.-  A WorldBankPublication
1 .estructuri,ng
Economies i:n
.   q      1  rt~~~~~~~c'LE CCPY
POl1yRIr  ana-_ 
the World Bank
4'~~~~~~
.   Vinod Thomas
i,  5Mansoor Dailami ^      -Xi
2 ~Jaime de Melo es\<t 






RESTRUCTURING
ECONOMIES
IN DISTRESS






RESTRUCTURING
ECONOMIES
IN DISTRESS
Policy Reform and the World Bank
edited by
Vinod Thomas
Ajay Chhibber
Mansoor Dailami
Jaime de Melo
PUBLISHED FOR THE WORLD BANK
Oxford University Press



Oxford University Press
OXFORD NEW YORK TORONTO DELHI
BOMBAY CALCUTTA MADRAS KARACHI
PETALING JAYA SINGAPORE HONG KONG
TOKYO NAIROBI DAR ES SALAAM
CAPE TOWN MELBOURNE AUCKLAND
and associated companies in
BERLIN IBADAN
C 1991 The International Bank for
Reconstruction and Development l THE WORLD BANK
1818 H Street, N.W., Washington, D.C. 20433, U.S.A.
All rights reserved. No part of this publication
may be reproduced, stored in a retrieval system,
or transmitted in any form or by any means,
electronic, mechanical, photocopying, recording,
or otherwise, without the prior permission
of Oxford University Press.
Manufactured in the United States of America
First printing June 1991
The findings, interpretations, and conclusions expressed in
this publication are those of the authors and do not necessarily
represent the views and policies of the World Bank or
its Board of Executive Directors or the countries they represent.
The World Bank does not guarantee the accuracy of the data
included in this publication and accepts no responsibility
whatsoever for any consequences of their use.
Library of Congress Cataloging-in-Publication Data
Restructuring economies in distress: policy reform and the World Bank
I edited by Vinod Thomas .. . [et al.].
p. cm.
Includes bibliographical references and index.
ISBN 0-19-520870-6
1. World Bank-Developing countries. 2. Balance of payments-
Developing countries. 3. Economic stabilization-Developing
countries. 4. Developing countries-Economic policy. 1. Thomas,
Vinod, 1949- .
HG3881.5.W57R47 1991
339.5'09172'4-dc20                                 90-25999
CIP



Foreword
THE WORLD BANK began lending for structural adjustment in 1980. The
balance of payments problems facing developing countries at that time
resulted from high oil prices and rising nominal interest rates, as well as
inadequate domestic policies. These problems intensified as real interest
rates rose, a global recession developed, and the debt crisis was recognized
as such.
Initially, adjustment lending for a given country was expected to last
three to five years, and its balance of payments was expected to become
sustainable in five to seven years. As the 1980s progressed, however, the
terms of trade for many developing countries deteriorated further, real
interest rates remained high, and the debt crisis persisted. In addition,
the sources of domestic inefficiencies proved to be intractable. Instead
of disappearing, adjustment lending intensified-both in depth, for a
given country, and in breadth, in the range of member countries receiving
such loans. Typically, the World Bank started adjustment lending with
structural adjustment loans and then moved on to a series of sectoral
adjustment loans to effect policy changes in particular sectors.
Almost a decade after the first structural adjustment loans, the World
Bank evaluated its experience with adjustment lending in a report that
was discussed by the Bank's Executive Directors in September 1988. This
volume, a product of the extensive work done for that report, provides
a detailed and frank discussion of adjustment lending programs, how
they have worked, and how they can be improved.
As external uncertainties and internal problems continue for member
countries of the World Bank, it is hoped that the objective of adjustment
efforts will shift from the short-term issues of crisis management and
stabilization that have dominated the past decade, to the more funda-
mental issues of long-term growth, sustainable development, and the
alleviation of poverty.
Barber B. Conable
President, The World Bank
v






Contents
Foreword, Barber B. Conable    v
Contributors   xi
Preface, John A. Holsen    xiii
1 Evolution and Lessons of Adjustment Lending,
Ernest Stern    1
Part I  Issues of Policy Reform      9
2 Introduction to Part I   11
3 Public Finance, Ajay Chhibber and
Javad Khalilzadeh-Shirazi   20
Comments, Peter S. Heller   44
4 Trade Policy Reform, Vinod Thomas   47
Comments, John Williamson   72
5 Financial Sector Reform, Alan Gelb and
Patrick Honohan    76
Comments, MaxwellJ. Fry   100
6 Reform of Public Enterprises, John R. Nellis   108
Comments, Leroy P. Jones   127
7 Agricultural Policy, Odin Knudsen andJohn Nash   131
Comments, Nurul Islam    152
8 Industrial Performance: Some Stylized Facts,
James R. Tybout   159
Comments, Mieko Nishimizu   187
Part II  Evaluation of Adjustment Programs         191
9 Introduction to Part 11   193
10 The Design and Implementation of Conditionality,
William A. McCleary   197
Comments, Elliot Berg   215
vii



Viii CONTENTS
11 Macro Performance under Adjustment Lending, Riccardo Faini,
Jaime de Melo, Abdel Senhadji-Semlali, and
Julie Stanton    222
Comments, Peter J. Montiel   243
12 The Social Costs of Adjustment, Elaine Zuckerman    247
Comments, Dennis de Tray   271
Part III Country Experience with Adjustment Lending
in the 1980s      277
13  Introduction to Part III   279
14 C6te d'lvoire: The Failure of Structural Adjustment,
Christophe Chamley    287
Comments, Shantayanan Devarajan   305
15 Ghana: Ending Chaos, Chad Leechor   309
Comments, Lyn Squire   328
16 Zambia: Reform and Reversal, Mobsen A. Fardi   332
Comments, Ben King   352
Comments on Africa, Edward V. K. Jaycox 358
17 Indonesia: Stabilization and Structural Change,
Sadiq Abmed    361
Comments, William H. Branson   382
18 Korea: Successful Adjustment, Mansoor Dailami   386
Comments, Larry E. Westphal   405
Comments on Asia, Attila Karaosmanoglu   412
19 Pakistan: Structural Adjustment and Economic Growth,
William A. McCleary    414
Comments, Mobsin S. Khan   435
20 Turkey: Structural Transformation and the Threat
to Sustainability, Faezeh Foroutan    440
Comments, Dani Rodrik   462
Comments on Pakistan and Turkey, Hans-Eberhard Kopp   466
21  Chile: Economic Crisis and Recovery, Cristian Moran    468
Comments, Sebastian Edwards   489
22 Mexico: Adjustment and Stabilization, John Nash    494
Comments, Vittorio Corbo   515
Comments on Latin America, S. Shahid Husain   516



Contents ix
Part IV   Assessment    521
23 Lessons of Experience and the Future of Adjustment
Lending    523
Joe Abbey
Stanley Fischer
Manuel Guitian
Gerald Helleiner
Pedro S. Malan
Moeen Qureshi
24 Summary and Conclusions   544
Index    549






Contributors
AFFILIATIONS are those at the time of the conference.
Joe Abbey, Ghanaian High Commissioner in the United Kingdom
Sadiq Ahmed, World Bank
Elliot Berg, Elliot Berg Associates, Virginia
William H. Branson, Princeton University
Christophe Chamley, Boston University
Ajay Chhibber, World Bank
Barber B. Conable, President, World Bank
Vittorio Corbo, World Bank
Mansoor Dailami, World Bank
Jaime de Melo, World Bank
Dennis de Tray, World Bank
Shantayanan Devarajan, Harvard University
Vinod Dubey, World Bank
Sebastian Edwards, University of California, Los Angeles
Riccardo Faini, Johns Hopkins University, Bologna Center, Italy
Mohsen A. Fardi, World Bank
Stanley Fischer, World Bank
Faezeh Foroutan, World Bank
Maxwell J. Fry, University of California, Irvine
Alan Gelb, World Bank
Manuel Guitian, International Monetary Fund
Gerald Helleiner, University of Toronto
Peter S. Heller, International Monetary Fund
John A. Holsen, World Bank
Patrick Honohan, World Bank
W. David Hopper, World Bank
S. Shahid Husain, World Bank
Nurul Islam, International Food Policy Research Institute
Edward V. K. Jaycox, World Bank
Leroy P. Jones, Boston University
Attila Karaosmanoglu, World Bank
Javad Khalilzadeh-Shirazi, World Bank
Mohsin S. Khan, International Monetary Fund
Ben King, formerly World Bank
Odin Knudsen, World Bank
Hans-Eberhard K6pp, World Bank
Chad Leechor, World Bank
Pedro S. Malan, Alternate Executive Director, World Bank
William A. McCleary, World Bank
Peter J. Montiel, International Monetary Fund
Cristian Moran, World Bank
John Nash, World Bank
xi



Xii CONTRIBUTORS
John R. Nellis, World Bank
Mieko Nishimizu, World Bank
Moeen Qureshi, World Bank
Visvanathan Rajagopalan, World Bank
Dani Rodrik, Harvard University
Abdel Senhadji-Semlali, World Bank
Murray A. Sherwin, Executive Director, World Bank
Lyn Squire, World Bank
Julie Stanton, World Bank
Ernest Stern, World Bank
Vinod Thomas, World Bank
James R. Tybout, Georgetown University
Larry E. Westphal, Swarthmore College
John Williamson, Institute for International Economics
Elaine Zuckerman, World Bank



Preface
THE GROWTH in the pace and scope of lending to developing countries
for economic adjustment has raised many concerns and provoked much
debate. The chief question is whether this instrument, designed essentially
to provide support for the balance of payments, can effectively serve the
broader objectives of policy and institutional reforms. Another important
issue concerns conditionality and whether it can be designed to integrate
short-term macroeconomic stabilization with longer-term reforms.
For the World Bank's report on adjustment lending, background pa-
pers were prepared on policy issues and country experiences, and a sym-
posium was held at Bank headquarters on April 4-5, 1989, to discuss
the background work. In the spirit of open debate, the symposium in-
cluded commentary and reviews by some of the World Bank's practi-
tioners and by outside critics of adjustment lending. This volume contains
the statements, papers, and comments presented at the symposium, as
well as two other papers commissioned after the conference-one to
analyze loan conditionality (chapter 10) and the other to round out the
case studies for the Asian region (chapter 17).
This book represents an assessment of nearly a decade of experience
with lending to support adjustment programs. Although the effectiveness
of structural adjustment and adjustment lending has varied considerably
across countries and policy areas, the overall judgment is one of qualified
optimism. Definitive conclusions are not yet possible, however, and ef-
forts to evaluate the success of adjustment programs and adjustment
lending must continue.
Opening this volume is an introductory statement to the symposium
by Ernest Stern, one of the architects of adjustment lending. The re-
mainder of the book is divided into four parts. The first covers major
policy and sectoral issues related to structural adjustment, or what can
be considered the content of adjustment programs. Part II is concerned
with the evaluation, conditionality, and social consequences of adjust-
ment operations. Case studies of the direct experience of nine countries
with structural adjustment and adjustment lending are presented in Part
III. Part IV contains the comments of various participants in the 1989
symposium and ties together many of the themes that are woven through-
out the preceding text.
This book has benefited from contributions by Bela Balassa, Suman
Bery, Fernando Clavijo, Simon Commander, Nadav Halevi, Arnold Har-
berger, Ishrat Husain, Donald Keesing, Enrique Lerdau, Robert Lieben-
thal, Dennis Mahar, Francis Ng, Peter Nicholas, Gerardo Sicat, Jose
xiii



XiV PREFACE
Sokol, and William Steel, editing by Meta de Coquereaumont, and as-
sistance from Karla Cabana and Rebecca Sugui. Valuable comments and
suggestions were received from many people inside and outside the Bank
as well as from participants in the symposium.
John A. Holsen
Director, Country Economics Department
The World Bank



1 Evolution and Lessons of
Adjustment Lending
Ernest Stern
WHEN THE WORLD BANK introduced the idea of structural adjustment
lending in 1979, I think it is fair to say, neither Robert McNamara, the
Bank president, nor I appreciated the full scope of the undertaking.
McNamara made his first reference to structural adjustment in a speech
to the U.N. Conference on Trade and Development (uNcTAD), and a brief
paragraph in his 1979 address to the annual meeting of the World Bank
proposed lending to help developing countries avoid balance of payments
crises.
Coming in the aftermath of the second oil shock, the proposal reflected
a great sense of frustration in the World Bank. Despite substantial lend-
ing, we had no real involvement with the major macroeconomic issues
that our members faced. Clearly, our project lending was not an adequate
tool. Our economic analysis, although it may have been as voluminous
then as now, generally was not directly linked to operational decisions.
We certainly did not have the kind of policy dialogue that has since
emerged.
Starting Up
Initially our thinking focused on ways to help countries develop greater
export capacity, given their need to deal with the tremendous drain of
resources generated by the oil price increase. When we started, however,
we were quite naive about just how profoundly distorted the development
strategies of many of the developing countries were. (Even now, this issue
is worth exploring at some depth.) The problem had been neglected not
only in our own analysis but also in much of the economic literature of
the time.
The oil shocks and subsequent occurrences in the 1980s revealed that
many of the development strategies being pursued simply did not promise
sustainable economic growth. That reality, although not yet fully com-
prehended, is still very much with us. It makes adjustment a subject of
continuing great importance and great urgency.
1



2 EVOLUTION AND LESSONS OF ADJUSTMENT LENDING
Whether we will be able soon to put the debt crisis-one of the symp-
toms of these distortions-behind us, in part or in whole, really has
nothing to do with the continuing need for developing countries to re-
think their development strategies. They must begin to move from highly
distorted price incentives and investment frameworks to something more
stable, more oriented to the market system of prices, and more open and
less protected.
In the early 1980s, we foresaw neither the full dimensions of the chal-
lenge nor the transformation that structural adjustment lending would
impose on the World Bank. A look at the kinds of activities the Bank
carried out in 1980, the demands on its staff, and its relationships with
developing countries then reveals not only dramatic differences from the
present but also that much of the change stems from the World Bank's
operationally significant involvement in macroeconomic and sectoral ad-
justment issues. In retrospect, it is clear that we did not anticipate just
how popular structural adjustment lending was going to become.
When we set out, it was a highly controversial subject. Indeed, our
first proposals for adjustment lending met a great deal of opposition from
the Board of Executive Directors and more generally, for a variety of
reasons. First, many thought the problems we were seeking to address
were short-term ones that did not require the attention of a long-term
development institution. Second, members of the World Bank staff and
of the Board felt strongly that we were essentially a project agency. They
held that we should stick to what we knew and not venture into macro-
economic issues where we had no special competence.
Such involvement, it was said, would create all kinds of problems,
among them, frictions between the World Bank and the International
Monetary Fund (IMF). Some of these issues are still being argued, but
perhaps not as vigorously as they were almost a decade ago. In the in-
tervening years, in any event, we have learned a set of lessons that remain
valid.
Lessons of Experience
The first lesson-a very important one that we learned fairly early-is
that structural adjustment has to be seen in a countrywide context. Ini-
tially, and perhaps to some extent still today, we tended not to carry over
into structural adjustment lending a truism about project lending: a
soundly designed project in a poor economic environment will not yield
expected results. In addition, we have at times been perhaps too com-
placent about the issues of comprehensiveness and sequencing, not rec-
ognizing that intermittent, partial, and gradual steps are insufficient in
themselves; they must fit into a well-articulated, well-understood overall
plan that has the full commitment of the government.



