kind of  seeks to study the adoption of techndlogy in a
research activity has tended to have a small share  series of agricultural development projects
of the Bank's research portfolio for at least two  initiated in Nigeria in 1975 with the assistance of
reasons. In the first place, the final product of  the World Bank.
social-science research is frequently not of a kind
that can be "packaged" for immediate               Finally, as examples of research projects that
analogy was drawn earlier between these latter     have undertaken reviews, one may refer to
research results and the lessons of history.       "Evaluation of Asian Data on Income Distribu-
Secondly, the time lag between basic research      tion" and "Evaluation of Latin American Data
and the design/development stage is generally a    on Income Distribution," both of which reviewed
long one. This is evidenced by the successful      existing data on the subject. Also in this category
Highway Design Study which commenced in           is the study of "China's Rural Development Ex-
1969, and which will not have been ushered        perience.
through its final design/development stage before  E. The "Operational Relevance" of
1983. Neither of these reasons suggests, however,  Bank Research: An Assessment
that there is no scope for a greater effort in this
direction.                                         It was noted in the introduction that the criteria
chosen for judging the relevance of the Bank's
Perhaps the best examples of sectorally oriented  research were of an ex ante, rather than an ex
research in the Bank have been in the field of    post, nature. To that extent, the criteria are
public utility pricing. Among these is "Pricing   somewhat incomplete, but they do provide one
and Investment in Electricity Supply," which had  useful way of approaching the task. The major
as its main objective the development of a practi-  disadvantage of this approach is that it precludes
cal method of incorporating marginal cost-pricing  us from making any judgments about the kinds
principles in electricity tariff analysis in the de-  and the quality of the studies undertaken under
veloping countries. As a result of this work, the  the categories suggested above (modeling, hypoth-
Bank's pricing and investment policy, originally  esis testing, sectorally oriented studies, etc.).
dominated entirely by financial and technical     Thus, it is entirely possible that, while the allo-
\NrORI I) It \k S



I() RESI: *LR(CH INEWS                                                                   FALL 1980
cation of research resources may be favorable       results that they would like to be able to use
from the point of view of the Bank's research ob-   with the aid of the Bank's research staff. Opera-
jectives, the kinds of studies undertaken leave     tional staff members are more interested in the
something to be desired. While this argument is     total stock of economic knowledge (based on
pertinent, there are two qualifications that        research done outside and inside the Bank) than
weaken it.                                          in the particular flow of research results that, at
,there is a significant degree of coupling     any given time, emanates from the Bank's
First, the Ba       research d   ofrcong            research efforts. This mismatch of incremental
between the Bank's research and operational         supply and current demand is then interpreted as
departments. Every research project that is fi-     arising from the research program's lack of oper-
nanced by the Research Committee is channeled       ational relevance.
through at least three separate review procedures,
at each stage of which operational staff are able   The stock-flow problem is undoubtedly a genuine
to influence and mould it. The Research Com-        one. To some extent, it arises because the demand
mittee itself comprises fifteen senior Bank staff,  for research applications and design/development
of whom about half are drawn from operations        studies outstrips the supply. Much has already
and half carry some responsibility for research     been said above about the need for a greater
management, with the Vice President, Develop-       effort in these directions, a need that was also
ment Policy, as Chairman.                           underscored in GRAP, and in respect of which
many initiatives have already been taken (such as
Secondly, i the past year, permanent Steerming      the allocation of a separate fund for research ap-
Groups have been established to determine           plication proposals, the systematic production of
research priorities (in the major research areas)   state-of-the-art reviews, etc.). However, it is well
for the Bank. To the extent that the other review   to note that the stock-flow problem can be ad-
procedures were inadequate, the guidance of the     dressed only marginally if one confines oneself to
Steering Groups (which are chaired by senior        chan      gin     ally if one research  elfrto
Bank operational staff) will help to ensure that    changing the composition of the research port-
the content of the Bank's research program is       folio. More fundamentally, the stock-flow prob-
relevant.                                           lem is one that can only be solved by either
reducing the time spent by Bank researchers on
The analysis of the preceding sections permits us   research per se and increasing their consulting
to claim that, prima facie, the Bank's research     activity for the operational departments, or by
program has, generally speaking, been operation-    adding new staff to perform the latter role, or by
pr~~~~ogrme cobnatiogneal ofaxg thee oprtwo.-
ally relevant. Why then is it not always perceived  some combination of the two.
to be so? And what remedial measures, if any,        A second problem is that of insufficient dissemi-
are implied by these perceptions? Three sets of      nation, the presence of which leads many non-
reasons suggest themselves, having to do, respec-    research staff to arrive at the judgment that the
tively with (   () the nature of the demand for      Bank's research is not operationally relevant.
research activity; (2) the nature of the supply of   This too is a genuine problem, and one that has
research activity; and (3) misperceptions of vari-   been repeatedly underscored in reviews of Bank
ous kinds.                                           research, including that of the General Research
The Demand*for Research Activity                    Advisory Panel. In coming to grips with this
problem, many new initiatives have been taken,
Certain characteristics of the nature of the de-    including training seminars for operational staff
mand for research support services may contrib-      on specific themes (such as "Management and
ute to the view that the Bank's research is not      Performance of Public Enterprises," "Employ-
operationally relevant. The first of these is what   ment Issues in Developing Countries," etc.) held
we may call the stock flow problem.9 As opera-       outside headquarters; state-of-the-art reviews;
tional staff see it, there exists a stock of research  and the publication of this bulletin thrice yearly.
9. 1 owe the conceptualization of this problem, and the termi-
nology adopted, to Gregory Ingram.
L: i O NORIl.) It XNK



l1 !-.. _1____ -----                    -__ -__ -     -     -                      ',i 1 x' "I   \ I I I
A third problem, related to the dissemination         In this category, one may finally refer to the
effort, is that of assimilation of research results by  problem of the genesis of research topics. There
operational staff. There are two views on this.      is a widespread belief among operational staff
One view is that, given the constraints on opera-     that the choice of research topics is influenced
tional staff, the only useful research is that which  more by researchers' interests than by the needs
is cost (time) saving. From this, it follows that     of the operational departments, and that the
innovative ways of doing things which involve         existing coupling arrangements only permit some
any increase in the efforts of operational staff are  tinkering with this predetermined supply of
not worthwhile, irrespective of their potential net   research proposals.
benefits. The other view is that, given the time
constraints on operational staff, the latter simply   Whether or not this is an exaggerated view
do not have the time to assimilate research           depends on one's judgments regarding the effec-
results, and unassimilated results are likely to be   tiveness of the coupling arrangements between
deemed not operationally relevant.                   research and the operational departments. How-
ever, the establishment of Steering Groups in the
It would seem that this last problem, too, calls      major research areas is partly a response to this
for greater efforts at research applications and      widespread belief, and bodes well for the future.
design/development projects. However, the as-        It is also likely that the activist role being
similation problem is likely to continue to exist     initiated by the research departments in soliciting
until nonresearch staff are afforded the time to     research questions from Regional Chief Econo-
address themselves to assimilation of research       mists and country economists will bear some
results.                                             fruit.
The Supply of Research Support Services              There is one other possible form of coupling that
has been mooted and that may bring a greater
There are also certain characteristics of the sup-   operational influence to bear on the genesis of
0ply of research that contribute to the perception   research topics, namely the institutionalization of
under discussion.                                    a greater involvement of research staff in opera-
tional work and vice versa. There are obvious
One of these is the problem of lags. The frequent    difficulties with carrying this too far, but there
long-time lags between basic and nonbasic            may be room for more of this form of involve-
research were referred to earlier. This explana-     ment than exists at present.
tion would suggest that the first eight years of
Bank research have been weighted in favor of ba-     Various Misperceptions
sic research which is now able to spawn a larger
number of research applications and design/de-       Finally, there are indeed some misperceptions at
velopment efforts than hitherto. This would im-      work, and we have identified two of these in pre-
ply that the operational relevance of the Bank's     ceding sections.
research should become increasingly obvious with
time.                                                The first of these is the problem of balance. As
noted earlier, there is a tendency to assume that
In addition, there is the question of the variance   all Bank research should be addressed to the ob-
of pay-off periods. This factor suggests that the    jective of supporting the Bank's operations rather
shorter gestation lags demanded of economic and      than to all four objectives of the Bank's research.
sector work lead staff engaged in it to treat        Of course, the question of the allocation of the
research activity as irrelevant to the Bank's work.  research effort over these four objectives remains
This also tends to be reflected in differences in    an open one, and the general feeling that design/
the work environments of research staff and op-      development studies should receive a greater
erational staff, causing some concern among the      share of the research effort has been emphasized
latter. As noted above, research activities have     above.
longer pay-off periods, and fewer deadlines. To
the extent that this is perceived as a problem, one  In addition, there is a possible misperception re-
can only hope that, with time, the relevance of      garding the nature of the research process in the
the research program will become clearer.            social sciences. The view that there can be defini-