Evolution and Lessons of Adjustment Lending 3
The second lesson concerns the great degree of economic dependence
that the policies of the 1960s and 1970s had bred. The ample supply of
capital, from both growing aid programs and the great liquidity created
by oil price increases, led to a systemic failure. Developing countries as
well as aid and export credit agencies-not to mention the World Bank-
failed to insist on any reasonable standards of efficiency for investment
capital and the management of economies.
The third lesson-and it became clear only gradually-was that ad-
justment postponed was pain compounded. This truth has been very hard
to learn. It is still not fully understood. After all, most development econ-
omists-and certainly most of those in the World Bank-grew up be-
lieving that development could proceed in a fairly linear fashion and fairly
painlessly. When we did, occasionally, deal with such problems as in-
equitable income distribution and the need for some drastic remedies,
we still tended to perceive the process as a very gradual one and the pace
of change as not very important.
We have learned since then that this approach is quite wrong. When
distortions are serious-for example, if the prices of energy or food are
only a fraction of world market prices-the longer adjustment is delayed,
the more difficult it becomes politically. The more difficult politically,
in turn, the more partial any adjustment is likely to be and the more
likely it is to fail.
The fourth lesson, which emerged as we became more deeply involved
with more and more countries in structural adjustment lending, was the
very limited institutional capacity to manage economies flexibly. This is
not to say that bureaucracies everywhere are inefficient or incapable,
only that the strong vested interests they harbor did not change easily
from a controlled administrative system or adapt readily to a more open
system.
In general, it is fair to say, we grossly underestimated the institutional
elements of adjustment at both the macro and the sectoral levels. Time
and again the best of policy intentions, the best of policy letters solemnly
agreed to and signed by the finance minister and the Bank, broke down.
This often happened because of a lack of adequately trained people in
the right places, the difficulty of changing bureaucratic structures, and
the difficulty of laying people off as functions were reduced.
The fifth lesson, learned gradually in the course of adjustment lending,
is the vital importance of macroeconomic flexibility. Adjustment lending,
whether structural or sectoral, is not going to be very productive unless
the country has a stable macroeconomic environment.
The arguments against this position can be very seductive politically.
It is often suggested that one ought to start slowly and that the macro-
economic issues will take care of themselves in time. I myself am convinced
that these are siren songs to which we listen at our peril. It is important



4 EVOLUTION AND LESSONS OF ADJUSTMENT LENDING
as part of any liberalization effort, any adjustment effort, that the mac-
roeconomic framework be sustainable. If it is not, we have learned time
and again, the imbalances created tend to overwhelm whatever structural
changes can be made at the sectoral level.
Finally, it is my personal conclusion-one which is certainly not uni-
versally shared-that structural adjustment lending ought to be anchored
in balance of payments problems.
It is important to distinguish between the need for policy reform and
the need for structural adjustment lending by the World Bank. They are
not identical. People often ask, "Why get involved at all in structural
adjustment lending? Why bribe governments to do what they ought to
be doing anyway?" The answer obviously is that one cannot bribe gov-
ernments to undertake basic adjustments. Any government that would
agree to structural changes just because there is a loan at the end of the
process does not understand the scope of the changes, is not committed
to the endeavor, would not resist the political opposition reforms might
generate, and will not ensure that the adjustment policies are sustainable.
In my view, the correct answer has to be that we provide quick-dis-
bursing loans because the actions being undertaken by the government
have some balance of payments impact, some additional costs that we
can help to defray. To me, that justification has been and continues to
be the basic anchor for structural adjustment lending.
This standard does not mean that there is no need for policy reform
in sectors in which the costs are not in foreign exchange. Indeed, countries
face major issues of reform in education and the other social sectors, in
transport, infrastructure, the environment, and urban development. Al-
though the World Bank should be an active participant in those policy
reforms, and a source of financial support, I believe quick-disbursing
assistance is not appropriate for these types of reforms.
Underestimating the Political Factor
Taken together, these lessons hold a moral for economists. It is that we
at the World Bank-and everyone else, I believe-underestimated the
political difficulty of protracted adjustment. Economists here and else-
where often tend to believe that we need only do our analysis, reach our
conclusions, and write a report; the rest will follow. We do not have
much experience with the political processes of change. We fail to give
full weight in our own thinking to the fact that structural adjustment
means a major redistribution of economic power and hence of political
power in many of the countries undergoing this process.
Such a shift is not neutral. It does not happen easily or by itself. The
politics of change is one of the reasons adjustment has taken a great deal



Evolution and Lessons of Adjustment Lending S
more time than expected in some countries and one of the reasons some
adjustment efforts have not been sustained.
I do not believe that the World Bank can do very much about this
process except to understand the problem. Hiring political scientists to
tell us what will or will not work in a country is not the solution. Those
judgments are properly the country's responsibility.
Nevertheless, in working out adjustment loans and time frames for
policy change, we need to give due attention to the political difficulty of
adjustment. We should not say that because the politics are difficult the
time has to be very protracted. Rather, we should take the difficulty as
part of the judgment to be made about the adequacy of both the gov-
ernment's commitment and its capacity to bring about the specific change
and achieve its objectives in the time agreed. We need to recognize that
either the commitment or the pace of its implementation may be less than
required.
Some Unresolved Issues
There remain a number of very important issues still to be resolved. I
will conclude by raising four of them.
First, we have not yet been able to define very well how to handle
policy reform in countries without significant balance of payments prob-
lems. Quick-disbursing lending seems to appeal to everyone. But I believe
that such loans must be justified by their contribution to the balance of
payments costs that adjustments incur and that the country lacks the
resources to meet-and this justification becomes weak if there is no
immediate balance of payments problem.
This issue is only a subset of a larger question about the transition
from adjustment lending to exclusively project-based lending in a coun-
try: how does the World Bank build on the present policy dialogue-
clearly most effective, most intimate, and most operationally relevant in
the adjusting countries-once the adjustment process is finished? Once
we no longer see a need for structural adjustment loans, what are the
means by which the World Bank will maintain a continuing discussion
with governments on their macroeconomic or sectoral management in
an operationally relevant way?
The second set of issues relates to the nature of this transition for the
countries now receiving adjustment loans. Transitions do take place.
Overwhelmed today by the great number of countries in deep difficulty,
we tend to forget that some countries have come out of that trouble.
Transitions have started or are well advanced in Chile and Turkey, for
example.
Yet it is easy to see the pressure building to perpetuate today's practices
indefinitely. Stopping adjustment lending, goes the argument, would be



6 EVOLUTION AND LESSONS OF ADJUSTMENT LENDING
wrong because it would reduce net disbursements to a country sharply.
The management of the World Bank's disbursement pipeline is absolutely
central, no matter how the transition will take place. During the period
of adjustment lending; project pipelines tend to dry up because investment
declines. In these circumstances, it is hard to get the governments' at-
tention, to get them to make the expenditures necessary to develop a
project pipeline. The result is that when adjustment lending, or quick-
disbursing lending, comes to a natural end, when most of the incentive
framework at the sector or macroeconomic level has been broadly put
to right, the country finds itself with a very sharp drop in disbursements
from the World Bank.
This drop, depending on the scale, obviously can create a major fi-
nancing problem. We and the governments involved need to pay attention
early on to managing the financial side of this transition. Otherwise, we
will find that the end of adjustment lending brings very serious financial
problems requiring a continuation of quick-disbursing lending, but for
very different and essentially inappropriate reasons.
A related issue is how to handle policy reform in sectors that have no
direct impact on the balance of payments. How can the World Bank and
other development agencies be of greater assistance to governments in
supporting the reforms necessary in the social sectors, urban areas, en-
vironmental management, and other spheres where quick-disbursing as-
sistance may not be the proper tool because the balance of payments
impact is small or because the financial benefits of the loan do not go
to the sector concerned? What kind of sectoral lending can we envisage
that will be useful to developing countries, will be the basis for effective
support for sectoral policy change, and yet will not simply finance general
imports?
A final question also relates to the transition. How do we assess the
appropriate balance between project lending and structural adjustment
lending?
One of the reasons the World Bank has been very effective-and be-
came effective very quickly-in structural adjustment lending at both the
macroeconomic and the sectoral levels is its very extensive database, the
profound knowledge accumulated from its project work and sectoral
analysis in many countries over many years. Structural adjustment lend-
ing, however, has a tendency to dry up those sources of information.
Policy reform by itself does not produce anything; it only makes more
efficient production and innovation possible.
The supply response is obviously crucial. In country after country in-
vestment levels have been dropping and in some cases have reached main-
tenance levels only. Clearly project lending ought to be seen not simply
as a residual activity nor only as a way to understand the economy, but
rather as a major tool to help begin the vitally needed increase in in-



Evolution and Lessons of Adjustment Lending 7
vestments. For countries that have made progress in reorienting their
policies, project lending should become increasingly important both for
World Bank financing and as a stimulus to resumed commercial financ-
ing.
We are grappling with a fascinating subject. The analyses and findings
assembled for this symposium make a valuable contribution to our un-
derstanding of the issues involved in adjustment lending and should stim-
ulate continuing, fruitful discussion.






PART I
Issues of Policy Reform






2 Introduction to Part I
"ADJusTMErN" has acquired a wide variety of meanings and interpre-
tations. For some people, it refers to stabilization of the economy through
measures to reduce or remove external and internal imbalances. For oth-
ers, it refers to measures to achieve structural changes that will contribute
to economic recovery and growth. At one extreme are those who suggest
that adjustment and sound policies for economic development are one
and the same, since both involve a wide range of desirable changes in
the economy. At the other extreme are the critics who associate adjust-
ment with policies they consider undesirable for development. It is useful,
therefore, to clarify some of these concepts and define how the term is
used in this volume.
Although long-term structural changes are an inevitable part or, better
still, an underlying determinant of economic development,' this is not
quite the sense in which adjustment is discussed here. A more time-bound
phenomenon is meant, which is best illustrated by the experience of the
1980s, when developing countries were hit with three major shocks and
the curtailment of external funding that came with the debt crisis in 1982.
The three shocks were the global recession of 1979-82, the increase in
real interest rates to positive levels for countries that had become ac-
customed to cheap financing, and the decline in the terms of trade for
producers of oil and primary commodities. These shocks led to large and
growing internal and external imbalances just when external financing
declined sharply because of the debt crisis.
Two types of policy response, both labeled "adjustment," were called
for. The first was stabilization, or managed reductions in expenditures
to bring about an orderly adjustment of domestic demand to the reduced
level of external resources available to the country. The second was struc-
tural adjustment, or changes in relative prices and institutions designed
to make the economy more efficient, more flexible, and better able to
use resources and so to engineer sustainable long-term growth. It was
envisioned that effective structural adjustment measures would reduce
the necessary extent of stabilization.2
Thus, in most countries adjustment involved, first, a rapid reduction
of the imbalances in the current account of the balance of payments and
11



12 ISSUES OF POLICY REFORM
the fiscal deficit-that is, stabilization. In this sense, adjustment also
aimed to reduce the rate of inflation to a sustainable level. But this was
not enough to improve growth because the reduction of imbalances at
current output levels implied a cut in living standards or a cut in in-
vestment. These cuts had adverse implications for future growth, best
illustrated with the familiar national income identity.
(2-1)           Y + NFI = C + I+ X - M  + NFI
where Y denotes gross domestic product (GDP), C consumption, I in-
vestment, X exports, M imports, and NFI net factor income. A reduction
in the external current account deficit (or an increase in the surplus X
- M + NFI) therefore involved a reduction in consumption C or in-
vestment I for the same level of gross national product (GNP). The sta-
bilization phase therefore involved a cut in the standard of living and
represented the harsh aspect of adjustment. One way to ameliorate this
was to bring about a change in the pattern of expenditure, that is, to
rely more on expenditure switching (rather than simply on expenditure
reduction). One channel was to switch expenditures and production in-
centives from nontraded goods to import substitutes and exports. The
level of imports (M) was a function of both income (Y) and the real
exchange rate (r), as shown in equation 2-2:
(2-2)                     M  = M (r,Y).
Import demand would be reduced by a real depreciation of the exchange
rate, so that the entire burden of stabilization would not fall on ex-
penditure reduction. Furthermore, the real depreciation of the exchange
rate would also induce a higher supply of exports.
The way to avoid lower standards of living was also to focus on the
growth aspects of adjustment, that is, on structural adjustment. Structural
adjustment includes reforms of policies and institutions-microeconomic
(such as taxes), macroeconomic (such as fiscal imbalance), and institu-
tional (such as public sector inefficiencies). These changes can improve
resource allocation, increase economic efficiency, expand growth poten-
tial, and increase resilience in response to future shocks. The economic
distortions are interconnected. For example, an overvalued exchange rate
is evidence of macroeconomic distortions, yet it also contributes to mi-
croeconomic distortions because producers of traded goods face a relative
price disincentive. Similarly, public enterprise pricing often leads to mi-
croeconomic distortions when prices do not reflect costs, to institutional
distortions in that managers have inadequate incentives to be efficient,
and to macroeconomic distortions as deficits become large.
In a world where resources had become severely constrained, the re-
forms of the 1980s tried to improve efficiency through the reallocation
of resources. Policies in this direction constituted the structural change



Introduction to Part 1 13
aspects of adjustment. In most countries, they involved changes in key
relative prices in the economy, such as the real exchange rate, the real
interest rate, and the internal terms of trade between agriculture and
industry. They also involved, among other things, the restructuring of
government operations and expenditures to fit the reduced resources and
to improve efficiency.
There was also a need for institutional changes, especially for a fun-
damental rethinking of the role of government in society. Past devel-
opment strategies called for a major and growing role for government
in the provision not only of public goods but also of a broad range of
manufactured goods as well as services. There was a need to refocus
government involvement in the economy in view of the poor performance
of state enterprises in many countries. The rethinking extended to the
regulatory framework and particularly to the objectives of regulation and
control in developing countries.
In many countries, the past neglect of the agricultural sector had been
very costly to economies. The important role of agriculture as a provider
of resources, a market for goods, and a supplier of food had been for-
gotten. Restructuring therefore involved the restoration of the terms of
trade in favor of agriculture. Such restructuring need not be anti-industry,
as is often suggested, since industrial growth cannot be predicated on a
squeezing out of agriculture.
In many cases there was also a need to rethink the inward-looking
approach that had been the hallmark of the 1960s and 1970s. Achieving
a reduction of imports during the stabilization phase was a short-run
objective. In the long run the objective was to expand exports in order
to increase imports and access to the technology needed to boost growth.
Again, the real exchange rate was a key variable in this strategy. The
successful export-oriented approach of the East Asian countries was a
clear indication that closed economies were not only bearing high costs
because of poor resource allocation, but-more fundamentally-were
lagging behind in the technological changes taking place in the rest of
the world. The most dramatic example of the failure of the inward-
looking approach was the collapse of the existing political regimes in
Eastern Europe.
These structural changes would obviously take time, since the changes
envisioned were for both policies and institutions. There were short-run
costs to be borne in return for longer-run benefits. If the goal was to
restore growth, it could not be achieved if economies were starved of
external financing. This was the rationale for the World Bank's adjust-
ment lending, which is reviewed in this volume. The purpose was to
provide medium-term financing over a finite period to help countries
undertake needed economic restructuring. The restructuring may involve
adjustment of both types-policies to achieve changes in internal and



14 ISSUES OF POLICY REFORM
external balances and changes in the structure of incentives and insti-
tutions.
Part I of this volume explores the six policy areas that typically are
featured in the design of adjustment programs: fiscal policy, trade, fi-
nance, public sector management, agriculture, and industry (see table 2-
1 for a summary of findings). Each chapter in Part I is devoted to one
of these policy areas and examines the design issues and other concerns
of that area. In addition, these chapters touch on the broader issues of
the overall design of adjustment programs and the interaction between
various policy reforms. They do not provide sweeping conclusions on
the adequacy of policy design, but they indicate areas of success and
failure and suggest improvements in the future.
FISCAL POLICY. Less progress has been made in fiscal reform than in
the external accounts, as is clearly brought out in chapter 3 by Ajay
Chhibber and Javad Khalilzadeh-Shirazi. Some countries reduced their
fiscal deficit primarily by cutting expenditures (particularly investment
expenditures), rather than by increasing revenue, but the policy was not
sustainable. In other countries, debt service payments rose sharply in the
budget-both because interest rates were rising and because devaluations
increased the domestic currency cost of foreign debt repayments. As a
result, a widening gap between the fiscal deficit and the current account
deficit put increasing pressure on the private sector to generate a net
surplus. This led to cuts in private investment and growth, often accom-
panied by higher inflation.
Some success has been achieved in rationalizing public investment, with
the help of the World Bank's investment reviews.3 There has been much
less progress in reforming taxes and reallocating current expenditures.
Given the urgency of reducing fiscal deficits, near-term revenue measures
have dominated initiatives to reform tax policy in most countries in recent
years. Exceptions are Indonesia, Mexico, and Turkey, which have un-
dertaken thorough tax reforms. Reforms to relieve the financial pressure
on the private sector and to generate increased domestic savings and
investments through growth-oriented fiscal policy have generally been
insufficient.
TRADE AND EXCHANGE RATE POLICY. Recipients of adjustment loans
have made considerable progress in achieving exchange rate flexibility
and strengthening the incentives for exports, as Vinod Thomas shows in
chapter 4. The domestic currencies of recipient countries depreciated in
real terms by an average of about 22 percent between 1981-83 and
1985-87 compared with 2 percent for nonrecipients. The adjusting coun-
tries' export volumes grew substantially, although export values in-
creased much less because of declining world prices for these exports.



Introduction to Part 1 15
Table 2-1. Assessment of Reforms in the Major Policy Areas Affected
by Adjustment Programs, 1981-88
Policy area            Extent of reform                    Assessment
Overall        External adjustment rapid, internal Policies have mattered a great
reforms        adjustment often unsustained,    deal, but constraints include
institutional reforms slow       conflicts in design (for example,
between fiscal and trade
policies)
Fiscal         Initial deficit reductions not   Expenditure cuts could not be
sustained                        maintained
Some success in rationalization of  More careful scrutiny of current
public investment (as in         expenditure needed
Pakistan and Turkey)
Tax reform not comprehensive     Short-run revenue concerns have
and with limited success;        dominated
Indonesia, Mexico, and Turkey
are exceptions
Trade          Progress in exchange rate        Strong responses in export
flexibility and export incentives    volume, but nonprice factors
need greater emphasis (as in
East Asia)
Progress in replacement of       Greater liberalization constrained
quantitative restrictions by     by stabilization requirements
tariffs and in tariff reform, but  and internal opposition; export
import liberalization slower     response assists in import
liberalization (as in Korea and
Turkey)
Financial      Reforms initiated in only a few  Reforms often protracted and
cases, but receiving greater     require sound macro framework
attention                        (as in Chile and Turkey)
Public sector   Institutional reforms and       Introduced in almost every
management   divestiture slow, particularly in    adjustment program
Sub-Saharan Africa
Some reduction in enterprise losses Gains more from price increases
than from improvements in
efficiency
Agriculture    Visible improvements in price    Improvements especially strong in
policy                           Sub-Saharan Africa
Reform of parastatals slower     Supply response to price
constrained by institutional
factors
Industry       Focus on trade policy and its    Reforms specific to industrial
impact on industry               sector not given sufficient
attention



16 ISSUES OF POLICY REFORM
Chile, Korea, and Turkey have made spectacular progress in boosting
export volume. Most countries, however, have barely begun to use the
price and nonprice instruments at their disposal to increase exports.
On the import side, several countries have replaced quantitative re-
strictions with tariffs and have reformed their tariff structure. The struc-
ture of protection provided by import barriers, with its implicit bias
against exports, has been more resistant to reform, however. The ten
major trade adjustment borrowers have typically reduced the discrimi-
nation against exports more than have other countries from the same
regions that were also studied. Liberalization efforts raise far fewer ob-
jections if exports are doing well and if the private sector believes that
liberalization is a medium-term program that will be sustained. Economic
analysis and country experience show that countries would benefit from
replacing quantitative restrictions on imports with tariff protection, re-
ducing tariff dispersion, and lowering protection. To improve a pro-
gram's credibility and the private sector's responsiveness, the main fea-
tures and goals of the program should be announced publicly and early.
FINANCIAL SECTOR. In the few countries that have undertaken financial
reforms, such as Argentina, Chile, Indonesia, and Turkey, an appropriate
macroeconomic policy framework has been critical to success. Large
budget deficits make financial reform more difficult because they lead to
high interest rates (as in Brazil and Turkey), a crowding out of the private
sector, and a weakening of the quality of bank portfolios. Efforts to
finance fiscal deficits have frequently created further problems for the
financial sector, as chapter S by Alan Gelb and Patrick Honohan shows.
Budget deficits financed through an "inflation tax" tend to increase the
opportunity cost of holding money, which causes the monetary base to
decline (as happened in Argentina). Some governments have increased
revenue by raising reserve requirements; others have relied on the stat-
utory sale of government bonds at low and often negative interest rates
to mitigate the adverse budgetary impacts. Financial sector repression
has often increased as external resources to finance budget deficits have
declined.
In any event, financial reforms have tended to be protracted; the or-
ganized financial system in many developing countries is technically in-
solvent because of poor management, insufficient supervision of bank
lending, and incorrect signals on lending as a result of financial repres-
sion. Much remains to be done in improving regulation and in financial
sector restructuring. The impact will be limited, however, unless the fi-
nancial repression that has accompanied the large fiscal deficits is re-
duced.
At the 1989 symposium on which this book is based, in the discussion
of the World Bank's role in policy reforms in the areas of trade and



Introduction to Part 1 17
finance, Vinod Dubey found the evidence encouraging. Progress in the
implementation of conditionality in trade and finance was respectable.
The evidence also suggested that adjustment lending is a useful tool in
promoting reforms in these areas. The findings of the papers on trade
and finance provide strong support for continuing to emphasize trade
and financial sector reforms in adjustment lending, while at the same
time addressing the constraints on better policy implementation.
PUBLIC SECTOR MANAGEMENT. Although almost every adjustment pro-
gram has promoted institutional reforms of the public sector, these re-
forms have been almost uniformly slow, as John Nellis shows in chapter
6. There has been some success in slowing or even halting the creation
of new enterprises and the growth of public enterprise employment. Some
public enterprises have reduced their losses, although largely through
price increases permitted by their monopoly position rather than through
improvements in efficiency. Improving efficiency in utility companies, for
example, has proved elusive, and their financial losses remain high. Few
attempted reforms have had much impact on planning, policy analysis,
or debt management in the public sector. After initial delays, several
countries-Jamaica, Niger, Panama, and Togo among them-have made
progress in divesting public enterprises; others, including Ghana, Senegal,
and Turkey, have not. It is too early to evaluate the performance of
enterprises since their privatization.
In the symposium discussion of reform of the public sector, Murray
A. Sherwin drew on the New Zealand experience to underscore the im-
portance of establishing clear commercial objectives for public enter-
prises. If public functions are to be performed by them, these functions
should be clearly identified and their funding made visible. In the case
of privatization it is essential to have a clear framework for monitoring
and accountability so that management can be left with a fair degree of
autonomy in its functions. The risks of privatizing into inefficient markets
should be minimized by policies promoting competition.
AGRICULTURE. Efforts to reduce disincentives to agriculture have been
significant in many countries, especially in Sub-Saharan Africa, as Odin
Knudsen and John Nash show in chapter 7. Although pricing policy is
a crucial element in the recovery of agriculture, supporting measures that
are also needed include both institutional reform and the strengthening
and improvement of government expenditure programs to reduce the
constraints farmers face in responding to better prices. A number of
studies show that public investment, carried out well, can be an effective
complement to price policy in efforts to increase the response of farm
production to changes in incentives.4 In particular, when expenditure
programs are being cut and government revenues from agriculture are



18 ISSUES OF POLICY REFORM
falling as producer prices rise, it is essential to ensure the continued fund-
ing of critical public investments in agriculture.
Other problems affect the agricultural sector in the many developing
countries in which public agencies monopolize commodity trade and
input distribution. Often these agencies are costly and inefficient, dis-
courage private trade, and tax farmers. One way to reduce inefficiencies
is to eliminate government market interventions, and some success has
been achieved in this area. When the immediate alternatives are limited,
however, because of past discouragement of private sector involvement,
it may not be advisable to dismantle public agencies quickly; instead,
adjustment programs have addressed inefficiency and overhead costs in
these agencies or have attempted to minimize the costs to farmers and
private trade. One mechanism is to set pricing rules-for example, by
linking farm prices to international prices-to avoid arbitrariness and
increase the likelihood of more sustained reforms. Some of the World
Bank's sectoral adjustment loans in agriculture, particularly in Mexico,
have introduced such automatic pricing rules. Other measures seek to
reduce overstaffing (Ghana), limit the scope of operations handled by
individual agencies (Morocco and Tanzania), and break up large and
unwieldy agencies into smaller and more specialized ones by, say, sep-
arating credit operations from the distribution of physical inputs (Co-
lombia). Experience shows that such reforms are more difficult to im-
plement but more lasting than those that merely realign prices
periodically.
INDUSTRY. Adjustment issues in the industrial sector vary widely, de-
pending mostly on the level of industrialization. In countries such as
Chile, India, Mexico, and Turkey, complex industrial structures need to
be made more competitive internationally. In Africa, however, the prob-
lems are much more basic, such as the extreme shortages of skilled labor
and modern technology.
Some countries, in trying to achieve rapid industrialization, invested
disproportionately and unwisely in heavy and chemical industries in the
1960s and 1970s; others established public enterprises in many areas or
tried to foster industrialization through distorted incentive systems re-
lying on heavy protection, subsidized credit, and other regulatory mech-
anisms. Frequently the industries thus promoted were viable only with
continuing support. In those cases, adjustments in industry meant re-
ducing losses, protection, and regulation so that a more efficient and
sustainable industrial structure could develop. With the exception of
Asian countries, adjustment operations for industry have generally em-
phasized trade and exchange rate policies. Measures have concentrated
on restructuring, pricing policies, investment incentives, and the redirec-
tion of investment in manufacturing from the public to the private sector.'



Introduction to Part 1 19
In chapter 8, James R. Tybout uses demand-side decomposition of
growth to study the changes in domestic demand and manufacturing
trade flows during the 1980s. The countries with adjustment loans appear
to have done slightly better than their respective regional averages in
terms of growth in manufacturing value added during 1984-86, when
most adjustment programs were under way. The same countries also
performed better than their regional averages during the precrisis period
of 1978-81. Tybout found weak evidence that, when loans featured con-
ditionalities related to trade liberalization or devaluation, countries
tended to exhibit increased import penetration during the adjustment
period. Similarly, an emphasis on conditionalities related to export pro-
motion policies and devaluation seems to have been associated with ex-
port expansion.
At the symposium, W. David Hopper and Visvanathan Rajagopalan
emphasized sectoral and micro-level issues. Improvements are needed in
three complementary areas: policies and regulations, institutions, and
human resources. The review of agricultural performance indicates that
improvements in productivity depend not only on introducing better pol-
icies but also on strengthening institutions and developing human re-
sources. Adjustment lending is useful in supporting policy reform, but
greater attention needs to be devoted to the complementary investments
required for long-term development. These include the recovery and de-
velopment of physical infrastructure, protection of the environment, and
the provision of technology.
Notes
1. See Simon Kuznets, Six Lectures on Economic Growth (Glencoe, Ill.: Free
Press, 1959).
2. It was always recognized that there was an overlap between stabilization
and structural adjustment and that certain measures-exchange rate realign-
ments, for example-played a role under both headings.
3. See Basil Kavalsky, "Reviewing Public Investment Programs," Finance &
Development, March 1986.
4. These studies are surveyed in Ajay Chhibber, "The Aggregate Supply Re-
sponse in Agriculture: A Survey," in Simon Commander, ed., Structural Ad-
justment in Agriculture: Theory and Practice (London: Overseas Development
Institute, 1989).
5. These and other measures are covered in Gerald M. Meier and William F.
Steel, eds., Industrial Adjustment in Sub-Saharan Africa (New York: Oxford
University Press, 1989).



3 Public Finance
Ajay Chhibber
Javad Khalilzadeh-Shirazi
THIS CHAPTER examines fiscal reform in adjustment programs and draws
major lessons for the design of reform and of conditionality in World
Bank adjustment lending. Although the scope of the chapter is broad,
coverage of the issues and countries is selective. The country references
are illustrative rather than exhaustive, since Part III of this book presents
detailed case studies. The main emphasis is on a group of thirteen de-
veloping countries that received three or more World Bank adjustment
loans since the Bank began adjustment lending in 1979. These adjustment
loan-intensive countries are sometimes compared with twelve countries
that received fewer than three adjustment loans and with fifteen countries
that received no adjustment loans from the World Bank.
The first part of the chapter focuses on the macroeconomic implica-
tions of fiscal deficits, the extent to which selected countries have man-
aged to reduce these deficits, the relation between growth and the sus-
tainability of fiscal policy, and the consistency of fiscal policy instruments
with other policy instruments. The next two sections go into issues of
expenditure and revenue. The section on expenditure policy relates in
particular to the level and composition of recurrent outlays and public
investment. This discussion is followed by an examination of tax systems
and experience with their reform. The final section summarizes key con-
clusions and outlines the set of growth-oriented fiscal policies appropriate
for inclusion in adjustment programs and lending.
Macroeconomic Aspects of Fiscal Adjustment
Fiscal crises, in the form of large budget deficits,' have been diagnosed
as a primary feature of external disequilibria and macroeconomic insta-
bility in many developing countries. The massive flow of financial re-
sources to developing countries in the 1970s (largely recycled oil ex-
porters' deposits) helped finance-or, some may say, caused-these
deficits. Foreign savings were used to cover shortfalls in public sector
savings.2 But after 1982 (see figure 3-1) these flows dried up as the
enormity of the debt buildup was revealed, and the underlying unsus-
20



Public Finance 21
Figure 3-1. Current Account and Public Sector Deficit as a Percent-
age of GDP for Thirty-Three Developing Countries, 1980 and
1982-87
10
0~~~~~0
h.)  7-
O   6             Current account                 -
1980   1982  1983  1984  1985  1986  1987
Source: World Bank data.
tainability of the fiscal deficits was exposed in a large number of devel-
oping countries. Moreover, external conditions deteriorated sharply; as
real international interest rates rose dramatically, oil and commodity
prices fell, and growth in industrial countries decelerated, demand for
the exports of developing countries (and others) was reduced. These fac-
tors exacerbated fiscal problems and the deterioration in economic per-
formance in developing countries. Heavy adjustments were required to
redress the budget deficit; government expenditures had to be cut sharply.
Fiscal Crises
The need for adjustment gives rise to several questions: How much of
adjustment should take place in the public sector as opposed to the private
sector? How much should be in the form of restraint on aggregate de-
mand, and how much should come from efforts to increase supply? When
reducing aggregate demand, an obvious need is to reduce the budget
deficit. One approach is to view fiscal policy as primarily an instrument
for reducing expenditure and to rely on other instruments, such as the
exchange rate and pricing, to shift expenditures and resources toward
tradable goods-expenditure switching. But the manner in which the



22 ISSUES OF POLICY REFORM
budget deficit is reduced, and by what amount, and the consistency of
other policies with fiscal policy significantly affect the ability of the econ-
omy to increase output and thereby diminish the need to cut demand.
Consequently, these factors affect as well the sustainability of fiscal
changes. The availability of external finance to smooth the process of
reform is, of course, a key factor.
Has fiscal adjustment been adequate considering the decline in external
financing many developing countries have experienced in recent years?
If not, how have budget deficits been financed, and what are the impli-
cations for the economy of alternative financing packages? Has fiscal
policy "crowded in" or "crowded out" the private sector and thereby
affected the sustainability of reform? A preliminary attempt is made to
answer these questions in this section, which also includes a broad as-
sessment of World Bank and International Monetary Fund (IMF) con-
ditionality related to macroeconomic fiscal issues, and a discussion of
the consistency of fiscal policy with other instruments of adjustment.
There is no doubt that excessive fiscal deficits were the proximate cause
of economic crises in many developing countries, such as Argentina, Ma-
lawi, Mexico, Nigeria, and the Philippines (see table 3-1). Once the crisis
occurred, the credit squeeze affected both the public and private sectors.
In the case of Malawi and the Philippines, recent data show that the
private sector was affected much more than the public sector by the credit
squeeze. Countries such as Indonesia, Korea, and Pakistan that restrained
the growth of credit to the public sector appear to have avoided this
problem. Although both Indonesia and Korea had temporary problems
in the early 1980s because of excessive expansion of credit to the gov-
ernment, both adjusted very quickly by, among other things, drastically
cutting the growth of credit to the public sector.
A large budget deficit can therefore have adverse implications for three
key macroeconomic targets: debt, inflation, and the growth rate of the
economy (for .a fuller discussion, see Buiter 1985; Anand, Chhibber, and
van Wijnbergen, 1990; and World Bank 1988). The arithmetic for this
goes as follows: The current account surplus is equal to net private sav-
ings (that is, private savings minus private investment) minus the public
sector deficit. A fiscal deficit can be financed in one of three ways:3 (1)
external borrowing or grants, (2) domestic borrowing from the banking
system, and (3) domestic borrowing from nonbank sources. Depending
on the method of financing, a large fiscal deficit creates one or more
macroeconomic problems. Large external borrowing creates a debt prob-
lem, financing through monetary creation leads to high inflation, and
heavy domestic borrowing leads either to insufficient credit for the private
sector (when credit is allocated by the government and the financial in-
stitutions it: controls) or to high real interest rates and less private sector
credit, which reduces private investment and future output growth.