12 RESEARCII NEWS                                                                        FALL 1980
tive answers to most questions may be true in the   rently being taken to implement these recommen-
physical sciences, but it is certainly much less    dations offered by GRAP. Other measures that
true in the social sciences which are characterized  were noted above, may be given further consider-
by essential complexity, institutions that vary     ation. What is certain is that, as in those
over time and place, and the importance of his-     industrial organizations in which good research
torical factors. Given this characterization,       departments have survived, the contribution of
research results akin to the "lessons of history"   research to the Bank's medium-term and long-
become important in their own right.                term goals has constantly to be kept to the fore.
Research is not solely a short-term activity, and
The preceding characteristics of the demand for     to judge it only on the basis of its contribution to
and supply of research activity in the Bank, and    immediate goals is to commit an error.
the misperceptions noted, lead us to a more qual-
ified view of the operational relevance of the
Bank's research. That is, while recognizing the
conceptual limits to judging the relevance of a
research program on the basis of the ex ante ob-
jectives of individual projects rather than ex post
evaluations, the operational relevance of the       COMPLETED                 RESEARCH
Bank's research may be described as being much
greater than commonly perceived. And, yet, it is
capable of being improved through greater efforts   Distributive Impact of Public
in (1) research applications and design/develop-    Expenditures
ment projects; (2) disseminating research results
(within and outside the Bank); (3) the assimila-    Ref. No. 670-96
tion of research results by operational staff. Also
of potential importance is a new form of coupling:  Because public expenditure accounts for such a
(4) a greater involvement of research staff in      large share of national income in developing coun-
operational work and vice versa, as a means of      tries-between 10 percent and 30 per cent-it has
influencing the genesis of research topics. And     important redistributive effects, whether intended
last but not least, it is important (5) to correct  or not. This also suggests that shifts in the pattern
the misperceptions that were noted previously.      of public expenditure may be an effective tool of
governments in aiding the poor.
F. A Concluding Note                                Where do the funds really go, and who benefits
The argument that the Bank conducts a research      the most? Detailed data invariably are sparse.
program that is more relevant to its operations     Much present analysis is based on statistics com-
than meets the eye rests essentially on (1) the     piled for other purposes. Frequently, the results
distribution of the research resources over the     are highly aggregated and the methodology is
program's three components of basic research,       awash with restrictive assumptions. The purpose
research applications, and design/development       of this research was to identify the beneficiaries
projects; (2) the distribution of the program's     of government services in two countries at similar
components over the objectives laid down for it;    per capita income levels-Colombia and Malay-
and (3) a proper understanding of the research      sia. The study covers the four sectors that have
process within the Bank. In attempting to recon-    large public expenditures which can also be allo-
cile the data in Tables I and 2 with the general    cated to households: education, medical care,
notion that much of the Bank's research is not      agriculture, and public utilities (for example, wa-
operationally relevant, various possible explana-   ter, electricity, sewage systems). In each country,
tions were offered and, where pertinent, remedial   two sets of estimates formed the basis for the
measures were suggested. To the extent that a       project: information on the costs of several major
greater effort at research applications (including  government programs and a specifically designed
design/development studies) and at disseminating    sample survey of the use of these services by
research results is warranted, measures are cur-    households.
MI"ORI.I) BANK



FAti.. 198X                                                                  RESIEAR(cI Nuws 1 3
The results of the household survey were ex-      outlays are highest for families in Selangor, in
panded to the national level and combined with     part because of the preponderance of postsecon-
the information on costs. Estimates were derived   dary schools in Kuala Lumpur. Broken down
of government spending per household in catego-   into communities, benefits per capita to Malays
ries of income, region, race, and other variables.  exceed the mean by 22 percent; those to Chinese
The factors explaining the distribution were      fall short of the mean by 30 percent. The public
identified.                                       costs of services to the lowest income quintile in
Malaysia are 15 percent below the mean, and
In both countries, the survey was able to trace   those in the highest quintile 10 percent above it.
the beneficiaries of one-third of total government  The costs of services to the remaining three-fifths
expenditures. In Colombia, major expenditures     of the population are very close to the mean. (Of
accounting for this fraction were the public sub-  course, this implies that the costs of services as a
sidies in the education and health sectors, and   share of income rapidly increase as income falls.)
investments in electricity, water supply, and sew-
age systems. In Malaysia, the segment of central   The distribution of services from Malaysian
government expenditure traced covered 56 per-      public utilities that are not subsidized shows a
cent of agriculture, health, and education bene-   distinct urban bias. Consumption of electricity
fits, distributed by household. The public utilities  and piped and treated water is weighted heavily
were also studied, although they receive no gov-  toward the Chinese, the cities, and the upper
ernment subsidies.                                 income families. Even so, the discrepancy be-
tween town and country is significantly smaller
In Colombia, expenditure on education IS distrib-  i   aasata      nClmi.A        otesbi
uted evenly across households, although subsidies  di Maayia th      an Coloi    to the, subs-
vary considerably at different educational levels.  dire servceso, suc  as l   dical  care, pi-
Expenditure for primary education is progressive,  mary education, and agricultural subsidies, the
e    xpenditure fr  r  y      X        *          poorest residents in the peninsula-the Malays,
expreas    ndiur forigher euation is      -       those who live in rural areas, and those who live
whereaso v mrrl ra, n toewh                                                      v
ally more unequally distributed than personal in-  in the North-benefit more than other residents.
come. The health subsidy is at relatively similar
rates across households. The National Health
System has a progressive effect, although the     This work, which was originally conceived as a
social security network favors middle-income       pilot study, indicates the merits of specifically
groups.                                           designed household surveys in assessing the dis-
tributive impact of large segments of government
The large difference in family size between rich   expenditure. Similar studies would be useful in
and poor households in Colombia also affects the   reflecting the distributive direction of government
findings. For instance, relatively equal health   investment, and the distributive consequences of
subsidies per household become heavily weighted   evolving government programs.
in favor of the richest segment of the community,
when expressed per capita. Low-income groups       A key aspect of the research was a methodology
benefit significantly less from expenditure on     to identify why particular households did not
sewerage, water supply, and electricity. Much of   consume particular services. Households that do
the explanation lies in an "urban bias" with al-   not consume services because of a lack of supply
most all consumers of public utilities concen-     are distinguished from those where the constraint
trated in towns.                                  is a lack of demand for a particular service. The
methodology then identifies the specific factors
In Malaysia, combined federal outlays per capita  that might have influenced the supply of or de-
for education, medical care, agriculture, and pen-  mand for the service. This is crucial because it
sions are highest in the rural areas, where        points out to what extent policies should encour-
Malays constitute the majority of the population,  age the supply of services to be expanded, or to
and lowest in the wealthiest state, Selangor,      what extent they should encourage factors to in-
where most people are located in, or near, the    fluence the demand, i.e., the utilization of exist-
national capital, Kuala Lumpur. Educational        ing services.
Vk *O)RLD BANK