Public Finance 23
Table 3-1. Growth of Real Domestic Credit in Selected Countries,
by Sector and Timing of Economic Crisis, 1980-81 to 1985-86
(annual percentage rate of change)
Country'               1980-81   1981-82    1982-83    1983-84    1984-85    1985-86
Argentina (1982)
Government             337.2        23.1       320.4    -135.9         44817        125
Public enterprise       98.2       895.4      -185.0    -174.91    -
Private sector          58.4        27.1       -27.5    -100.8       -300.0        -6.4
Malawi (1981)
Government              74.6        16.6        12.0        -5.9         9.2        23.0
Public enterprise       -7.4       -3.7       -13.0         14.4        20.0       -8.2
Private sector         -12.2         4.9         4.8      -23.9       -20.8          0.1
Mexico (1982)
Government              22.8       113.1      -37.7       -30.1         35.8        n.a.
Public enterprise       33.5        92.1        18.5         9.5       -24.9        n.a.
Private sector          18.6      -36.5       -48.5         23.4       -1.3         n.a.
Nigeria (1983)
Government              63.4        53.9         43.9        6.9        -7.0        n.a.
Public enterprise       16.2      -13.0          98.5     -34.6        -9.2         n.a.
Private sector          17.6         8.4        -1.1       -1.5         -1.2        n.a.
Philippines (1983)
Government              51.3        74.9         14.7     -44.2       -27.5         69.2
Public enterprise        5.3        25.3        34.3      -19.4         10.0      -48.2
Private sector          10.7         5.1         15.3     -50.0        -35.2      -23.8
Indonesia (1981)
Government              52.3         5.6       -18.1        51.4         6.6        n.a.
Public enterprise      -21.4        11.0       -14.0        -7.4         3.2        n.a.
Private sector          21.4        35.6         15.3       23.5        16.3        n.a.
Korea (1981)
Government              69.5        18.1        -1.6         0.6         4.5         8.9
Public enterprise       21.5        20.5         17.4      -3.9         -4.0       -2.4
Private sector          10.8        18.6         13.7       10.3        15.0        12.7
Pakistan (1981)
Government, public
enterprise            -4.0        19.2          3.2        2.5       -1.4          9.5
Private sector          10.0        10.8         14.0        7.6        22.2        12.9
n.a. Not available.
a. Year of economic crisis in parentheses.
Source: IMF, International Financial Statistics.
Fiscal Deficits in Adjustment Loan-Intensive Countries
The evidence for the loan-intensive countries, the thirteen countries that
received three or more World Bank adjustment loans, shows that the
reduction in the budget deficit closely followed the decline in the current
account balance from 1982 to 1985 (figure 3-2).4 Until 1985 the loan-
intensive countries reduced their fiscal deficit more than did countries
that received no adjustment loans. Since 1985, however, the loan-inten-
sive countries appear to be having greater difficulty than nonrecipients



24 ISSUES OF POLICY REFORM
Figure 3-2. Current Account and Public Sector Deficit as a Percentage
of GDP for Thirteen Adjustment Loan-Intensive Countries, 1980 and
1982-87
i                 Publc sector deficit, a
6-
4 
Current account
1980   1982   1983   1984   1985   1986   1987
Source: World Bank data.
in controlling fiscal deficits (figure 3-3). The divergence between the pub-
lic sector deficit and the current account balance has become substantial,
which implies increased pressure on the private sector to generate a net
savings surplus. Therefore domestic financing of the budget deficit rose
sharply in a number of countries.' As a result, some countries such as
Mexico and Zambia faced problems of high inflation, whereas others
such as Turkey experienced high real interest rates.'
Why were the loan-intensive countries, all of which had World Bank-
IMFadjustment programs with strong conditionality on fiscal deficits, not
able to maintain a lower fiscal deficit as well as nonrecipient Countries?'
One possible reason is that loan-intensive countries suffered larger terms
of trade shocks. A disaggregated look shows that among the loan-inten-
sive countries, only five-C6te d'lvoire, Malawi, Mexico, Philippines,
and Zambia-suffered a deterioration in the fiscal deficit. In C6te d'l-
voire, Malawi, Mexico, and Zambia fiscal deficits deteriorated as wors-
ening terms of trade led to falling revenues, while expenditure programs
continued because they could not be easily cut. In the Philippines the
increase in fiscal deficits was due largely to the increased losses of public
sector financial institutions, which had provided loan guarantees under
the Marcos regime. Since the loan-intensive countries suffered larger



Public Finance 25
Figure 3-3. Public Sector Deficit as a Percentage of GDP in Selected
Country Groups, 1980 and 1982-87
12 -- *-                                            
II..~~~~15 non~~~recipens
10 
a    9             JO
0                        25 adjustmn loan recipients
to   8 
-   7 -
6 -* 
S13 adju stment          _      /
loan-intensive countries
1980   1982  1983  1984   1985   1986  1W87
Source: World Bank data.
terms of trade shocks (see chapter 11) than did nonrecipient countries,
they faced greater difficulty in reducing fiscal deficits as well.
The link between the fiscal deficit and the terms of trade, both in levels
and changes, is verified by the following equations, which use pooled
time-series and cross-country data for 1980-87 for twelve loan-intensive
countries (t statistics in parentheses). The results show a significant neg-
ative relationship between the terms of trade index and the fiscal deficit.
Expenditures rose when the terms of trade improved, but were difficult
to cut quickly once terms of trade and consequently revenues fell (see
also figure 3-4).
(3-1)        PSD = 12.1478 -  0.0891 TOT + ZibiDi
(3.06)    (2.30)
R2 = 0.52    D.W. = 1.24
(3-2)     PSD =  1.99  -  5.8135 T6T +  0.6581 PSD(-1)
(3.20)   (1.86)         (8.63)
R2 = 0.55    D.W. = 1.96



26 ISSUES OF POLICY REFORM
Figure 3-4. Expenditure and Revenue as a Percentage of GDP among
the Thirteen Adjustment Loan-Intensive Countries, 1980 and
1982-87
28               - 00% 1 
Expenditure\ 0'
U                     ~~~~~~~~--
26-
24.
22-
20 - ~Revenue  
1980 1982 1983 1984 1985 1986 1987
Source: World Bank data.
where PSD = public sector deficit as a percentage of GDP
TOT = terms of trade index (1980 = 100)
TOT = percentage change in the terms of trade index
Di = country dummy variable.
Sustainable Fiscal Deficits and Growth
In the post-1982 adjustment period considered here, the adjustment to
lower budget deficits has come primarily through expenditure cuts (figure
3-4). These expenditure reductions have cut heavily into public invest-
ments (which experienced the largest percentage of reductions) and the
noninterest component of current expenditure (figure 3-5 and table 3-
2). These cuts have affected social and economic services and subsidies
and have important implications for future growth and equity that sug-
gest the need for a careful consideration of alternative policies.8 Because
of a variety of social and political pressures on the government, the re-
duction in expenditure has not always been rational and efficient.
In choosing between a reduction in expenditure or an increase in rev-
enue, some important macroeconomic implications need to be kept in
mind. One is the'possibility that an increase in revenue could lead to a



Public Finance 27
Figure 3-5. Capital and Current Expenditure as a Percentage of GDP
for Thirteen Adjustment Loan-Intensive Countries, 1980 and
1982-86
35
30 
25 25
o  20-
8  15 
10 
S
0
1980  1982  1983  1984  1985  1986
Capital           Current
expenditure       expenditure
Source: World Bank data.
reduction in the aggregate savings rate of the economy. Known as the
Please effect, this occurs when the public sector propensity to save is
lower than that of the private sector (Please 1967; Chhibber 1985).9 A
recent study (Balassa 1988) has shown a negative correlation between
growth and public consumption as a share of GDP. Many Asian econo-
mies, however, have raised taxes to finance public investment and have
thereby raised savings and investment rates in the economy. The choices
made in cutting expenditures or raising revenues also have wide-ranging
effects on trade balances and income distribution. Who benefits and who
loses from government expenditure and the manner in which expenditure
choices are financed are key factors in the ability of the government to
undertake reforms and address poverty and the social costs of adjustment
(Cornia, Jolly, and Stewart 1988). These issues have been neglected or
given insufficient attention in the design of adjustment programs.'°
Although in general it is not difficult to identify when fiscal deficits
are unsustainable-inflation accelerates, interest rates are high, and the
private sector gets crowded out or external imbalances grow-it is less



28 ISSUES OF POLICY REFORM
Table 3-2. Change in Real Public Expenditure by Functional Category
for Thirty-Four Randomly Selected Developing Countries, 1982-86
(unweighted average)
Percentage
Account                                    rate
Capital account                            -10.3
Current account                              0.1
Goods and services                        -4.3
Wages                                     - 7.0
Interest                                  81.3
Subsidies                                 - 1.7
Other                                     - 1.2
Change in real total expenditure            -2.4
Current account (functional categories)      0.1
General                                   -5.3
Defense                                   - 5.8
Education                                 - 5.7
Health                                    -0.8
Social security                           -4.9
Econoniic services                        - 7.7
Housing                                    4.4
Other social                              10.8
Other (including interest)                41.9
Source: iwF, Government Financial Statistics, various years.
easy to decide what constitutes a sustainable level. This level depends,
among other things, on how the private sector adjusts to changes in fiscal
policy. If a reduction in the fiscal deficit leads to lower private savings,
its effectiveness in lowering the current account deficit is reduced. Con-
versely, a country can run a higher budget deficit without adversely af-
fecting the current account if the private sector is willing to generate
additional net savings surpluses at reasonable rates of inflation and real
interest rates, as happened in Malaysia and Thailand, for example.
Whether the improvement in the current account is brought about by
a reduction in investment or an increase in savings is fundamental to the
medium-term success of the adjustment program. It is in this area that
potential conflicts arise between stabilization and adjustment. There is
little merit in having a low budget deficit and a low external account
deficit if the outcome is a low level of savings and investment in the
economy. Thus the quality of fiscal adjustment is as important as the
quantitative improvement in the deficit. The composition of public ex-
penditures and the nature of the government's tax and credit policies are
key factors determining the outcome of fiscal adjustment and are ad-
dressed in more detail in later sections of this chapter.



Public Finance 29
The sustainable level of the fiscal deficit, therefore, depends on such
factors as the cost and availability of external and internal finance, the
quality and composition of public resource mobilization and use, ac-
companying adjustment measures, and the behavioral responses of the
rest of the economy. Turkey's adjustment experience provides an inter-
esting example of the need to assess these factors carefully in judging the
sustainability of fiscal policy.
Turkey repeatedly exceeded IMF fiscal targets in the early stages of its
adjustment program. Undoubtedly, the financing needs commensurate
with larger public sector deficits generated higher medium-term inflation
and real interest rates. But the thrust of the program was growth-oriented,
centering on export performance and on keeping savings and investment
rates up. A key contributing factor was the substantial excess capacity
resulting from the heavy investments in the 1970s, which allowed for a
quick improvement in output and exports once the exchange rate was
realigned. A recent study (Anand, Chhibber, and van Wijnbergen 1990)
argues that in terms of forgone growth the cost of a more conservative
fiscal policy might have been high in the early stages of the adjustment
program. After 1985, however, as net external inflows declined, further
increases in the fiscal deficit could not be sustained. This combination
of higher fiscal deficits and lower external inflows has led, particularly
since 1985, to high real interest rates and inflation. Thus the concept of
sustainability requires careful case-by-case analysis and a focus on both
the short- and long-term consequences of deficit financing.
Consistency between Fiscal Policy and Other Measures
Consistency among various components of an adjustment program is a
major concern when designing such programs. Almost every component
of an adjustment program has fiscal implications. In particular, exchange
rate and trade policies, financial liberalization, and price liberalization
in selected sectors have large effects on the budget. Unless these effects
are anticipated and their influence incorporated into fiscal programming,
they can be destabilizing and disruptive and possibly lead to a collapse
of the program and to policy reversals.
EXCHANGE RATE POLICY. Consistency between fiscal and exchange rate
policies is a major issue in the design of adjustment programs. In general,
the need for coordination between the two is obvious. Devaluation is
unlikely to be effective unless accompanied by appropriate demand man-
agement policies, including a tight budget, to reduce aggregate demand
and imports and thereby decrease the pressure on the current account.
Without appropriate fiscal restraint, inflation increases and the real ef-



30 ISSUES OF POLICY REFORM
fective devaluation is quickly eroded, with few effects on resource allo-
cation.
The coordination of exchange rate and fiscal policies can become com-
plex since changes in the exchange rate affect the size of the budget deficit
through changes in revenue generated by customs duties and export taxes,
foreign debt service, transfers to enterprises, and subsidies (see the ap-
pendix to this chapter for the case of Zimbabwe). The effect of deval-
uation on debt service costs can be seen in table 3-2, which breaks down
real expenditures by category for thirty-four randomly selected devel-
oping countries. The table shows the large contribution of rising interest
costs to fiscal problems. If the impact of devaluation on the budget is
likely to be adverse, it may be counterproductive to recommend free
floating of the currency (such as an exchange auction). Movements in
the exchange rate and the budget deficit may reinforce each other in a
destabilizing manner. This happened in Bolivia, Sierra Leone, and Zam-
bia, where the fiscal situation deteriorated following the devaluation be-
cause of the loss of the black market premiums that had previously ac-
crued to the government (Pinto 1987; Kharas and Pinto 1987). In such
cases, a series of managed exchange rate changes programmed within an
overall reduction in the budget deficit may be a much safer alternative,
with some priority in sequencing given to deficit reduction.
TRADE POLICY. Important issues of consistency also arise with respect
to trade policy. The first stage of the World Bank's recommendations
for trade reform usually involves a shift from quantitative import re-
strictions to tariffs, which improves the fiscal position of the government.
Subsequent reforms that call for reductions in the tariff rate, however,
may require compensatory fiscal measures. Policy reversals may take
place if governments do not implement such compensatory measures, as
happened in Morocco.
PRICE POLICY. Sectoral price liberalization usually benefits the govern-
ment budget. In many cases, the benefit comes from a decline in the losses
of state enterprises. If a sector is a net taxpayer, however, price decontrols
can have an adverse fiscal impact. This is often the case for the agricul-
tural sector, where input subsidies are the rule and the prices received
by farmers are kept low through export taxes and the payment of below-
market procurement prices (see Chhibber 1985; Braverman and Hammer
1986). Thus price increases to the farm sector will adversely affect the
budget unless they are compensated for by a reduction in input subsidies.
Comprehensive price liberalization in several sectors often presents fewer
problems because the government's fiscal losses in some sectors could be
balanced by gains in others. Individual sectoral adjustment loans would,



Public Finance 31
however, need to direct attention to this issue to ensure that there were
no adverse effects on the government's fiscal position.
Some countries have imposed selective price controls in an attempt to
control inflation, often in the mistaken belief that inflationary expecta-
tions alone had contributed to high inflation. In Brazil under the Cruzado
Plan, price controls were put in place as a vigorous expenditure program
was being undertaken. In this case, price controls in the face of an 8
percent increase in wages led to an increase in public enterprise deficits.
A consumer boom eroded the trade surplus, and the program collapsed
with much higher inflation than before (Cardoso and Dornbusch 1987).
Recent studies (Cukierman 1988; Simonsen 1988) suggest, however, that
even orthodox stabilization programs are insufficient to control inflation
when inflationary expectations set in and, furthermore, that they are
often accompanied by a prolonged recession with heavy social costs. The
issue remains unsettled since the answer depends to a large extent on the
somewhat nebulous concept of the "credibility" of policies and how best
to attain it (Blejer and Cheasty 1988).
An issue of great concern to many governments is the short-run infla-
tionary consequences of price liberalization-the release of repressed in-
flation (Feltenstein and Farhadian 1987). In this case the monetary ov-
erhang which often accompanies price controls must be gradually
dissipated, and careful coordination of public enterprise price policy and
fiscal policy is needed. The immediate burst in prices following decontrol
and liberalization can lead to policy reversals if the eventual impact of
deficit reduction on inflation takes effect over a longer period. This is
what has led to the interruption in China's economic reforms and is a
key issue in planning reforms in Algeria, Angola, and much of Eastern
Europe.
Resource Allocation and Use in the Public Sector
The preceding section argued that stabilization and adjustment typically
require considerable changes in government expenditure policy both to
reduce unsustainable fiscal deficits and to enhance the impact of public
spending on growth and income distribution. Thus adjustment programs
have devoted considerable attention to four broad issues of public ex-
penditure: (1) the level and composition of recurrent expenditures, (2)
the size and allocation of public investment, (3) institutional processes
and mechanisms for budgeting, and (4) the fiscal consequences of public
enterprise operations. Each of these areas is discussed below.
Recurrent Expenditures
Table 3-2 shows the changes (unweighted averages) in real expenditures
in different categories for thirty-four randomly selected developing coun-



32 ISSUES OF POLICY REFORM
tries during 1982-86. Cuts have been made in every category of ex-
penditure except housing and other social services. Increasingly, concern
is being expressed that budgetary allocations for operations and main-
tenance are inadequate. The World Bank, on the basis of its detailed
sectoral knowledge, has at times set specific targets for increases in op-
erations and maintenance expenditures in the belief that the payoff from
such outlays would be high. In some cases, structural adjustment loans
have stipulated a target for increased real spending on social sectors, even
when the loan has also included a condition requiring a reduction in the
overall deficit or in recurrent expenditures.
Because of the large amounts that are involved, subsidies-their ef-
fectiveness, relevance, and costs-are among the recurrent expenditure
categories most often subject to scrutiny in adjustment programs (World
Bank 1987). Both explicit and implicit subsidies are covered and involve
a wide range of items, including foodstuffs. For most products, the even-
tual elimination of subsidies should be the goal. Where appropriate, sub-
sidies and cost recovery ought to be handled jointly. For some items,
particularly foodstuffs, however, the focus might justifiably be on better
targeting of beneficiaries rather than on the complete elimination of sub-
sidies, both to improve the distributional impact and to reduce or restrain
budgetary outlays. Although World Bank adjustment lending often con-
tains conditions related to a reduction in and rationalization of subsidies,
reform has proved to be difficult in some cases, such as fertilizers.
Public Investment
In many developing countries the level, composition, and institutional
processes of public investment are key aspects of fiscal adjustment. Re-
source constraints often dictate a relatively sharp cutback in the mag-
nitude of public investment (see figure 3-5). Such cutbacks generally have
a significant impact on the balance of payments because the import com-
ponent of this spending generally exceeds the recurrent expenditures com-
ponent. Within a reduced public investment program, priority is usually
given to projects with shorter gestation and higher rates of return. Many
developing countries lack a systematic process for evaluating public
projects and for establishing multiyear budgeting requirements.
In the early phases of the adjustment process, the emphasis is generally
on scaling down the size of the public investment program and shifting
the focus from the manufacturing sector to infrastructure and social sec-
tors. A core investment program consisting of a small number of projects
that can be rapidly completed is often devised. In Pakistan and Turkey,
for example, the governments have increased the emphasis on infrastruc-
ture, which has important complementary linkages with private sector
productivity. Some structural adjustment loans, such as one for Turkey,



Public Finance 33
contain highly specific quantitative conditions on public investment con-
cerning, for example, the size of the gap between public investment and
savings, a ceiling on the growth of public investment, and a reduction
in the number of projects in the program. Generally, however, structural
adjustment loans contain no explicit quantitative targets for public in-
vestment. Rather, conditionality takes the form of a requirement for re-
view and agreement on the size and composition of the program at some
future time.
As the process of adjustment and expenditure compression has pro-
gressed, concern has been expressed about the insufficiency of public
investment, particularly for infrastructure and social sectors. Thus some
recent adjustment loans have specified an increase in, or a minimum level
of, public investment; those for Chile and Ghana are examples. Reducing
the share of public investment allocated to the manufacturing sector in
order to free resources for these other sectors has sometimes been a related
feature of conditionality.
Budgetary reforms, particularly improvements in the process of in-
vestment planning and priority setting and in capital budgeting and con-
trol, are widely emphasized in adjustment programs and loans. For ex-
ample, one for Ghana required implementation of the IMF technical
assistance recommendations on budgeting and expenditure control and
stipulated further measures for strengthening institutional capacity for
budgeting. One for C6te d'Ivoire contained provisions for strengthening
the capacity for project identification, economic evaluation, and super-
vision in technical ministries as well as the use of "agreed criteria" for
appraisal of economic and social projects. This focus on institutional
issues in the budget area arises from the belief that expenditure reforms
are unlikely to be durable unless the related institutional capacity is
strengthened. This type of institutional change is often difficult to effect
given the shortage of skilled personnel and the low government pay scales
in most developing countries. Accordingly, sufficient time must be al-
lotted for the implementation of such changes.
Fiscal Consequences of Public Enterprise Operations
In many developing countries, the financial performance of public en-
terprises has a major bearing on the government budget and the allocation
of resources between public and private sectors.1" Budgetary transfers to
finance the investment outlays of public enterprises or their operating
losses are often a major drain on public finances. Accordingly, the reform
of public enterprises, including investment size and financing, budgetary
transfers, pricing, personnel, and divestiture, is an important component
of fiscal reform and conditionality in structural loans.



34 ISSUES OF POLICY REFORM
Price and rate increases figure prominently in conditionalities related
to public enterprises. The intent of such increases is gradually to phase
out subsidies, which are particularly common for petroleum products,
electricity, and fertilizers, and to bring the prices of petroleum products
and other homogeneous tradables to international levels. Considerable
progress has been made in raising petroleum prices, but phasing out
fertilizer subsidies has proved to be more difficult, as the cases of Pakistan
and Turkey show.
The financial performance of public enterprises, often measured rela-
tive to a targeted rate of return or level of self-financing, is a common
conditionality. What seems often to be overlooked is the possibility that
price or rate increases by public enterprises enjoying market power may
actually work against improvements in efficiency. This issue arises par-
ticularly in the context of electric power utilities and other natural mo-
nopolies. Improvements in resource utilization among utility companies
have proved elusive, for example, in Bangladesh, Colombia, and Indo-
nesia, where rates of power loss remain very high despite efforts to stem
the losses.
Reform of Taxation
In many developing countries the tax system distorts incentives by its
influence on the relative prices of commodities and factors of production.
Moreover, tax revenues are often insufficiently buoyant with respect to
growth in output and income and do not adequately serve equity objec-
tives. Typically, a large proportion of tax revenues is raised through taxes
on international trade, particularly imports, and excise taxes. For ex-
ample, 38 percent of tax revenues in low-income countries in 1985 orig-
inated from taxes on international trade and 13 percent from excise taxes
(table 3-3).
Existing Structure of Taxation
Changes in tax systems that reduce distortions in production and improve
incentives for investment and savings can go a long way toward pro-
moting structural adjustment and economic growth. Moreover, the rev-
enue elasticity and equity aspects of the tax system can be enhanced by
changes that broaden the tax base, simplify the tax system, and strengthen
tax administration. Thus adjustment programs have increasingly come
to include elements of tax reform. In designing a tax reform package,
however, tradeoffs among various objectives are inevitable. These ob-
jectives include revenue generation, efficiency, equity, and administrative
simplicity. In addition, administrative constraints often significantly limit
the overall design and pace of reform. Nevertheless, theoretical insights



Public Finance 35
Table 3-3. Composition of Tax Revenue for Sample Groups of
Countries by Income Level, 1975 and 1983
(unweighted average)
Low-incomeb       Middle-income'       Industriald
Tax category'           1975     1985      1975     1985      1975      1985
Domestic income          28       26        30        32       35        35
Individual              9         9        8        10       27        27
Corporate              18        15       17        17        7         7
Other direct taxes        5        4        18        18       33        34
Social security         1         1       12        11       29        31
Property                2         1        3         2        2         2
Domestic commodity       28        32       27        31       29        29
Sales, VAT, turnover   13        17        9        13       16        17
Excise                 13        13       12        12       10        10
International trade      39       38        25        19        4         2
Import                 25       28        20        17        4         2
Export                 11         8        4         1        0         0
a. Figures for subcategories do not add up to the total given for each category because of the
presence of small unallocated taxes.
b. Includes eleven countries.
c. Includes thirty-six countries.
d. Indudes nineteen countries.
Source: World Bank (1988).
and experience provide broad guidelines for reforming tax systems. The
desirable changes include:
* Shifting away from taxation of production to taxation of consumption
in the form of a value added tax coupled with excise taxes on a few
luxury items so as to achieve greater equity
* Broadening personal income taxation to include income from all
sources, with a top marginal rate significantly below that found in
many industrial and developing countries, and with very few exemp-
tions and rates
* Simplifying and rationalizing corporate taxation so that there are very
few rates, a relatively low marginal effective rate, and neutrality with
respect to sectors, assets, and the financing composition of investment
* Eliminating the double taxation of dividends
* Introducing inflation accounting for both personal and corporate taxes
* Reducing and simplifying tariffs so that they are based on ad valorem
rates and serve protective rather than revenue-raising purposes, and
replacing high tariffs on luxury goods with excise taxes
* Strengthening the tax administration system overall.
Even though few countries have successfully carried out such com-
prehensive tax reforms, it nevertheless remains important that even par-



36 ISSUES OF POLICY REFORM
tial reforms be based on a well-designed overall tax system. In particular,
reform of international trade taxes needs to be closely coordinated with
reform of domestic trade taxes.
Experience with Tax Reform in Adjustment Programs
Because of large and unsustainable fiscal deficits, many developing coun-
tries have allowed short-run considerations of revenue to temper their
fundamental tax reform plans. Indeed, given the urgency of reducing the
fiscal deficit, short-term revenue-enhancing measures have often domi-
nated tax policy initiatives in recent years. These measures have fre-
quently included an increase in excise and international trade taxes be-
cause of the relative ease of administering such taxes. Malawi, for
example, increased import duties and excise taxes on selected goods in
the early 1980s. Other countries with similar experiences include Ar-
gentina, Bangladesh, the Philippines, Thailand, and more recently Mo-
rocco, which increased a special surtax on imports. These ad hoc changes,
however, further complicate the tax structure and, because of increased
protection and cascading effects, have adverse implications for incentives.
Changes in categories of tax revenues as a percentage of GDP from
1982 to 1986 for a selected group of countries are shown in table 3-4.
Few countries have significantly increased their tax/GDP ratio. This find-
ing underscores the earlier obseration that expenditures have borne the
brunt of fiscal deficit reduction. Simply to maintain the tax/GDP ratio,
many countries have introduced ad hoc measures, as noted above. The
inelasticity of the tax systems in many developing countries has given
rise to the need for frequent ad hoc changes in the tax rates or the tax
base to ensure that tax revenues increase in line with GDP.
Reform of international trade taxes in the context of trade liberali-
zation programs has been the most common tax change aimed at im-
proving resource allocation. Generally, the emphasis has been on con-
verting specific duties to ad valorem duties and reducing the level and
dispersion of tariff rates. Converting quantitative restrictions to tariffs
has typically been an accompanying measure. Virtually all structural ad-
justment loans and trade or industry sector adjustment loans have con-
tained extensive conditionalities with respect to tariff regimes. By and
large, significant progress has been made on this front. However, in some
cases (for example, Kenya in 1983), trade liberalization has suffered set-
backs in part because tariff reductions negatively affected revenue during
a period when a foreign exchange shortage caused imports to decline. In
cases such as Kenya, the reduction and rationalization of tariffs were
initiated but compensatory revenue measures were not instituted. In other
cases, such as the Philippines and Thailand, tariff reform was not un-
dertaken because the negative effect on revenue was foreseen. Similarly,



Public Finance 37
Table 3-4. Broad Categories of Government Revenue as Percentage of
GDP in Selected Countries, 1982 and 1986
Inter-
Domestic   Domestic   national
direct    indirect     trade    Total               Total
Country         taxes      taxes       taxes     tax    Nontax   revenue
Chile
1982            9.0       12.3        1.4      22.7      9.0       31.7
1986            6.3       14.1        3.0      23.4      5.6       28.9
Colombia
1982            2.6        2.8        1.8       7.4      0.9        8.3
1986            3.3        3.5        2.2       9.6      0.5       10.1
C6te d'lvoire
1982           4.1         5.6       10.1      21.3      6.4      27.7
1986            3.7        5.7        9.6      20.3     11.4       31.7
Jamaica
1982           11.9        8.9        3.0      23.9      0.9       24.8
1986           11.9       11.1        5.2      28.4      1.0       29.4
Kenya
1982           6.5         8.2        S.0      19.7      3.4       23.1
1986           6.8         9.9        4.4      21.4      3.2      24.6
Malawi
1982            6.4        S.9        4.2      16.7      2.9       19.6
1986            7.1        6.4        3.5      17.1      4.5       21.6
Mexico
1983           10.5        5.6        0.5      16.8      1.0       17.8
1986            8.0        6.1        0.9      15.0      1.0       16.0
Morocco
1982           4.6         8.7        6.4      20.1      2.6      22.7
1986            5.0        8.1        4.2      17.9      3.8       21.7
Pakistan
1982            2.9        4.9        S.9      13.6      2.9       16.5
1986            2.2        5.8        5.8      13.8      3.7       17.5
Thailand
1983            3.6        6.7        3.7      14.6      1.5       16.1
1986            3.4        7.9        3.5      15.6      1.5       17.1
Source: World Bank and IMF data.
the removal of export taxes on agriculture is sometimes impeded by the
expected loss of revenue, particularly since imposition of a land tax to
balance the loss encounters powerful political opposition (for example,
in Argentina).
Reform of domestic indirect taxes has been less extensive than reform
of tariffs. Several countries, including Indonesia, Korea, and Turkey, have
recently introduced a value added tax (vAT) as the main instrument for



38 ISSUES OF POLICY REFORM
rationalizing their system of indirect taxation. More than thirty devel-
oping countries now have some form of VAT. Some VATS apply to the
manufacturing stage (for example, in Indonesia) while others extend to
wholesale and retail stages (for example, in Korea and Turkey). Expe-
rience with the VAT has been favorable, even in the short-run generation
of revenue. Although the IMF has long advocated the introduction of a
VAT as the most desirable, broad-based, efficient form of commodity
taxation, not until the mid-1980s did the World Bank take a more active
role in advocating VATS. Until then, the World Bank had generally ac-
corded only limited attention to the taxation of goods and services, except
for tariffs, in its economic work and adjustment lending. An important
exception was the explicit support in structural adjustment loans to Tur-
key for the introduction of a VAT as early as 1981.
Progress in reforming personal and corporate income taxation has been
very limited. The inertia in streamlining income taxation derives from
the low share of income taxes in overall tax revenues, the difficulty of
enforcement, the opposition of powerful vested interests, and a genuine
desire to influence savings and investment decisions. Two broad areas
are particularly problematic in the income tax systems of most developing
countries. First is the large number of exemptions and incentives, which
are often intended to stimulate investment and savings or to serve other
policy objectives. Such provisions reduce the tax base and revenues (to
an often unknown extent), but there is very little empirical evidence that
they produce socially desirable results, at least in the case of investment
incentives, which are pervasive in some countries (for example, Morocco
and Turkey). A second problem with income taxation, particularly in
high-inflation countries, is the absence or inadequacy of indexation of
income and assets for purposes of taxation. High inflation can have other
negative impacts on revenue because collections lag or interest expenses
are fully deductible.
Growth-Oriented Fiscal Programs: Lessons of Experience
Fiscal adjustment and reform programs are difficult to design and im-
plement, given the complexity of underlying economic and institutional
factors and the presence of social and political sensitivities. The discussion
presented in this chapter points to several broad conclusions about the
nature and extent of fiscal adjustment in developing countries in this
decade:
* Fiscal policy reforms have often been triggered by actual or impending
budgetary and balance of payment crises. Under such circumstances,
short-run considerations have dominated the policy measures intro-
duced.