14 RESEARCHI NEWS                                                                             FAI.A 198O(
The work in Colombia was carried out with the          affect the allocation of resources, and govern-
assistance of the Empresa Colombiana de Datos,         ments often intervene in their operations to influ-
and that in Malaysia with the aid of the Depart-       ence this process. Despite the importance of com-
ment of Statistics and other government offices.       mercial banks, there has been insufficient study
of why they behave as they do. Still less is
Further information may be obtained from Jacob         known about how government policy affects their
Meerman and Marcelo Selowsky in the Develop-           behavior.
ment Economics Department.
This research project undertook case studies of
Reports                                                three countries -Brazil, India, and Sri Lanka. It
describes the way banking systems mobilize re-
Meerman, Jacob. "Are Public Goods, Public Goods?"      .sorces        the   structurenofsthei liabilites
Public Choice 35 (1980).                               sources, and how the structure of their liablities
-  . "The Definition of Income in Studies of    changes. Bank portfolios are analyzed, and con-
Budget Incidence and Income Distribution." The Re-  clusions are drawn about the allocative decisions
view of Income and Wealth (December 1974): 515-     of each system-taking into account the bank's
522.                                                internal policies and the influence of the govern-
. "Do Empirical Studies of Budget Incidence  ment on their decision making. The analysis used
Make Sense?" Public Finance 3 (1978).               existing data to view the banking systems as a
.__         "The Household Distribution of Govern-     whole, as well as broken down into sectors ap-
ment Services for Agriculture and Education in Pen-  propriate for each country. In the Brazilian sur-
insular Malaysia in 1974." The Review of lncome     vey, data deficiencies dictated a less econometric
and Wealth 24 (June 1978): 161-176.                 approach than the authors had envisaged, and in
. "The Incidence of Sales and Excise Taxes,  the case of Sri Lanka, the absence of any well-
or Where Do We Put the Transfers?" Journal of Po-   de      sate    fo credit abctnle to a
litical Economy 88, no. 6 (December 1980).          defined strategy for credit allocation led to a
. "Income Redistribution in West Malaysia    qualitative-rather than quantitative -approach.
and Recent Empirical Work on Budget Incidence: A
Comment." Public Finance 30, no. 1 (1975): 131-     The Brazilian study incorporates material on the
135.                                                growth of financial assets within the economy,
. "Meeting Basic Needs in Malaysia: A        and on the structure of the assets and liabilities
Summary of Findings." World Bank Staff Working      of commercial banks. Since 1967, legislation has
Paper No. 260. April 1977.                          enabled the government to attempt to channel
. Public Expenditures in Malaysia, Who       credit toward particular sectors. The report
Benefits and Why? London: Oxford university Press,  assesses the success of these policies in directing
1979.                                               funds to agricultural production and export-
___---. "Public Services for Basic Needs in Malay-  oriented industry. It finds that banks cease to be
sia." World Development 7 (1979).                   innte     in an   inflathat banment         be
. "Subsidies of Major Malaysian Public       innovative i an Inflationary environment and
Utilities." Kajian Ekonomi Malaysia-Malaysian Eco-  become virtual agents of the government, which
nomic Papers (March 1974): 57-71.                   is the source of most of their funds. Even in this
Meerman, Jacob, and Shome, Partha. "Estimating Coun-   context, the philosophy which has been an un-
terfactual Incomes in Studies of Budget Incidence."  derlying factor in the banking system's evolution
Public Finance (1980, forthcoming).                 remains valid: banks prefer 1. -ge borrowers to
Selowsky, Marcelo. Who Benefits from Government Ex-    small and medium-sized ones
penditure? A Case study of Colombia. London: Ox-
ford University Press, 1979.                        Administrative costs remain prohibitive to small
borrowers, and banks have been slow to intro-
duce innovations that could help in lowering
Ref. No. 671-25                                        these costs. The Sri Lanka analysis concentrates
on the mobilization of resources by the banking
Size, geography, and an ability to mobilize de-        system. Estimates of the formation of assets by
posit resources all make the commercial banking        households and businesses between 1950 and
system the dominant financial entity in most           1974 provide a framework for assessing the be-
developing countries. Commercial banks critically      havior of the system as a whole and of classes of
IV WORLD BANK



d ......-                                                                     Rl.sf I   \R(  NII  S  1 5
banks. Sri Lanka's case illustrates bank behavior      For further information contact Vinayak V.
in many developing countries. The banking sys-         Bhatt and Shahid Yusuf of the Development
tem's resources are adequate for its foreseeable       Economics Department.
functions, but excessive preoccupation with short-
term financial considerations has stifled the crea-    Reports
tivity of Sri Lanka's bankers. The system lacks
the initiative to assume a promotional role-that       The following studies are available from the Public Fi-
is, to stimulate demands for credit to be used for     nance Division of the Development Economics Depart-
productive purposes.                                   ment:
By contrast, the Indian study showed the success       Bhatt, V.V. "Interest Rate, Transaction Costs and Fi-
of an ambitious policy of rapid branch expansion          nancial Innovations." Domestic Finance Studies No.
in urban areas. Small entrepreneurs and house-            47. January 1978.
holds (both farm and nonfarm) gained access to         Gandhi, J. "Bank Behaviour and the Control of Credit
banking services. The increase in the real resource       Flows: Lessons from the Sri Lanka Experience."
cost of the services was far smaller than antici-         Domestic Finance Studies No. 41. June 1977.
pated. Indeed, the report suggests it was less than    Patel, V.G. "Innovations in Banking: The Gujarat Ex-
the reduction in transaction costs to borrowers and       periments." Domestic Finance Studies No. 51. Au-
savers in the small.enterprises and the household         gust 1978.
sector. The Indian case study focuses on the im-       Saito, Katrine. "An Estimation of the Household Sector
pact of credit guarantee schemes on lending to            Savings of Sri Lanka." Domestic Finance Studies
smaller industries, and the cost of banks' inter-         No. 27. August 1976.
mediation role.                                        Thingalaya, N.K. "Innovations in Banking: The Syndi-
cate's Experience." Domestic Finance Studies No.
Two studies were also made of specific financial          46. January 1978.
innovation in India. Both have significance for        Yusuf, Shahid. "The Commercial Banking Industry in
developing countries that are devising policies to        Brazil 1964-74." Domestic Finance Studies No. 39.
promote small and medium-size enterprises. The            May 1977.
first study describes how the rapidly growing
Syndicate Bank of India, with its head office in a     Country Case Studies of Agricultural
village, has introduced new schemes to expand its      Prices, Taxes, and Subsidies
services for attracting savers and providing credit
in small amounts to small enterprises in agricul-      Ref. No. 671-42
ture and industry. The second study analyzes the
package of services (entrepreneurial training,         Developed countries tend to tax consumers and
managerial and technical guidance, and financial      industry, but to subsidize agricultural production.
support made available to small and medium-            In developing countries, the situation often is the
sized industrial enterprises by the Industrial In-     reverse: Agricultural producers are taxed, some-
vestment Corporation, the Industrial Develop-          times heavily, in order to subsidize urban con-
ment Corporation, and the State Finance Corpo-         sumers and manufacturing industry. There are
ration of the State of Gujarat.                        many forms of government intervention in this
area, including administered prices, taxes, subsi-
The research project was undertaken with the          dies, and quotas. Each may affect resource allo-
assistance of researchers at the Graduate School       cation, public savings, and economic growth, as
of Economics (EPGE) of the Fundacao Getulio            well as raise important equity considerations.
Vargas, Rio de Janeiro; the Indian Institute of       What happens to agricultural output (especially
Management of Ahmedabad; and with Professor           food production), income distribution, employ-
Jamshed Gandhi of the University of Pennsylva-         ment, and the balance of payments? What eco-
nia, who worked closely with the Central Bank         nomic change will be brought about by reducing
of Sri Lanka.                                          (or eliminating) intervention?
-     -___ _   -_         _ __  _ - - -  - . t )K l  I)  1B  \,.\k