Public Finance 39
* A number of countries have undergone considerable adjustment with
respect to reductions in fiscal and external deficits, although several
countries experienced a reversal on the budget side after 1985. As
expected, the short-run impact of budgetary retrenchment has gen-
erally been recessionary, with a fall in investment and growth.
* Reduction of the fiscal deficit has been achieved primarily through
cutbacks in expenditures relative to GDP, particularly in public in-
vestment. These cutbacks have been accompanied by more focused
public investment programs, often a core set of projects and activities
in high-priority areas such as infrastructure. The allocation of recur-
rent expenditures has also been improved, although the reduction and
elimination of some subsidies (for example, on fertilizers) has proved
to be more difficult.
. With a few exceptions, revenues have not been very significantly in-
creased. Tax systems are not sufficiently elastic, and reductions in
imports due to foreign exchange shortages have resulted in the loss
of tariff revenues, a main component of tax receipts. Indeed, to main-
tain or modestly increase tax/GDP ratios, many countries have intro-
duced ad hoc tax measures that adversely affect efficiency and incen-
tives.
* Fundamental tax reform, except for tariffs, has not been a significant
aspect of fiscal adjustment.
Growth-oriented adjustment programs go considerably beyond the tra-
ditional stabilization approach to fiscal reform, which emphasizes mea-
sures to reduce aggregate demand. By definition, fiscal policy changes
aimed at augmenting aggregate supply, like other structural reforms, take
longer to bring about the desired results. Consequently, adjustment pol-
icies accenting fundamental fiscal reform require a medium-term per-
spective. Countries suffering from a severe macroeconomic imbalance
need to restore a reasonable degree of stability before structural measures
can be expected to elicit the desired response. But what constitutes a
reasonable degree of macroeconomic correction depends on a number
of factors, including the availability of external finance to support an
appropriate adjustment program. Countries committed to fundamental
fiscal reform-as well as to reform in other key policy areas-should
have access to the necessary financial support within a time frame that
allows them to initiate and sustain the process of change."2
In the case of the World Bank, structural adjustment lending, because
of its economywide approach, remains the most appropriate vehicle for
supporting fiscal reform. In principle, there is no reason why a series of
fiscal-based structural adjustment loans could not be designed for a coun-
try willing to undertake structural and institutional reforms in expend-
iture policy, public sector pricing, and taxation. Considerable prepara-



40 ISSUES OF POLICY REFORM
tion, in the form of policy studies, would need to precede such lending
operations, however. The World Bank has in recent years prepared many
reviews of public investment (expenditure), some of which have served
as background for structural adjustment loans. More sharply focused
studies of this type, including reviews of the revenue side, are needed to
underpin adjustment programs and lending. Loan conditionality would
need to be geared toward reform actions as opposed to necessarily short-
run improvements in key macro targets (for example, the budget deficit).
In countries with an acute stabilization problem, measures to reduce
aggregate demand would also be needed. Finally, multiyear financing
commitments would be required to help persuade governments to un-
dertake and sustain the reforms (Fischer 1986).
Appendix: Exchange Rates, Inflation, and Fiscal Adjustment
in Zimbabwe
Coordination of exchange rate and fiscal policies requires a careful as-
sessment of the various linkages through which each affects the other.
High fiscal deficits lead to high inflation. If the real exchange rate ap-
preciates and trade balances deteriorate, there will be a need for increased
foreign borrowing, a loss of foreign reserves, or import controls. Even-
tually, a country is forced into a devaluation in order to realign the real
exchange rate. How effective the devaluation will be in bringing about
a depreciation of the real exchange rate depends to a large extent on how
it affects domestic inflation. A key factor, as shown in a recent empirical
study on Zimbabwe (Chhibber and others 1989), is how the government
adjusts fiscal policy.
In the model used for the Zimbabwe study depreciation of the exchange
rate affects inflation in three ways. The first is a direct cost-push effect
through increased import costs. On average, a 1 percent increase in im-
port costs as a result of devaluation would raise domestic inflation by
0.2 percent. The second way is through the fiscal deficit: the change in
the exchange rate affects various components of the budget, particularly
foreign debt service, trade taxes, and import costs of public enterprises;
and there is a second-round effect on the fiscal deficit from the initial
increase in prices (Tanzi effect; see Tanzi 1977). The impact of these
changes in the fiscal deficit on inflation depends at least in the medium
term on how the budget deficit is financed. The third way is through the
positive effect on output of depreciation of the real exchange rate; these
changes in output then contribute to a reduction in inflationary pressures.
The key parameter here is the magnitude and speed of the export response
to changes in the real exchange rate.
On the assumption that 25 percent of the increase in the budget deficit
due to a devaluation is monetized (this is the historical average), that the



Public Finance 41
balance is financed by foreign and domestic borrowing, and that the
export elasticity is around 0.2 in the short run and 1 in the long run, the
Zimbabwe model shows that a 1 percent annual increase in the rate of
devaluation raises domestic inflation by 0.4 percent each year. Given
foreign inflation, the real exchange rate depreciates by 0.6 percent a year.
Note that 0.2 percent of inflation comes from a direct cost-push effect,
while the remaining 0.2 percent comes from linkages through the fiscal
deficit and effects on real output.
If the option to borrow (foreign or domestic) is limited, the results
change dramatically. When half instead of a quarter of the budget deficit
is monetized, domestic inflation rises by 0.6 percent a year for a 1 percent
annual increase in devaluation. This simulation shows one reason why
adequate financing of an adjustment program is crucial and why deval-
uations erode quickly when they lead to higher budget deficits. The real
effects of devaluation are completely eroded if adjustment lags on rev-
enues are longer. For example, if customs duties are not collected ad
valorem and other revenues adjust to inflation over four years instead
of about two years, the higher fiscal deficit would completely erode the
positive effects of the devaluation. There is no change in the real exchange
rate.
Notes
John Brondolo and Reza Firuzabadi provided excellent research assistance. The
authors are grateful to Bela Balassa, Francis Colaco, Jaime de Melo, Catherine
Mann, William McCleary, Pradeep Mitra, and Vinod Thomas for comments on
an early draft.
1. The term "budget deficit," unless stated otherwise, refers to the public
sector borrowing requirement. There is a growing body of literature (see Tanzi,
Blejer, and Teijeiro 1987) on different concepts of the fiscal deficit. Although
these concepts are all useful in revealing the fiscal stance of the government, it
is the public sector borrowing requirement that reflects the financing requirement.
2. The link between the growth of debt and the fiscal deficit has been em-
pirically verified (Tanzi and Blejer 1986).
3. A fourth way is through the buildup of arrears, which has emerged in recent
times as a significant fiscal issue. Arrears can be treated as forced borrowing by
the government, but the accumulation of arrears can be only a temporary measure
since it destroys the government's credibility.
4. The data for this section are from the World Bank country briefs and not
from International Financial Statistics or from central World Bank data sources.
Accurate information on the public sector deficit was available for only twenty-
five adjustment loan recipients (those with one or more adjustment loans since
1980) and for fifteen nonrecipients. Of the twenty-five adjustment loan countries,
thirteen received three or more adjustment loans and are referred to here as
adjustment loan-intensive countries.
S. Results are available from the authors.
6. Turkey's inflation rate was also high, but not explosive, averaging 30 to
40 percent a year between 1980 and 1986. It has since accelerated to over 60
percent a year.



42 ISSUES OF POLICY REFORM
7. An important qualifier of this comparison is that many nonrecipient coun-
tries, and the less loan-intensive countries, had IMF programs in place.
8. In the thirteen adjustment loan-intensive countries considered here, neg-
ative growth in GDP per capita was observed in Chile, C6te d'lvoire, Jamaica,
Malawi, Mexico, the Philippines, and Zambia between 1980 and 1986. Only
Korea, Pakistan, Thailand, Turkey, and Morocco (marginally) showed positive
growth in GDP per capita.
9. For an empirical verification of the Please effect, see Chhibber (1985).
10. For details see chapter 12 by Elaine Zuckerman.
11. This discussion is limited to fiscal issues. Chapter 6 treats the issues related
to public enterprise reform in more depth.
12. The World Bank and IMF (see Tanzi 1987) have increasingly recognized
the need for refocusing fiscal reform packages to include supply-side measures.
The World Bank is proposing a change in its policy in this direction, as evidenced
in recent internal policy documents.
References
Anand, Ritu, Ajay Chhibber, and Sweder van Wijnbergen. 1990. "External Bal-
ance, Fiscal Policy and Growth in Turkey." In Torsun Aricanli and Dani Rod-
rik, eds., The Political Economy of Turkey: Debt, Adjustment, and Sustain-
ability. London: Macmillan.
Balassa, Bela. 1988. "Public Finance and Economic Development." PPR Working
Paper 31. Policy, Planning, and Research, World Bank, Washington, D.C.
Blejer, M. I., and Adreane Cheasty. 1988. "High Inflation and Heterodox Sta-
bilization and Fiscal Policy." World Development 16 (8):867-81.
Braverman, Avishay, and Jeffrey Hammer. 1986. "Multi-Market Analysis of Ag-
ricultural Commodities." Agriculture Department, World Bank, Washington,
D.C.
Buiter, W. 1985. "A Guide to Public Sector Debt and Deficits." Economic Policy
1 (November):13-19.
Cardoso, Eliana, and Rudiger Dornbusch. 1987. "Brazil's Tropical Plan." Amer-
ican Economic Review 77 (May).
Chhibber, Ajay. 1985. "Taxation and Aggregate Savings: An Econometric Anal-
ysis for Three Sub-Saharan African Countries." CPD Discussion Paper Series
1985-35. Country Policy Department, World Bank, Washington, D.C.
. 1989. "The Aggregate Supply Response in Agriculture: A Survey." In
Simon Commander, ed., Structural Adjustment in Agriculture: Theory and
Practice. London: Overseas Development Institute.
Chhibber, Ajay, Joaquin Cottani, Reza Firuzabadi, and Michael Walton. 1989.
"Inflation, Exchange Rates and Fiscal Policy in Zimbabwe." PPR Working
Paper 192. Policy, Planning, and Research, World Bank, Washington, D.C.
Cornia, Giovannini, Richard Jolly, and Frances Stewart, eds. 1987. Adjustment
with a Human Face. Oxford, U.K.: Clarendon Press for UNICEF.
Cukierman, Alex. 1988. "The End of High Israeli Inflation: An Experiment in
Heterodox Stabilization." In Michael Bruno, Guido di Tella, Rudiger Dorn-



Public Finance 43
busch, and Stanley Fischer, eds., Inflation Stabilization: The Experience of
Israel, Argentina, Brazil, Bolivia, and Mexico. Cambridge, Mass.: MIT Press.
Feltenstein, Andrew, and Ziba Farhadian. 1987. "Fiscal Policy, Monetary Tar-
gets, and the Price Level in a Centrally Controlled Economy: The Case of
China." Journal of Money, Credit, and Banking 19 (2).
Fischer, Stanley. 1986. "Issues in Medium-Term Macroeconomic Adjustment."
World Bank Research Observer 1 (2) July: 163-82.
International Monetary Fund. Various years. International Financial Statistics.
Washington, D.C.
Kharas, Homi, and Brian Pinto. 1987. "Exchange Rate Rules, Black Market
Premia, and Fiscal Deficits: The Bolivian Hyperinflation." cpD Working Paper.
Country Policy Department, World Bank, Washington, D.C.
Pinto, Brian. 1987. "Black Market Premia, Exchange Rate Unification, and In-
flation in Sub-Saharan Africa." Country Policy Department, World Bank,
Washington, D.C.
Please, Stanley. 1967. "Savings Through Taxation: Reality or Mirage." Finance
& Development.
Simonsen, M. H. 1988. "Price Stabilization and Incomes Policies: Theory and
the Brazilian Case Study." In Michael Bruno, Guido di Tella, Rudiger Dorn-
busch, and Stanley Fischer, eds., Inflation Stabilization: The Experience of
Israel, Argentina, Brazil, Bolivia, and Mexico. Cambridge, Mass.: NIT Press.
Tanzi, Vito. 1977. "Inflation, Lags in Collection, and the Real Value of Tax
Revenues." IMF Staff Papers 24.
. 1987. "Fiscal Policy, Growth and Stabilization Programs." Finance &
Development (June).
Tanzi, Vito, and M. I. Blejer. 1986. "Public Debt and Fiscal Policy in Developing
Countries." IMF Working Paper WP/86/5. Washington, D.C.
Tanzi, Vito, M. I. Blejer, and M. 0. Teijeiro. 1987. "Inflation and the Mea-
surement of Fiscal Deficits." IMF Staff Papers 34 (4):711-38.
World Bank. 1987. "Bank Treatment of Subsidies." Country Policy Department,
Washington, D.C.
. 1988. World Development Report 1988. New York: Oxford University
Press.



44 ISSUES OF POLICY REFORM
Comments
Peter S. Heller
IN THIS CHAPTER, Ajay Chhibber and Javad Khalilzadeh-Shirazi analyze
the role of fiscal reform in adjustment programs. They initially discuss
the macroeconomic links between fiscal reform and the adjustment pro-
cess, arguing that the crises that engender adjustment programs have
often been related to exogenous circumstances (for example, a collapse
in world commodity prices, a surge in real interest rates, or weak export
demand). Fiscal deficits that were previously viable have become unsus-
tainable and have brought on some combination of excessive external
borrowing, inflation, and constraints on credit to the private sector. The
authors argue that stricter fiscal discipline is not a sufficient condition
for sustainable adjustment, but that a comprehensive policy package is
required to stimulate savings and investment. They then discuss public
expenditure policy and tax reform in the adjustment process, arguing
that fiscal adjustment has been accomplished primarily through expend-
iture cuts, since the scope for increasing revenue has been limited. The
authors draw on a cross-country analysis to suggest a large number of
stylized facts that are used as the basis for policy recommendations.
My comments fall into two broad categories, one relating to substan-
tive issues and the other relating to the methodology of the study. On
the substantive side, I have five points to raise. First, the authors argue
that the quality of fiscal adjustment matters as much as, if not more than,
the quantity of adjustment. This is a view that the IMF has also come to
recognize and consider in the design of programs. A short-run emphasis
on deficit-cutting measures may lead to the adoption of policies that in
the medium run exacerbate rather than reduce the budget deficit, thus
weakening the prospects for growth. The problem, of course, is that in
the short run the adoption of important fiscal reforms or of a higher
investment effort may be costly and, in the absence of additional budget
financing, not feasible.
The authors seem to suggest that when additional investments can be
financed with external funds on concessional terms a larger budgetary
deficit may be acceptable within the framework of an adjustment pro-
gram than would otherwise be the case. Yet even when additional in-



Public Finance 45
vestment can be financed fully from a highly concessional loan or grant
there may be reasons to be cautious about accepting a higher budget
deficit. In the medium term, can the government meet the higher oper-
ational and maintenance costs associated with the investment? What ad-
ditional domestic outlays are needed to complement the investment? Will
the government be able to meet the higher debt-servicing that may be
required in the medium term?
Second, the authors seem to equate policies of adjustment with in-
vestment-led growth. They argue that "there is little merit in having a
low budget deficit and a low external account deficit if the outcome is
a low level of savings and investment in the economy." Certainly, ad-
justment ought to facilitate an economic environment that is supportive
of higher medium-term growth. Such higher growth may derive from
more investment, but it may also stem from a strengthening of the per-
formance of markets, the more rapid introduction of technological
change, a higher current expenditure on operations and maintenance,
and so on. The temptation to opt for higher investment and to neglect
operations and maintenance is one of the biases of an excessive emphasis
on public savings and investment.
Third, in the case of Turkey, the authors seem to assert that the initially
high deficit, in excess of IMF targets, was desirable because of its positive
impact on growth but that it ultimately contributed to higher real interest
rates and accelerating inflation. If the latter is true, then the cost of achiev-
ing the higher growth may in fact have been too high relative to its
medium-term effects. The policy lessons that emerge from this example
are unclear. If it is difficult for a country to fine-tune its fiscal policies,
it may prove more sensible in the medium run to err on the conservative
side in the initial policy stance.
Fourth, in discussing the adverse budgetary effects of a devaluation
coupled with price liberalization, the authors fail to mention how det-
rimental, in the long run, artificial prices can be in eroding the tax base.
To suggest that price liberalization for tax-paying sectors is likely to be
bad for the budget neglects the fact that price liberalization in repressed
sectors can lead to an improvement in public finances.
Fifth, the authors appear to treat the consequences of unsustainable
deficits as separable, depending on how they are financed: inflation if
financed through money creation, constrained credit to the private sector
if through domestic debt creation, or excessive external debt if through
external financing. However, the manner in which the deficit is financed
is likely to be of less importance than the many consequences of the deficit
(inflation, depreciation, and so on).
Methodologically, the authors rely heavily on cross-country analysis
rather than on individual case studies. This approach has its limitations.
For example, the fact that adjustment loan-intensive countries were able



46 ISSUES OF POLICY REFORM
to maintain their cuts in the deficit relative to the experience of other
countries does not really say much about the adjustment process. In many
respects an approach using case studies would be more useful, since it
would allow a coherent description of the fiscal experience of the coun-
tries in question and the possibility of a careful analysis of the relation-
ships between the fiscal and overall adjustment processes. Similarly,
rather than simply focusing on the postcrisis period, it would be useful
to have a discussion of the events leading up to the crisis that precipitated
the adjustment process. This discussion would allow the reader to better
understand the way a country's fiscal policies help precipitate such crises
and would, in turn, assist in an analysis of the appropriate fiscal response.
This approach would also allow some assessment of the "counterfac-
tual," that is, of what might have happened in a particular country in
the absence of an adjustment program.



4 Trade Policy Reform
Vinod Thomas
DURING THE 1980s, many developing countries received financial and
policy support from the World Bank and the International Monetary
Fund (IMF) for trade reforms. This chapter evaluates these policy reforms
using cross-country data and country studies. The analysis considers re-
forms in the forty countries that received trade adjustment loans during
1980-87 and assesses reform implementation in the twenty-four coun-
tries for which sufficient data are available.' The effect of reforms on
incentives is also examined, as are changes in economic performance in
countries that have received trade adjustment loans and have carried out
reforms.
Extent of Reforms
Halevi (1989) considered the following restrictions in commercial policy
in the forty countries: export impediments,2 import impediments on in-
puts used in export production, quantitative restrictions3 on both non-
competitive and competitive imports, and tariff rates4 and rate dispersion.
The countries were grouped into three categories according to judgments
on the relative antiexport bias before reform for each country: low, me-
dium, or high.' Chile and the Republic of Korea had a relatively low
level of restrictions; 60 percent had a high level, and 35 percent had a
medium level.
Degree of Restrictiveness
A comparison with the restrictiveness of industrial countries provides a
perspective on that of developing countries. The weighted average tariff
rate for fifty developing countries was 26 percent at the end of 1985
according to Erzan and others (1986). Adding other import charges raises
the figure to 34 percent. For members of the Organisation for Economic
Co-operation and Development (oEcD), average tariffs on industrial
goods were estimated to be about 5 percent according to a report of the
General Agreement on Tariffs and Trade (GATr 1980) and Finger and
Laird (1987) and are roughly of that order today. Erzan and others es-
47



48 ISSUES OF POLICY REFORM
timated the coverage of nontariff barriers in the same fifty developing
countries to be 40 percent (unweighted) of import items corresponding
to all tariff positions at the end of 1985. Finger and Laird provide a
similar estimate for thirty-eight developing countries for 1982. They also
estimated that 15 percent of the product categories of the eleven industrial
countries were subject to nontariff barriers in 1984. Laird and Yeats
(1988) provide a similar figure (15.9 percent) for all products in fourteen
industrial countries in 1986.6 These estimates indicate that developing
countries have a much more restrictive trade regime than do industrial
countries.
What Was Proposed
The intensity of reform proposals was also grouped into three categories:
low, medium, or high. This intensity was considered as strong as the
initial degree of restrictiveness in more than half the forty countries that
received trade adjustment loans. Of the twenty-four countries with high
initial restrictiveness, reform proposals were strong in twelve (for ex-
ample, Ghana, Jamaica, Mexico, and Turkey), while in six they were
only mild (for example, Brazil, Guyana, and Pakistan). Among the four-
teen cases with moderate initial restrictiveness, nine had moderate or
strong proposals. In general, the intensity of the proposals was com-
mensurate with the initial degree of restrictiveness more frequently in
export policy than in import policy.
The main components of trade policy proposals under adjustment lend-
ing are summarized in table 4-1. Although policy packages are not uni-
form there are common threads. One is a reduction in restrictions on
exports and imports, and another is greater reliance on the price mech-
anism, that is, on exchange rate depreciation and the use of tariffs in
place of quantitative restrictions.7 The loan proposals were most con-
sistent in their attempt to reduce direct impediments to exports and re-
strictions on imported inputs for export production. Reform of exchange
rate policy was almost always a stated, or unwritten but important, goal.
Almost all loans supported a greater use of price mechanisms, as well as
reductions in the level and dispersion of tariff rates. Proposed reductions
in quantitative restrictions on imports were large in some cases but mod-
est on average in the case of items both competitive and noncompetitive
with domestic production.
There has been less attention, however, to reforms that would promote
greater internal competition. Because most proposals were put together
quickly, as is usual in loans for direct balance of payments support, they
often included plans for studies to identify needed future actions. Not
much evidence is available for assessing progress on these studies. Pro-
posals to reduce protection for import substitutes have been cautious.



Trade Policy Reform 49
Table 4-1. Intensity and Distribution of Proposals for Trade Policy
Reform  among Forty Countries Receiving Trade Adjustment Loans
(number of countries)
Mild
Not                       or
Area of reform                        Present present Strong Moderate absent
Exchange rate'                          38       2
Export promotionb                       33       7
Protection studies                      28      12
Overall export policy                                   15       15       10
Imports for exports                                   17       15        8
Overall import policy                                   14       15       11
Nonprotective quantitative restrictions               14       16       10
Protective quantitative restrictions'                 14       15       11
Tariff levelc                                          7       21       12
Tariff dispersion                                      7       24        9
Schedule of future action                              6       29        5
Overall reduction in antiexport bias                    17       12       11
a. Often these were not explicit conditions, but understandings under the program.
b. Removal of restrictions and provision of export credits, insurance, guarantees, institutional development,
and the like.
c. Where reforms include a replacement of quantitative restrictions, they are counted in both these lines.
Source: World Bank data.
Most programs envisioned that some level of effective protection would
continue indefinitely. In some cases, particularly in Sub-Saharan Africa,
additional incentives were introduced for import substitution-for ex-
ample, higher duties on imported inputs that compete with domestic
production. (Increasing the duties on imported inputs reduces the pro-
tection provided to finished goods that use them.)
Implementation Record
The twenty-four countries with data on implementation were ranked in
five categories according to the degree of implementation of the reform
proposals. Some examples are given below. The two countries that had
a low level of restrictiveness (Chile) or antiexport bias (Republic of Korea)
at the beginning of the 1980s were also assigned the highest ranking on
implementation. Success in implementation for the eight countries judged
to have a moderate level of restrictiveness covered the range from the
lowest (Malawi), through the middle of the range (Panama), to the highest
(Mauritius). Among the fourteen countries that initially had high levels
of restrictiveness, six (of the nine countries with strong proposals for
commercial policy reform) had the highest or next to highest ranking in



50 ISSUES OF POLICY REFORM
implementation (Ghana, Madagascar, Mexico, Philippines, Senegal, and
Turkey). Table 4-2 presents these examples in a three-way classification
that takes into account both the intensity of proposals and the degree of
compliance, that is, as a product of the constructed indices of proposals
and of implementation.
Although implementation was swift in exchange rate adjustment and
the removal of export restrictions, countries in the sample have been slow
to liberalize imports. Overall, trade reforms were moderately significant.
Substantial actions were taken in reducing export restrictions (licensing,
prohibitions, and export taxes). Restrictions on imported inputs for ex-
ports have also been significantly reduced. On the import side, several
countries Jamaica, Mexico, Senegal, and Turkey) have made progress
in switching from quantitative restrictions to tariffs. Many countries have
adopted tariff reform programs. Progress has been most notable in re-
ducing maximum tariff rates, limiting the number of tariff classes, es-
tablishing a (low) minimum tariff, and reducing tariff exemptions.
The lowering of protection, however, has been modest on average.
Most trade regimes continue to maintain escalated tariff structures, with
higher tariffs (and quantitative restrictions) on final goods than on capital
goods and low rates (and exemptions) for intermediate and raw materials.
Tariff dispersion has usually been reduced, but dispersion in effective
protection is still large. This experience supports the conclusion of Pa-
pageorgiou, Michaely, and Choksi (forthcoming) that commercial lib-
eralization is a drawn-out process. For instance, of the fourteen countries
with highly restrictive trade regimes in the early 1980s, only four (a-
maica, Mexico, Senegal, and Turkey) had achieved a high degree of com-
mercial liberalization by 1987-88.
Stronger reforms have been implemented in exchange rate policy than
in commercial policy. There was a larger depreciation in the real exchange
rate in most of the countries receiving trade adjustment loans than in
most of the others. In part this was because of the higher debt and greater
external shocks in the former, which required a larger depreciation, and
in part because of exchange rate and macroeconomic reforms accom-
panied by some trade liberalization.8 A real depreciation of the currency
is an important liberalization measure. In the presence of binding quan-
titative restrictions on imports, it increases the price not only of tradables
relative to nontradables but also of exportables relative to importables,
thereby reducing antiexport bias. Moreover, a large depreciation can
eventually make quantitative restrictions redundant and thereby result
in a de facto liberalization of the import regime.
Table 4-2 shows the extent of exchange rate depreciation and the in-
tensity of commercial policy reforms among the twenty-four countries.
The latter summarizes the outcomes of policy changes-taking account
of what was proposed and what was implemented-in three categories:



Table 4-2. Intensity of Reforms in Trade Regimes in Twenty-Four Sample Countries, 1980-87
Reduction in antiexport bias through commercial policy (export and import)
Exchange rate                                                      --
depreciation'                                  Mild                                   Moderate                             Substantial
Mild (no depreciation or         Guyana*                             C6te d'lvoire, Senegal
appreciation)
Moderate (less than 20 percent)    Kenya, Malawi, Togo, Yugoslavia,'   Bangladesh, Madagascar, Morocco, Panama,    Jamaica, Korea," Mauritius
Zimbabwe                            Philippines, Thailand
<'     Substantial (20 percent or more)   Pakistan, Zambia*                  Colombia, Ghana                                Chile,b Mexico, Turkey
. Aborted, reversed, or no reform of commercial policy.
Note: The table indicates changes after reform compared with the prereform situation. In the case of commercial policy, the judgment of the prereform situation is
derived from accounts in World Bank reports. In the case of the exchange rate, it is based on trends in purchasing power parity over the long term. Some important
changes since 1987 are not captured in this table, such as improvements in Morocco and reversals in Turkey, but some major improvements that occurred in 1986-87
are emphasized (for example, Jamaica and Mexico in commercial policy). Some important reforms are not included in this sample because adjustment lending for trade
policy was too recent (Indonesia, Uruguay) or because the reform was not in connection with adjustment lending (Bolivia, Haiti).
a. Based on a measure of the average change in real effective exchange rate during the period after the first loan compared with the period from 1965 to the year before
the first loan.
b. Chile and Korea had already achieved substantial reforms by the early 1980s.
Sources: World Bank and IMF data.