16 HIAtxI ii \I N                                                                          FM
This project takes a broad view of these issues.     taken with the case studies includes desk studies
The use and consequences of administered prices,     of price elasticities of demand and supply and a
taxes, and subsidies are assessed in eight diverse   functional review of the treatment of agricultural
countries: Argentina, Arab Republic of Egypt,        prices and subsidies in the Bank's country eco-
Kenya, Mexico, Pakistan, Portugal, Thailand,         nomic and rural sector work, along with assess-
and Yugoslavia. Each country case study de-          ments of countries' experiences with varied agri-
scribes the way these policy instruments have        cultural price, subsidy, and tax regimes.
been used, based on the Bank's economic, agri-
cultural sector, and special study reports. Further   Reports
information is derived from official government
publications and from field investigations. The       Bertrand, Trent. "Thailand: Case Study of Agricultural
project investigates means of evaluating the con-        Input and Output Pricing." World Bank Staff Work-
sequences of government policy through the use           ing Paper No. 385. April 1980.
of border prices, which define the opportunities      Cuddihy, William. "Agricultural Price Management in
afforded by world trade, as a frame of reference.        Egypt." World Bank Staff Working Paper No. 388.
May 1980.
A number of indicators are computed to test the       Gotsch, Carl, and Brown, Gilbert. "Prices, Taxes and
effect of intervention measures. These include the       Subsidies in Pakistan Agriculture, 1960-1976."
nominal protection coefficient, the effective pro-       World Bank Staff Working Paper No. 387. April
tection coefficient, the domestic resource cost, and     1980.
the net economic benefit. Various other indicators    Reca, Lucio G. "Argentina: Country Case Study of Agri-
are computed in individual case studies. Implicit        cultural Prices and Subsidies." World Bank Staff
social welfare transfers associated with the devia-      Working Paper No. 386. May 1980.
tion from world prices are estimated for 24 com-      Scandizzo, Pasquale L., and Bruce, Colin. "Methodolo-
modity cases drawn from the case studies. In             gies for Measuring Agricultural Price Intervention
some cases, the net social loss, due primarily to        Effect." World Bank Staff Working Paper No. 394.
the policy-induced reduction in potential interna-       June 1980.
tional trade, exceeds 10 percent of the value of
production and consumption. However, welfare          Consequences of Risk for Agricultural
transfers appear to be much larger than the           Plc
aggregate social loss to the economy. In one-third    Pe. oli6cy  4
of the cases studied, the welfare gain (or loss) ofReNo67-4
producers is more than 10 times as large as the       This research investigated some of the conse-
net social loss in production. With positive pro-    qecso       rdcinrssi         giutr     o
tection in only three cases, the producers incur      marete efficiencyiand thek inee  foricgovuefrnmn
welfare losses in the remaining 21 instances-a        intrventefiontoinuencen the  suppl fof ovrmmodi
dramatic example of how heavily agriculture is        itiresnwhose producionflec is e rsky. lyConsiderable
penalized in developing countries. Both con-          te   hs    rdcin' ik.Cnieal
sumers and government revenues benefit from the       research has been undertaken in recent years on
welfare losses of producers, with the consumer       the importance of price and yield risks in deter-
gain generally exceeding that of the government,      mining farmers' decisions. Risk considerations
have been shown to affect cropping patterns, the
The study includes a summary of available infor-      volume of output, levels of pesticide and fertilizer
mation on producers' responses to changes in          use, and the rates and level of adoption of new
product and factor prices, and in factor availabil-   technologies. This is partly because farmers are
it.In the cases of Mexico and Portugal, linear      averse to risk, but also because it is difficult to
programming models of the agriculture sector are      forcas yilsadpie.oretyi            nucr
used to assess the indirect as well as the immedi-    tamn environment. Surprisingly, little work has
ate impact of taxes and subsidies, and to test the    been done on the implications of these findings
legitimacy of the assumption that indirect effects   for the efficient working of markets in the aggre-
may omeimesbe gnord i les deaile  anly-gate. This study attempts to fill part of that void.
sis. The extensive supplementary work under-



The research focuses on how different types of       of different levels of risk aversion on the part of
price forecasting behavior determine the behavior    farmers. This study showed that the social cost of
of aggregate supply over time, and thence the        using forecasts of prices rather than per unit rev-
efficiency of markets. The authors of the research   enue may be quite large, but that this cost falls
study undertaken at the Bank's Development Re-       rapidly as farmers become more risk averse. This
search Center, Jock R. Anderson, Peter B.R.          is because risk-averse behavior leads to more di-
Hazell, and Pasquale L. Scandizzo, contend that      versified cropping patterns in the Mexican situa-
unless farmers properly take account of the cor-     tion, thereby reducing the production of risky
relation between prices and yields in forming        crops in much the same way as does the use of
their price forecasts, competitive markets will      the expected per unit revenue forecasting.
often be socially inefficient. Social welfare may
then be considerably advanced through appropri-      The authors argue that the importance of market
ate intervention policies aimed at reducing the      inefficiencies in risky situations essentially de-
supply of commodities that are risky to produce      pends on the variability of yields, the nature of
and increasing the supply of less risky ones.        farmers' price forecasting behavior, the degree of
private risk aversion, and the elasticity of de-
In general, a simple and practical price forecast-   mand. Where markets are very inefficient, a gov-
ing rule for farmers to follow which leads to effi-  ernment can choose between three basic policy
cient market behavior is the expected per unit       approaches. First, production quotas could be
revenue (i.e., the average value of the gross        imposed to limit the average output levels of
return per hectare divided by the average yield      risky crops to their socially desired norms. Sec-
per hectare). This forecast adequately captures      ond, market information services could be estab-
the correlation between price and yield, and may     lished or improved to help farmers attain their
be easily calculated from historical data available  expected revenue per unit of output. Third, the
to the farmer. In many cases, it is also the price   government could attempt to reduce br eliminate
forecast which is optimal from the farmers' point    risks from the market through price stabilization
of view. The analysis expands on earlier work        schemes. The latter approach has the advantage
published by Hazell and Scandizzo in the Ameri-      of exploiting opportunities for pooling risks over
can Journal of Agricultural Economics. Theoreti-     space and time, thereby increasing the socially
cal results are generalized from a partial equilib-  desired output of risky commodities. However, on
rium analysis of a single linear market to non-      the basis of a general simulation study, the
linear multimarket and general equilibrium situ-     authors found the social return from price stabili-
ations. The authors have also attempted to deter-    zation to be rather small, when farmers plan on
mine the empirical relevance of their work and       the basis of expected per unit revenue. Where
the effectiveness of certain kinds of policy inter-  less appropriate price forecasting is pursued, the
ventions.                                            social gain from stabilization may be substantial,
but in those cases the largest part of the gain
Time-series data from a wide range of countries      may be more easily attained through improved
were used to test the hypothesis that market sup-    market information services.
ply is influenced by the covariance of price and
yield as well as by lagged market prices. While      This research has implications for the way in
conclusive statistical results proved elusive,       which econometric models of agricultural supply
Anderson in collaboration with Professor Richard     should be specified, and for the formulation of
Just of the University of California, Berkeley,       agricultural sector models in which the prices of
provides evidence that farmers in industrialized     risky commodities are endogenized. The authors
western economies do take account of price and       also find that for a wide class of welfare func-
yield correlations, whereas farmers in developing    tions, the expected per unit revenue is the rele-
countries and in the centrally planned economies     vant shadow price to use in the appraisal of risky
do not.                                              markets. Shadow pricing in this way will tend to
penalize risky projects compared with the more
An agricultural sector model of Mexico was also       conventional approach of using forecasts in mar-
used to evaluate the social cost of incorrect price   ket prices.
forecasting behavior, and to determine the effect
.. .. .. .. ., , ....................... . . ........ -- -- . ................ -- .......- - - - - -



II SLAW 1R Ni lSs                                                                       FALI 1980
The authors conclude that their study has merely      tion alone, while three countries (Upper Volta,
scratched the surface of an important and largely      Ghana, and Mali) experienced a net loss of 1.2
untouched area of research. While some initial         million. About 4.4 million people migrated
policy implications have been explored, more           within their respective countries. For both inter-
work of both a theoretical and applied nature is       national and internal movements, the direction of
required before the full implications of the study     the flow-from  the interior toward the coast and
can be assessed.                                       from rural to urban areas-affects the regional
development of the individual countries and con-
Reports                                               tributes to the problems of urban growth.
Anderson, J.R.; Hazel, P.B.R.; and Scandizzo, P.L.     For these reasons, the World Bank and the Or-
"Considerations in Designing Stabilization Schemes."  ganisation for Economic Co-operation and Devel-
American Journal of Agricultural Economics (Decem-  opment (OECD) agreed, in 1975, to sponsor a
Hazell, P.B.R; Norton, R.; Parthasarathy, M.; and Po-  research project to investigate regional migration
mareda, M. "The Importance of Risk in Agricul-     patterns and trends i Western Africa. The
tural Planning Models." World Bank Staff Working   research was directed by K.C. Zachariah of the
Paper No. 307. November 1978.                      Bank's Development Economics Department,
Hazell, P.B.R., and Scandizzo, P.L. "Farmers Expecta-  with the assistance of Julien Conde of the
tions, Risk Aversion and Market Equilibrium Under  OECD. It is the first such study to be attempted
Risk." American Journal of Agricultural Economics  in this region.
(February 1977).
. "Market Intervention Policies When Pro-   The nine countries included in the research proj-
duction is Risky." American Journal of Agricultural  ect were selected according to the availability of
Economics 57 (November 1975).                      recent reliable census or survey data on migra-
Roumasset, J.A.; Boussard, J.M.; and Singh, 1. "Optimal  tion. The governments of The Gambia, Ghana,
Price Intervention Policies When Production is               g
Riky"Rik Ucetinyand Agricultural D7eo-Ivory Coast, Liberia, Togo, and Upper Volta
Risky." Risk, Uncertainty an  giutrlDevelop-       provided unpublished tabulations from their
ment. Agricultural Development Council, New York,  lates cnsuses     her analyeb research-
U.S.A.                                             latest censuses, which were analyzed by research-
Scandizzo, P.L. "Project Evaluation in Risky Markets."  ers in the region and at the Bank and OECD.
Paper presented to the International Federation of  For some countries, notably Mali and Senegal,
Operational Research Societies, special conference on  data from recent censuses are not yet available
Operations Research in Agriculture, Jerusalem, Is-  though survey data are. Supplementary informa-
rael (November 1979).                              tion on emigration from several other countries
. "Risky Agricultural Markets: Expectations,  in the region was available through data supplied
Welfare and Intervention." Draft monograph. Devel-  by the countries of destination. Country reports
opment Research Center, World Bank (1979).         were prepared for The Gambia, Ghana, Ivory
Migration in Western Africa                            Coast, Liberia, Mali, Senegal, Sierra Leone,
Togo, and Upper Volta. These studies reviewed
Ref. No. 671-26                                        national demographic background, international
migration, rural-urban and interregional migra-
Migration patterns across and within national         tion, and the socioeconomic characteristics of the
boundaries affect the distribution of resources        migrants. A regional analysis based on the find-
especially human ones. When, as in Western            ings of the country studies is being published this
Africa, millions of people are involved, the eco-      fall.'
nomic and social effects of migration are signifi-
cant for the development of both the countries of
origin and destination. Data from the 1960 and          I. K.C. Zachariah and Julien Conde, Migratzon in West Af-
1970 series of censuses and surveys in the region     rica: Ihe Demographic Aspects (London: Oxford University
indicate the magnitude of the flow. From               Press, 1980, forthcoming).
1965-75, six countries (Ivory Coast, Senegal,
Sierra Leone, Togo, Liberia, and The Gambia)
had a net gain of 1.3 million people from migra-



9           FAILL. 1980                                                                     R'SEARtCl NE'S 1t9
The project finds that while the populations of      long-distance migrants include more males. The
the nine countries vary in size and growth rate,     dependency ratios of migrant groups in Western
they have fairly similar fertility rates, mortality  Africa are increasing over time, and the average
rates, and rates of natural increase. The growth     duration of residence of migrants at destination is
rates range from 1.7 percent annually in Upper       also increasing. These factors indicate an increas-
Volta to 6.0 percent in Ivory Coast; these differ-  ing tendency for migrants to settle permanently
ences are caused by migration. The data avail-      in the host country or area of in-migration.
able were not adequate to analyze the causes and
consequences of migration, but the study noted a     For further information, contact K.C. Zachariah.
general positive relationship between the eco-
nomic status of a country and the rate of lifetime   Reports
immigration. A similar positive relation was ob-     Zachariah, K.C. "Demographic Aspects of Migration in
served between regional economic conditions             West Africa." World Bank Staff Working Paper No.
within countries and the rate of in-migration,          414 -"Volume I: English-Speaking Countries."
while out-migration rates were negatively corre-        World Bank Staff Working Paper No. 415 -"Vol-
lated with prevailing economic conditions.              ume Il: French-Speaking Countries." August 1980.
Zachariah, K.C., and Conde, Julien. Migration in West
The total number of foreign nationals living in         Africa: The Demographic Aspects. London: Oxford
the nine countries was 2.8 million in 1975, or          University Press, 1980 (forthcoming).
about 7 percent of these countries' population.      "International Migrations in West Africa: Demographic
Until recently, Ghana was the major country of          and Economic Aspects." Published in Economic and
immigration in Western Africa, but economic             Demographic Change: Issues for the 1980s 3 (1978):
and political instability have decreased its flow of    237-46 (International Union for the Scientific Study
migrants over the last decade. Ivory Coast is now       of Population, Helsinki).
the principal recipient of immigrants, coming                                          X
mainly from Upper Volta, Mali, Guinea, and
Togo. For Upper Volta, emigrants represented as
much as 17 percent of the population, while im-      NEW        RESEARCH
migrants contributed as much as 21 percent of
the population in Ivory Coast.                       Brazilian Income Distribution and
Growth
In most countries, the principal areas of in-
migration are the areas in which the capital city    Ref. No. 672-21
is located. Between 1965 and 1975, net rural-ur-
ban migration in the region was 1.7 million, with    Brazil's rapid growth in recent years is univer-
the total growth of the urban population during      sally acknowledged. There is less agreement on
the period at 3.6 million (about 5.8 percent a       how the benefits of that growth have been dis-
year). A little less than half the growth was from   tributed, and on whether the poor have gained
rural-urban migration and the balance from a         much in absolute terms. Until now, data have
natural increase. One of the most direct economic    been difficult to interpret. In the absence of any
consequences of migration is the redistribution of   common interpretation, there has been minimal
the working population. The estimated total em-      progress in understanding the social and demo-
ployment in 1975 in eight of the Western Afri-       graphic factors which have led to Brazil's com-
can countries was 11.5 million. Of these, 1.3 mil-   bination of growth and inequality. Much of this
lion were immigrants and 2 million were lifetime     may now change, thanks to the release by the
interregional migrants. Thus, excluding short-       Brazilian Census Bureau of a one percent public
distance migrants, about 3.3 million workers or      use sample of the 1970 population census. The
almost 30 percent of the labor force were migrants.  sample covers 910,000 individuals cross-classified
into 176,000 households and 117 identifiable geo-
As in other parts of the world, migrants in West-    graphic areas throughout Brazil. The sample is
ern Africa include a relatively high proportion of   unusually large and affords an opportunity to
young adults of working age; short-distance mig-     obtain much detailed information about the links
-           rants include more females than males, while         between crucial variables.
0



2() Ii>\1 sl: s   '.i Nk '.                                                            FALL 1980
This research is really a combination of three     of income and social development, and coincided
projects working with the census material, in an   with the introduction of the family planning pro-
attempt to delineate some of the links between     gram. The Indonesian Government has set ambi-
demographic variables at the household level and   tious targets for further fertility reduction. Two
income distribution. As a group, the three proj-   priorities are to maintain the impetus of the
ects represent a first step toward modeling the    Java-Bali family planning program, and to see if
interactions between macroeconomic growth and      its success can be repeated on the Outer Islands.
structural change on the one hand, and changes
at the household level in labor supply and hu-     This research will attempt to understand the
man capital accumulation on the other. In addi-    process of fertility decline and to plot future
tion, it is hoped that the research will establish a  growth rates for the Indonesian population. It
baseline for comparison with the 1980 census.      will appraise the impact on fertility of changes in
marital patterns, incomes, education, labor force
Project One focuses on the relation of household   participation of women and children, migration,
size and composition (age structure, age of the    urbanization and modernization, and, of course,
head of household, work status and members,        the provision of family planning services. Target
and so forth) to inequality and on the way in      populations with high fertility will be identified
which households have responded to economic        by socioeconomic characteristics and location; and
pressures-many of which have increased as          policy options other than family planning will be
growth has become more unequal.                    weighed.
Project Two focuses on schooling decisions, and    Although Indonesian aggregate income has
the way in which poverty can be transmitted        grown very rapidly in recent years, there have
across generations, either because education is not  been doubts about the equity of its distribution.
available or because poor children must work to    Urban-rural disparities are growing, and overall
supplement family incomes. The relative impor-     inequality appears to be increasing. With 50 mil-
tance of family characteristics as opposed to      lion people unable to maintain the low consump-
school availability will be weighed on the basis of  tion level of US$90 a year, the problem of abso-
the census data.                                   lute poverty is severe. The project will attempt to
establish broad trends in consumption levels, up-
Project Three looks at infant and child mortality  dating earlier Bank research by taking into ac-
and its links with income distribution. The cen-   count the new Multipurpose Household Survey.
sus material provides an opportunity to isolate    It will also define the poor in terms of various
social, economic, and demographic characteristics  characteristics: demographic patterns; consump-
by household. This makes it possible to explore,   tion levels and caloric intake; housing conditions,
for instance, the relative importance of material  access to social services; concentration by region;
education, family size, type of water system,      urban/rural residence; labor-force participation;
house size, and sanitary arrangements in securing  income-earning activities and ownership of assets.
infant and child health.
The second round of the Intercensal Population
For further information contact Nancy Birdsall     Survey, 1976 (SUPAS II) and the Multipurpose
or Constantino Lluch in the Bank's Development     Household Survey (SUSENAS 1978/79) will
Economics Department.                              provide basic data. They provide household-level
Poverty, Fertility, and Human                      information on the socioeconomic characteristics
of household members, living conditions, incomes
Resources in Indonesia                             and consumption expenditures, fertility behavior,
and family planning practice. Data from these
Ref. No. 672-19                                    sources will be supplemented with community-
level information on the availability of family
This project traces the links between poverty and  planning and other government services.
fertility in Indonesia. Between 1971 and 1976,
fertility rates on Java and Bali fell significantly.  For further information contact Oey Astra Mee-
The reduction took place at relatively low levels  sook in the Development Economics Department.
L -  .i O\ \   ) R  i)  ! i  h--                       - - - - - - . . . . . ....  --_ _ _ - _ - . - _ _ _ - - . - - - - - - - - - _ _ _ _



Case Studies of Determinants of                     fertility and marriage measures on the one hand
Recent Fertility Decline in Sri Lanka               and such independent variables as personal,
household and community level characteristics on
and South India                                    the other.
Ref. No. 671-70                                     The project is being carried out in collaboration
e s n dwith the Department of Census and Statistics,
her signic     del ine  fri   level among          Government of Sri Lanka, Colombo; the Bureau
lower incoe groupsedconsiderabe intLnkatand        of Economics and Statistics, Kerala Government,
India has aroused considerable international       Tiadu;adteIsiuefrSca                  n
interest. Between 1960 and 1974, the birth rate      conaomcd Chandge Bnga   eor K arnat   C
in Sri Lanka fell from 37 births to 27 births per   Zacharia of the Bank's Dlopment EconomniC
thousand population. In some parts of Kerala,       Departis              of theestudy, assisd
the rate fell to 22 per thousand or even lower.     NKNamboodiri, Unive     of  North aroina
Income per capita in Sri Lanka is US$190 a
year, and in Kerala is even below the all-India     and A. Thavarajah and S.L.N. Rao of the
o    a       . (  M   ,      -United Nations Fund for Population Activities,
average of US$ 180 a year. (In Mexico, by con-      New York.
trast, the birth rate is around 40 per thousand at
an income level of over US$1,000 per person).      Improved Technology for Animal-
Fertility declines in such poor populations are    Powered Agriculture
virtually without precedent. Furthermore, neither  Ref No. 672-17
area has spent large sums on family planning.
Isolating successfully the factors that contribute  The capital costs of equipment, combined with
to these declines may be of considerable help in   rising prices for petroleum-based fuels, are a
designing population projects for developing       major impediment to mechanization among small
countries, and in identifying components of other  farmers in poor countries. Although the creation
schemes-rural development projects, for in-        of employment in rural areas is often an impor-
stance-which create a demand for family plan-      tant development objective, the availability of
ning services. The development policies of the     energy for agriculture still needs to be increased
two areas may shed some light on the subject.      throughout the developing world to realize higher
Public services have become increasingly avail-    yields and production.
able to the majority of the population at low cost
or, indeed, free. Government policy has also led    For many small-scale farmers in large areas of
to a significant redistribution of wealth in favor  Asia, Africa, and the nonoil producing Middle
of the very poor. Some of the traditional eco-      East, the draft animal continues to have a key
nomic rules of thumb in these societies may no     role in production-and one that may well grow
longer apply. For example, the balance of bene-    more significant. In countries that still rely prin-
fits in marrying young and adding extra children   cipally on human labor in agriculture, the use of
has tilted: the economic advantage of a large      animals will increase. In the countries where ani-
family is now negligible.                          mal power is widely used, techniques of improv-
e p t w              ing its efficiency need to be developed. These
The project will test this hypothesis i different  include better feeding and care, selective breeding
ways. In Sri Lanka, available data from the 1975   and genetic upgrading, the improvement of tradi-
World Fertility Survey have been supplemented      tional equipment-yokes and other implements
by more up-to-date and detailed information on     -      t
fertility, family planning, household income and   equipment. Efforts to develop such improved
expenditure, and other factors influencing demo-    eques      E    be  deren imayoue        -
graphic variables. In South India-Kerala and       tries, ut relatvel lttle has been done to dssem-
Karnataka-existing data are inadequate and         miate the results.
sample surveys have been conducted, which are
now being analyzed. Statistical techniques will be  This research project aims at evaluating the
used to test multivariate relationships between    "shelf" of improved implements and associated
technology against traditional technologies and



.ti M ii O \X '                                .. FALL 1980
those still in the process of development. The     These data will be collected for six countries and
mechanisms for the adoption of improved imple-    the results will be assessed against the ICP Phase
ments will be studied, both within particular      IV benchmark study. Tentatively, the countries
farming systems and among different systems.       examined will be Colombia (or Jamaica), India,
Efforts will be made to determine the most im-     Indonesia, Kenya, Nigeria and Tanzania. Other
portant incentives which lead farmers to switch    possibilities are Bangladesh and Australia. In
to improved equipment, and what constraints        addition, the European Economic Community is
they and the equipment producers face. Initially,  conducting (and financing) similar surveys in
the study will focus on India, where animal-       Morocco, Senegal, and Tunisia. Finally, it is
powered agriculture is widespread and where        hoped consultants will be engaged to make fur-
considerable experience exists with efforts to im-  ther surveys in Guatemala, Costa Rica, Panama,
prove traditional technologies. The Indian Insti-  and, possibly, the Dominican Republic. All the
tute of Management in Bangalore will assist in     material will be collected and analyzed at the
the implementation of the study.                   World Bank.
For further information contact Peter Brumby,      The project will produce a report on the metho-
New Delhi Office, World Bank.                      dology of reduced information comparisons, as
well as estimates for the sample of countries.
Reduced Information Methods of Real                Further reports will be issued on the integration
Income Comparisons                                 of reduced information results with other
regional and global estimates.
Ref. No. 6 72-16
For further information contact Sultan Ahmad in
It is widely agreed that relative levels of income  the Economic Analysis and Projections Depart-
in different countries should be compared with     ment.
national income converted to a common currency,
using purchasing power parity rather than          Pricing of Indigenous Energy Resources
exchange rates. However, the time and effort
required to make reliable estimates of purchasing  Ref. No. 672-15
power parity (PPP) mean that such data are
quite expensive and available for relatively few   One of the major constraints on development, in
countries,                                         recent years, has been the supply and price of
energy. In the last years of the 1970s, crude oil
The World Bank has supported the International     prices doubled to almost US$30 a barrel after
Comparison Project (ICP) for the past 11 years.   registering substantial increases in 1973 and
The ICP has laid the foundations for a world-      1974. The oil import bill of the developing coun-
wide income comparison system, producing PPP       tries has increased more than tenfold since 1972.
estimates for 34 countries-by 1985, it will cover  The domestic price of energy products is a key
over 60 nations.                                   Tedmsi       rc   feeg     rdcsi      e
variable both in stimulating indigenous produc-
It is believed that estimates approaching the     tion and in managing domestic demand. Yet, no
quality of the ICP comparisons can be produced     general framework has been developed for pric-
with much less than the full set of data currently  ing indigenous energy resources. Past research
being used. This project will examine three        has concentrated on electricity pricing, and is not
reduced-information approaches to estimating       directly applicable to tradeable and/or depletable
purchasing power parity. The first involves        energy products which have an internationally
working with data regularly collected by national  determined opportunity cost.
statistical organizations for their published price
indices. The second selects a small sample of      This present research project will draw together
prices based on expert judgment. The third de-    the lessons of available literature and of country
termines analytically, for instance with multiple  case studies to evaluate the role of energy pricing
regression, the best subset of items that will pre-  pomlcies in achievIng both sectoral and develop-
dict PPP at various levels of GDP aggregation.     mental objectives. It has two specific purposes:
r-                                    ------
r72;  ;



first, to develop and test approaches to deriving  NEW   BOOKS
"economic" prices which take into account the
tradeable and depletable nature of some energy     Egypt: Economic Management in a
resources; and second, to explore the scope for    Period of Transition
establishing market prices for energy products
that further fiscal, social, or other development  Khalid Ikram and others
objectives, even though they may differ from
"economic" prices.                                 The Johns Hopkins university Press, 1980
(forthcoming). About 568 pages.
The project will begin with a literature review
and analysis, based on particular aspects of en-   LC 80-552
ergy pricing that are significant for developing   ISBN 8018-2418-4     $32.50 hardcover
countries-for example, valuing natural gas in            8018-2419-2    $10.50 paperback
the ground or estimating the opportunity cost of   Egypt's relatively fixed amount of usable land
firewood. This will be followed by country case    Egypts rapidly  fing pountion usabe con-
studies that will derive economic and market       and its rapidly growing population must be con-
prices tailored to the energy resource endowment   sidered in any discussion of this country's eco-
and developmental objectives of each country.      nomic prospects. So, too, must the legacy of past
Aand heveslopmentl objectves fof acrecount.        international hostilities, which have left much of
Approaches will raLnge from macroeconomicthpysclnfaruuefodvlpmt
modeling exercises to a product-by-product anal-   the physical infrastructure for development
ysis of elasticities and substitution effects. Guide-  neglected and deteriorating. Nonetheless, Egypt
lines will be proposed for evolving appropriate    has considerable potential for long-term economic
short-term and medium-term energy pricing          growth, based on earnings from petroleum,
strategies for particular products in developing   revenues from the widened Suez Canal, rapidly
countries,                                         increasing remittances from EgyptiaAs working
abroad, opportunities for tourism, cultural and
For further information contact DeAnne Julius      political links with capital-surplus Arab coun-
in the Energy Department.                         tries, the accessability of European markets, and
a reasonably well-trained labor force.
To realize this potential will require policy re-
forms and structural changes, since not all exist-
ing policies and institutions are appropriate for
the new "open door" policy and more market-
NEW       AND                                      oriented economy. Strategic choices concern the
quantity, quality, and proportion of resources to
F ORTHCOMVIING                                     be allocated to development and to economic
PUBLICATIONS                                       management.
The full range of World Bank publications is de-   This study is the most detailed examination of
scribed in the Catalog of World Bank Publica-     the Egyptian economy to appear since the mid-
tions, issued annually. The new books listed be-   1960s, and the first to emphasize economic man-
low, which are published by outside publishers     agement and policy. It examines issues related to
for the World Bank, are obtainable through book-   population, human resources, and the main prod-
sellers or by writing to the publishers. Prices are  uctive sectors; reviews the evolution of financial
subject to change. The other items listed, and the  resources; and describes issues of the physical
Catalog, are available free of charge from:        mfrastructure and mfrastructure mvestment.
Publications Unit
World Bank
1818 H Street, N.W.
Washington, D.C. 20433, U.S.A.



-  di 'l \;it1 1 Ni  <                                                             F .A   1980i
Urban Labor Markets and Income                    WORLD BANK STAFF WORKING
Distribution: A Study of Malaysia                 PAPERS
Dipak Mazumdar                                    No. 379. Managing Information for Rural De-
velopment: Lessons from Eastern Africa.
Oxford University Press, 1980 (forthcoming).      Guido Deboeck and Bill Kinsey, Agriculture and
About 320 pages.                                  Rural Development Department. March 1980.
ISBN 0-19-520213-9   $18.95 hardcover             No. 380. Women in the Urban Labor Markets
0-19-520214-7   $ 7.95 paperback            of Africa: The Case of Tanzania. Nwanganga
Shields, Development Economics Department.
Thils book is a detailed case study of the working  April 1980.
of the urban labor market and the factors affect-   P
ing the distribution of urban income in Peninsu-  No. 381. Some Aspects of Wheat and Rice
lar Malaysia. Its main sources are two surveys    Price Policy in India. Raj Krishna and G. S.
conducted by the World Bank in 1973 and 1974      Raychaudhuri (consultants), South Asia Pro-
with the active cooperation of the Government of  grams Department. April 1980.
Malaysia; it also makes use of a wide variety of
published government statistics pertaining to the  No. 382. Trends in Rural Savings and Private
early 1970s. The analytical apparatus of the      Capital Formation in India. Raj Krishna and
study, as well as the nature of the primary data  G. S. Raychaudhuri (consultants), South Asia
collected, will be of much interest to those      Programs Department. April 1980.
designing future studies of labor markets in the
developing world.                                 No. 383. Fertility and Its Regulation in Bang-
ladesh. R. Amin (consultant) and R. Faruqee,
Although Malaysia is a member of the middle-      Development Economics Department. April
income group of developing countries, the prob-   1980.
lems of the growing labor market and of relative
poverty in this sector are of general concern in  No. 384. Small-Scale Enterprises in Korea and
the study of the process of development, for aca-  Taiwan. Sam P. S. Ho (consultant), Development
demics as well as policy makers. The importance   Economics Department. April 1980.
of the rural-urban migration in the process of
urbanization, the coexistence of a modern and a   No. 385. Thailand: Case Study of Agricultural
traditional sector in the urban labor market, and  Input and Output Pricing. Trent Bertrand
the significance of formal education in urban     (consultant), Agriculture and Rural Development
earnings and unemployment, are examples of        Department. April 1980.
such common features. The concentration of        No. 386. Argentina: Country Case Study of
relative poverty in a particular racial group of  Agricultural Prices and Subsidies. Lucio G.
the country is also a feature of many developing          Agriculturar nS    d Rucil G.
countries. There have been several studies on     opment Department April 1980
specific aspects of the urban labor market in
some developing countries, but this study is prob-  No. 387. Prices, Taxes and Subsidies in Paki-
ably the first to attempt an integrated analysis of  stan Agriculture, 1960-1976. Carl Gotsch (con-
interrelations. Starting with the distribution of  sultant) and Gilbert Brown, Policy Planning and
income among urban households, it works its       Program Review Department. April 1980.
way from the determinants of employment rates
of different types of household members to the    No. 388. Agricultural Price Management in
major factors affecting personal earnings.        Egypt. William Cuddihy, Agriculture and Rural
Development Department. April 1980.
' .  .WiI                   . l



No. 389. Poverty and Growth in Kenya. Paul       No. 400. Rural Poverty Unperceived: Prob-
Collier (consultant) and Deepak Lal, Develop-    lems and Remedies. Robert Chambers (consul-
ment Economics Department. May 1980.             tant), Development Policy Staff. July 1980.
No. 390. The People of Bogota: Who They          No. 401. Levels of Poverty: Policy and
Are, What They Earn, Where They Live.            Change. Amartya Sen (consultant), Development
Rakesh Mohan, Development Economics De-          Policy Staff. July 1980.
partment. May 1980.
No. 402. Education and Income. Edited by
No. 391. A Dynamic Simulation Model of the       Timothy King, Development Economics Depart-
World Jute Economy. Jock Anderson, Charles       ment. July 1980.
Blitzer, and Tom Cauchois, Development Re-
search Center, and Enzo Grilli, Economic Analy-  No. 403. Implementing Programs of Human
sis and Projections Department. May 1980.        Development. Edited by Peter T. Knight, De-
velopment Economics Department. Prepared by
No. 392. Shadow Prices for Project Appraisal     Nat J. Colletta, Eastern Africa Projects Depart-
in Turkey. Afsaneh Mashayeki (consultant),       ment; Jacob Meerman, Development Economics
Europe, Middle East, and North Africa Country    Department; Milton Esman, Joseph Mbindyo,
Programs Department I. May 1980.                 John Montgomery, Everett M. Rogers, Constan-
tina Safilios-Rothschild, and Norman Uphoff
No. 393. Food Security in Food Deficit Coun-     (consultants), Development Policy Staff. July
tries. Shlomo Reutlinger and Keith Knapp (con-   1980.
sultant), Development Economics Department.
June 1980.                                       No. 404. Population and Poverty in the De-
veloping World. Nancy Birdsall, Development
No. 394. Methodologies for Measuring Agri-       Economics Department. July 1980.
cultural Price Intervention Effects. Pasquale
L. Scandizzo and Colin Bruce. Agriculture and    No. 405. The Effects of Education on Health.
Rural Development Department. June 1980          Susan H. Cochrane, Development Economics
Department, and Donald J. O'Hara and Joanne
No. 395. Income Distribution and Poverty in      Leslie (consultants), Development Policy Staff.
Mexico. Joel Bergsman, Latin America and the     July 1980.
Caribbean Regional Office. June 1980.
No. 406. Poverty and the Development of
No. 396. Structural Change in Trade in Man-      Human Resources: Regional Perspectives.
ufactured Goods between Industrial and De-       Willem Bussink, East Asia and Pacific Country
veloping Countries. Bela Balassa, Development    Programs Department; David Davies, Western
Research Center and The Johns Hopkins Uni-       Africa Projects Department; Roger Grawe, South
versity. June 1980.                              Asia Country Programs Department; Basil
Kavalsky, Europe, Middle East, and North
No. 397. Human Factors in Project Work.          Africa Country Programs Department I; and
Heli Perrett (consultant) and Francis J. Lethem,  Guy Pfeffermann, Latin America and the Carib-
Projects Advisory Staff. June 1980.              bean Regional Office. July 1980.
No. 398. Publishing for Schools: Textbooks       No. 407. Human Resource Development and
and the Less Developed Countries. Peter H.       Economic Growth in Developing Countries: A
Neumann (consultant), Education Department.      Simultaneous Model. David Wheeler (consul-
n  1tant), Development Policy Staff. July 1980.
No. 399. Primary Schooling and Economic          No. 408. Economic Growth and Human Re-
Development: A Review of the Evidence. 1
Deveopmnt:A Reiewof he Eidece.sources. Norman Hicks, Policy Planning and
Christopher Colclough (consultant), Education    Pograeview Dearmet July        1980.
Department. June 1980.



No. 409. TI'he Changing Nature of Export          Number 121. The Economic Theory of the
Credit Finance and Its Implications for Devel-    Household and Impact Measurement of Nu-
oping Countries. Albert C. Cizauskas, Economic    trition and Related Health Programs. Dov
Analysis and Projections Department. July 1980.   Chernichovsky. From Evaluating the Impact of
Nutrition and Health Programs, Robert E. Klein
No. 410. Trade in Non-Factor Services: Past       et al., eds. (New York: Plenum Publishing Cor-
Trends and Current Issues. Andre Sapir (con-      poration, 1979), pp. 227-67.
sultant) and Ernst Lutz, Economic Analysis and
Projections Department. August 1980.              Number 122. Pesticides, Information, and Pest
Management under Uncertainty. Gershon
No. 411. Population Policy and Family Plan-       Feder. From American Journal of Agricultural
ning Programs: Trends in Policy and Admin-        Economics (February 1979), pp. 97-103; copy-
istration. Kandiah Kanagaratnam, Population,      righted by the American Agricultural Economics
Health, and Nutrition Department, and Cather-     Association.
ine S. Pierce, Development Economics Depart-
ment. August 1980.                                Number 123. The Behavior of a Dual Econ-
omy under Different "Closing Rules." Clive
No. 412. Health Problems and Policies in the      Bell. From Journal of Development Economics,
Developing Countries. Fredrick Golladay,          vol. 6 (1979), pp. 47-72.
Transportation, Water, and Telecommunications
Department, and Bernhard Liese, Projects Advis-   Number 124. Economic Rent: Incidence in Se-
ory Staff. August 1980.                           lected Metals and Minerals. Helen Hughes and
Shamsher Singh. From Resources Policy (June
No. 413. Eastern and Southern Africa: Past        1978), pp. 135-45.
Trends and Future Prospects. Ravi Gulhati,
Eastern Africa Regional Office. August 1980.      Number 125. Accounting and Fixed Price Mul-
tipliers in a Social Accounting Matrix Frame-
lV4orld Bank Staff Working Paper Nos. 400-412    work. Graham Pyatt and Jeffery I. Round.
are background studies for World Development      From The Economic Journal, vol. 89 (December
Report, 1980.                                     1979), pp. 850-73.
Number 126. Hierarchial Location Analysis
for Integrated Area Planning in Rural India.
WORLD BANK REPRINT SERIES                         Shyamadas Banerji and H. Benjamin Fisher.
From the Thirteenth European Congress of the
The following recent articles, arising from      Regional Science Association, vol. 33 (1974),
research undertaken at or for the World Bank,     pp. 177-94.
have been reprinted with permission:
Number 127. Target Group Oriented Food
Number 119. The Oil Price Revolution of           Programs: Cost Effectiveness Comparisons.
1973-1974. Salah El Serafy. From The Journal      Marcelo Selowsky. From American Journal of
of Energy and Development, vol. 4, no. 2 (Spring  Agricultural Economics, vol. 61, no. 5 (December
1979), pp. 273-90, by permission of the Interna-  1979), pp. 988-94.
tional Research Center for Energy and Economic
Development (ICEED), University of Colorado.     Number 128. Costs Incurred by Residential
Electricity Consumers Due to Power Failures.
Number 120. Engineering Economics of the          Mohan Munasinghe. From Journal of Consumer
Maintenance of Earth and Gravel Roads. Asif       Research, vol. 6 (March 1980), pp. 361-69.
Faiz and Edgardo Staffini. From Transportation
Research Record, no. 702 (proceedings of a con-
ference on low-volume roads conducted by the
Transportation Research Board, August 20-23,
1979) (Washington, D.C.: National Academy of
Sciences, 1979), pp. 260-68.



Number 129. Demographic Factors and the            Number 137. Consumer Search and Alterna-
Distribution of Income: Some Issues. Pravin        tive Market Equilibria. Avishay Braverman.
Visaria. From Economic and Demographic             From Review of Economic Studies, vol. 47
Change: Issues for the 1980s. Proceedings of the   (1980), copyrighted by the Society for Economic
Conference (Liege: IUSSP, 1979), vol. 1,          Analysis Limited.
pp. 289-320.                                      Number 138. The Evaluation of the Benefits
Number 130. Market Access for Semi-manu-          of Basic Needs Policies. Pasquale Scandizzo
factures from Developing Countries. Deepak        and Odin K. Knudsen. From American Journal
Lal. From Commercial Policy Issues, no. 5         of Agricultural Economics, vol. 62, no. I (Febru-
(Leiden: Sijthoff, for the Graduate Institute of  ary 1980), pp. 46-57. Copyrighted by the Ameri-
International Studies, Geneva, and the Trade      can Agricultural Economics Association.
Policy Research Centre, London, 1979).
Number 131. Shadow Pricing and Wage and
Employment Issues in National Economic
Planning. Deepak Lal. From The Bangladesh          WORLD BANK COUNTRY STUDIES
Development Studies, vol. 6, no. 3 (Monsoon
1978), pp. 233-56.                                 World Bank country studies are prepared mainly
for the Bank's own use, with distribution restricted
Nrumbr T132 Ac        Nt forw  eco                to member governments and internatlonal organi-
Growth: The Case Of Norway. Bela Balassa.         zations that deal with development problems.
From Oxford Economic Papers, vol. 31, no. 3        Where the issues studied have attracted a wide
(November 1979), pp. 415-36.                      interest, where it appears that the Bank's study
Number 133. Incentive Policies in Brazil. Bela    could contribute substantially to knowledge and
Balassa. From World Development, vol. 7 (1979),   understanding of these issues, and where the
pp. 1023-42. Also available in Portuguese, Num-   authorities of the country concerned art' agreeable,
ber 133a: Politicas de Incentivo no Brasil. From  such reports are made available to a wider audi-
Pesquisa e Planejamento Econ6mico, vol. 9, no. 3  ence. Potential readers are advised that these are
(dez. 1979), pp. 739-82.                          working documents, not prepared with a view to
broad distribution.
Number 134. The Tokyo Round and the De-           Employment and Income Distribution in In-
veloping Countries. Bela Balassa. From Journal    donesia. Mark Leiserson, Chief of Mission, et
of World Trade Law, vol. 14, no. 2 (March!        al., East Asia and Pacific Regional Office. July
April 1980), pp. 93-118.                           1980.
Number 135. Commercial Bank Lending to
Developing Countries: Supply Constraints.         Seychelles: Economic Memorandum. Robert
Chandra S. Hardy. From World Development,         Maubouche and Naimeh Hadjitarkhani. Eastern
vol. 7 (1979), pp. 189-97.                        Africa Regional Office. July 1980.
Number 136. A "Stager" Approach to Com-
parative Advantage. Bela Balassa. From Eco-
nomic Growth and Resources, edited by Irma
Adelman. Volume 4 of the Proceedings of the
Fifth World Congress of the International Eco-
nomic Association, Tokyo, Japan, 1977,
pp. 127-156.






0             9            0



I



0



World Bank
Headquarters
1818 H Street, N.W.
Washington, D.C. 20433, U.S.A.
Telephone: (202) 477-1234
Cable Address: INTBAFRAD
WASHINGTONDC
European Office:
66, avenue d'16na
75116 Paris, France
Tokyo Office:
Kokusai Building
1-1. Marunouchi 3-chome
Chiyoda-ku, Tokyo 100, Japan