52 ISSUES OF POLICY REFORM
mild, moderate, and substantial accomplishments. The table focuses on
changes after the trade loan. But, by comparing these changes with the
situation prior to the loan, it also gives a rough idea of how much of the
problem was being addressed. However, the table does not fully capture
the extent of the initial problem. For example, both Korea and the Phil-
ippines depreciated the real exchange rate moderately compared with
their long-term levels, but the adequacy of the long-term levels in each
case may have been different. Subject to this caveat, the table distinguishes
strong reformers of commercial policy and reformers of exchange rate
policy from weak reformers. (For a qualitative discussion of rankings
according to both the initial problem and the reform, see table 13-3.)
Progress and Constraints to Implementation
The extent of implementation has varied greatly among countries and
policy areas. Overall, price reforms have been relatively substantial under
trade adjustment programs. Examples include removal of export taxes,
introduction of duty drawback schemes for exporters, and more uni-
formity in tariffs. But there has been less success in institutionalizing and
sustaining some of the price changes. By and large, institutional reform
has been limited. There are many instances of abandonment, reversals,
and flip-flops in price policies. Despite modest goals, Yugoslavia aban-
doned reforms, C6te d'Ivoire and Kenya made slow progress, Morocco
and Thailand partially reversed their tariff policy reform, Argentina re-
versed its reform of quantitative restrictions, and Sierra Leone, Somalia,
Uganda, and Zambia reversed their policies of exchange rate auctions.
Unless changes appear to be sustainable, the credibility of actions and
the supply response to them are likely to be limited (Rodrik 1988).
Four factors have constrained more thorough implementation and sus-
tainability: weak macroeconomic performance, inadequate government
commitment to reform, difficulties in implementation, and conflicts
among policy reforms and weaknesses in design.
WEAK MACROECONOMIC PERFORMANCE. Economic instability and ex-
ternal imbalances are serious constraints to liberalization, while export
growth makes liberalization easier. Recession, inability to address infla-
tion, and real appreciation of the currency have inhibited trade reforms
to varying degrees in Costa Rica, Jamaica, Mexico, and the Philippines.
Balance of payments problems resulting from a fall in copper prices and
faulty exchange rate management contributed to the reversal of reform
policy in Zambia. Export performance and the availability of foreign
exchange also offset the sustainability of reform. A strong and rapid
supply response improves the sustainability of reforms by reducing their
transition costs associated with the release of resources from previously



Trade Policy Reform 53
highly protected sectors. Slow export expansion hurt Kenya's liberali-
zation attempts. Export diversification is just beginning in Costa Rica
and C6te d'lvoire, which made more rapid progress in commercial policy
reform but were also vulnerable to declining terms of trade. Even in Chile
the rapid growth of exports and the availability of foreign exchange have
been important in preventing policy reversals. In Jamaica the availability
of financing has been crucial for maintaining the liberalization effort in
the face of a worsening current account balance.
INADEQUATE GOVERNMENT COMMITMENT. In a number of cases in which
the government has not been fully committed to the program (Kenya,
Malawi, Zambia), implementation has been weak. The slow pace of re-
form has in turn sometimes hurt the credibility of the program for the
private sector, thereby diminishing its sustainability. Inadequate com-
mitment has limited the sustainability of reforms, particularly in the
highly indebted countries and in Sub-Saharan Africa. Changes in political
regimes and leadership have often compounded these problems and have
led to policy reversals. A related constraint is internal opposition to re-
form. There are always winners and losers from policy changes. Resis-
tance from losers, as in Zimbabwe, has delayed or reversed reductions
in protection. In Yugoslavia, despite modest goals related to trade and
the foreign exchange regime, political opposition (in addition to mac-
roeconomic instability) led to a dilution or reversal of most elements of
the program.
DIFFICULTIES IN IMPLEMENTATION. Sometimes a country's limited ad-
ministrative capacity has been a critical constraint. Bangladesh and C6te
d'lvoire made slow progress in part because of administrative difficulties.
The introduction of export tax rebates, duty drawback systems, and
bonded warehouses has been subject to administrative delays in many
cases. If changes in policy are to be successfully implemented, they often
require changes in administrative arrangements and capabilities (for ex-
ample, import administration may need to be reorganized to implement
tariff reforms). Sometimes, policy changes were predicated on the com-
pletion of studies, which were delayed for various reasons in a number
of the cases reviewed (for example, Colombia and Kenya). Inadequate
capacity for medium-term policymaking is a common problem.
CONFLICTS AMONG POLICY REFORMS. Inadequate stabilization efforts
have constrained trade reform in Pakistan and Panama. In contrast, sta-
bilization and trade reforms in the 1980s have reinforced one another
in Chile, Colombia, and Korea, where the ability to quickly regain sta-
bility in the external sector has helped to sustain trade reforms. The
targets of structural adjustment and stabilization have at times conflicted.



54 ISSUES OF POLICY REFORM
For instance, the imposition of customs duties and tariff surtaxes to in-
crease revenues for stabilization purposes in the Philippines has conflicted
with attempts to liberalize imports. To increase revenue Morocco raised
import tariffs, which had been reduced in an earlier phase of reform.
These conflicts may sometimes be unavoidable. Trade taxes create dis-
tortions, so less-distorting, alternative sources of revenue are preferable.
When a country has a weak tax system, however, some trade taxes may
remain necessary in the short term to generate revenue.
Effects of Policy Changes
The previous section reviewed the main elements and intensity of changes
in trade policy. In this section I survey the evidence on the effects of those
changes.
Change in Incentives
REAL EXCHANGE RATE. An indicator of the incentives for the production
of tradables relative to nontradables is the real exchange rate. Exchange
rate misalignments were significant in the early 1980s for the recipients
of trade adjustment loans. Subsequent adjustments were also substantial
in a large number of cases. They involved a series of devaluations or
institution of a crawling peg, supported by macroeconomic adjustments.
Figure 4-1 compares changes in a trade-weighted multilateral real ex-
change rate relative to major trading partners for a group of twenty-one
industrial countries, the forty recipients of trade adjustment loans, and
forty-eight nonrecipients. The domestic currency depreciated in real terms
by more than 22 percent between the periods 1981-83 and 1985-87 for
the group of forty trade adjustment loan countries, in contrast to 2 per-
cent in the nonrecipient countries and a slight appreciation in the in-
dustrial countries. This implies that the price of traded goods relative to
that of nontraded goods increased on average in the trade adjustment
loan countries.
Real exchange rate indices indicate the change in bias against tradable
goods, but they seldom distinguish between exportable commodities and
import substitutes.9 To make that distinction, measures of changes in the
levels of effective protection for the different sectors would be needed or
measures of effective exchange rates for exporting activities as opposed
to import-substituting activities. Individual country studies of effective
protection and antiexport bias exist (for example, for Chile, Colombia,
Kenya, Korea, Mexico, Morocco, Pakistan, Philippines, and Turkey), but
the results are not comparable across countries. Comparisons of even
nominal protection rates or the coverage of quantitative restrictions are
difficult. In a few cases, changes in protection levels over time have also



Trade Policy Reform 55
Figure 4-1. Real Exchange Rate Indices for Selected Country Groups,
1978-88
(unweighted averages)
Index 115-
110 
105-
100-_
90 \\
85 -
80" -
75 - 
70-                                
1978   79    80    81    82    83    84    85    86    87   88
-       21 industrial countries
- -   40 recipients of trade adjustment loans
47 nonrecipients of trade adjustment loans
Note: An increase in the index indicates real appreciation. This figure does not
indicate initial currency misalignments. For a discussion of "proper" levels of exchange
rates, see Williamson (1985).
Source: The trade-weighted multilateral index of rhe real exchange rate for the various
countries is based on IMF data.
been assessed, but intercountry comparisons of the changes are even more
difficult than comparisons of the levels.
IMPORT LIBERALIZATION AND PROTECTION. Overall import/GDP ratios
reflect the effects of the depreciation of the real exchange rate and the
availability of financing in addition to import liberalization. During the
1980s import levels in developing countries declined (in current and con-
stant prices) on average because of balance of payments problems, as did
import/GDP ratios. The ratio of nonfuel imports to GDP declined as well,
although the extent of the fall was less than for total imports. The re-
duction in the import/GDP ratio was significantly less, however, for coun-



Table 4-3. Imports of Goods and Nonfactor Services in Current Prices as a Percentage of GDP for Selected Country
Groups, 1980-88
(unweighted averages)
Percentage change
1984-861      1985-871
Sample group              1980      1981      1982     1983      1984      1985      1986      1987      1988a      1980-82       1981-83
10 intensive trade loan
cs     recipients               32.7      33.3      29.4      30.1      31.1     32.0      29.5      31.1      33.1         -2.9*         -0.1*
26 trade loan recipients  34.3      34.7      31.0     29.7      30.5      30.9      29.2      31.7      30.7        -9.3*          -3.8*
40 trade loan recipients  33.0      33.4      31.6      30.3     30.7      30.8      29.0      30.3      30.6        -7.7*          -5.5*
47 nonrecipients          38.8      40.4      38.8     35.6      34.1      33.5      33.1      32.2      32.3       -14.6         -14.0
87 developing countries   36.1      37.2      35.5     33.2      32.5      32.3      31.2      31.3      31.4       - 11.8t       - 10.St
21 industrial countries   35.5      35.7      35.1     34.5      36.5      36.9      33.2      32.6      31.5          0.3          -2.5
The difference in means between the recipients and nonrecipients of trade adjustment loans is significant at the 5 percent confidence interval.
t The difference in means between the recipients and nonrecipients of trade adjustment loans is significant at the 10 percent confidence interval.
a. Preliminary estimates.
Source: World Bank estimates.



Trade Policy Reform 57
tries associated with trade reforms and adjustment lending. As indicated
in table 4-3, the ratios declined systematically less among countries that
received trade adjustment loans than in the other countries, and these
differences were statistically very significant.
Direct examination of the conditions in trade adjustment loans and
their implementation records suggest that import protection on average
has fallen only modestly in most of these countries. By and large, tariff
structures remain escalated, with the highest protection afforded to final
goods. This seems consistent with the evidence on changes in the com-
position of nonfuel imports since 1980. If protection of the most pro-
tected goods (consumer goods) had been reduced substantially, they
would have increased as a fraction of total imports while imports of
intermediates used in their domestic production would have decreased
as a share of the total. Instead, intermediate goods, and capital goods to
a lesser extent, have increased relative to consumer goods in the total
(table 4-4).
Individual countries show considerable variation in changes in imped-
iments to imports. Chile, Korea, Mexico, the Philippines, and Turkey are
among the countries that undertook import reform. Chile's import lib-
eralization, which began in 1975, has been the most extensive in recent
time. Quantitative restrictions were rapidly replaced by uniform tariff
rates of 10 to 15 percent by mid-1979. Commercial policy reversals were
corrected and coupled with a substantial devaluation during 1983-87,
when the export/GDP ratio nearly doubled. In Mexico a major reduction
in import restrictions in the mid-1980s substantially reduced antiexport
bias and significantly increased exports. Korea is an example of sustained
liberalization and export development over a long period. Turkey carried
Table 4-4. Composition of Nonfuel Imports at Current Prices in
Countries with Trade Adjustment Loans 1980-87
(percentage share of total nonfuel imports)
Component         1980    1982    1983    1984    1985    1986    1987
Consumer goods    22.4    20.4    20.2    19.7    18.0    17.9    16.5
Capital goods     31.0    32.6    33.7    32.7    33.1    32.2    32.2
Intermediate
goods            46.6    47.0    46.1    47.6    48.9    49.9    51.3
Total nonfuel
imports         100.0   100.0   100.0   100.0   100.0   100.0   100.0
Average value
(millions of
U.S. dollars)   4,260   3,871   3,492   3,568   3,517   3,818   4,379
Note: Data are averages for the thirty-seven countries for which data were available.
Source: World Bank data.



58 ISSUES OF POLICY REFORM
out a major trade reform in the first half of the 1980s, transforming the
economy from its inward orientation to a more outward-looking one and
nearly tripling its export/GDP ratio during 1980-87. The Philippines
began with a tariff reform in the early 1980s, followed by substantial
reductions in quantitative restrictions in the mid-1980s.
Milder reform and even reversals of reform occurred in some cases.
Colombia, whose trade regime has been characterized by stability over
the past thirty years, undertook some export promotion along with mod-
est import reform in the 1980s. Kenya and Pakistan, among many others,
undertook only mild reforms, although their existing trade regimes were
quite restrictive. In other cases-such as Yugoslavia and Zambia-policy
improvements have been followed by abandonment or reversals of re-
forms. Estimates on the effects of incentives are available in a few cases
(Colombia, Kenya, Korea, Mexico, Morocco, Pakistan, the Philippines).
The estimates, however, are not comparable across countries. In general,
the effect on antiexport bias has varied, ranging from very significant
reduction in Mexico to little change in Pakistan.
Relative Performance Before and After Lending
Changes in performance indicators for recipients of trade adjustment
loans compared with nonrecipients (the control group) are presented in
table 4-5. To allow some time after the first loan to a country, only the
twenty-six countries that received a trade adjustment loan before 1986
are considered. Average changes in indicators for trade loan recipients
relative to nonrecipients are shown for 1985-87 compared with 1981-
83, as well as average changes for the recipients for the three-year period
following the first loan (excluding the year of the loan) compared with
the three years before the loan relative to the changes over the same
periods for nonrecipient comparators (see Balassa 1988). The numbers
show how many recipient countries in each classification performed bet-
ter on each indicator than their comparators after the start of trade ad-
justment lending. The plus and minus signs indicate an improvement or
a worsening of the average value of an indicator for the trade adjustment
loan group in comparison with the average value of the same indicator
for the comparator group."0
There are important limitations to this type of comparison (see Khan
1988 for a discussion)."1 Most important, perhaps, is that the countries
with adjustment loans are not necessarily selected randomly. Many fac-
tors affect performance other than the presence of this type of lending
and reforms. Subject to these and other caveats, the trade indicators show
that, on average, the change in performance between 1981-83 and 1985-
87 was better for the loan recipients than for the forty-seven nonrecipients.
The last three rows in the top half of the table show the total percentage



Trade Policy Reform 59
of cases across all nine indicators in which those countries with trade
adjustment loans (in three different classifications) did better than the
others. The relative performance of the twenty-six countries three years
after as opposed to three years before the loan is considered in the lower
half of the table.
The relative improvements in the trade adjustment countries are most
apparent in the indicators of trade and growth. These improvements are
probably attributable to the additional financing provided by the loans,
exchange rate adjustments, and improvement in the incentive regime.
There was a relative worsening, however, of some debt indicators as well
as investment ratios. This is not altogether surprising because these coun-
tries borrowed more heavily than the others, but since they also made
major adjustments, the weak improvements in the debt indicators may
have ramifications for the sustainability of the improvements.
Difference in Export Performance
Exports of developing countries expanded rapidly in the 1970s but
growth decelerated in the 1980s to half that of the 1970s. Although
overall export growth was weaker in the 1980s (3.6 percent average
annual rate) than in the 1970s (6.8 percent), performance was stronger
for recipient countries (4.5 percent) than for nonrecipients (2.8 percent)
in the 1980s. About one-third of the trade loan recipients also managed
to increase their exports in recent years (1982-88) compared with the
longer-term trend (1965-8 1, figure 4-2). Among the ten intensive recip-
ients (those that received three or more trade loans), more than half
increased their shares in total exports to industrial countries from nonoil-
exporting developing countries.
Manufacturing exports from developing countries grew at an average
annual rate of 7.6 percent during 1982-87, while the rate of increase
for all exports was only 3.1 percent. Particularly rapid rates of increase
in manufacturing exports during 1982-87 were recorded by Turkey (29
percent a year), Mauritius and Mexico (25 percent), Thailand (20 per-
cent), Korea and Zambia (around 15 percent), and Ghana and Morocco
(10 percent). Growth in manufacturing exports was stronger in the coun-
tries with trade loans than in the other countries, even when calculated
using unweighted averages (table 4-6). The annual growth rate for 1982-
87 was 9 percent for the loan recipient group compared with 6 percent
for the nonrecipient group. Although exports of primary products and
services have been very important for some countries, over time the main
contribution to export performance at the margin probably came from
manufactured exports.
To improve performance in this area, a realistic exchange rate policy
that yields competitive production costs, given the productivity of labor



Table 4-S. Indicators of Performance before and after Trade Adjustment Lending: Twenty-Six Recipients of Loans before
1986 and Forty-Eight Nonrecipients
Low- and          Sub-                           Exporters
Low-         Middle-         middle-        Saharan         Highly               of
Indicator                                  income        income          income          Africa        indebted        manufactures
Number of trade loan recipients             9             17             26              11              10                 7
Number of nonrecipients                    20             27             47              18               4                 8
1985-87 compared with 1981-83
GDP growth                                  9+*           12 +*          21              10+b             3_                5 +
Investment/GDP                              5 +           14 +           19               9 +             6-                7 +
Real exchange rate                          8+t           15 + s         23               9+t             8 +               7 +
Manufacturing exports growth                7+*           12+             19             10+*             1-*               4-
Import growth                               8+            12 + *         20               8 +             4-                6 +
Resource balance/GDP                        2 -           12 +           16               5 +            10+b               1-
Inflation                                   8+t           14 +           22              10+r             7                 I -
External debt/exports                       6+            17+t           23               8+             10+*               6+
Debt service/exports                        S +           10 +           15               4 -             3-                7 +
Share showing improvementa                 0.74           0.77            0.76            0.74           0.58              0.70
10 intensive recipients                    0.83           0.72            0.74           0.78            0.53              0.56
All 40 recipients                          0.63           0.74            0.71            0.68           0.58              0.56



Three years after compared with three years before
GDP growth                                    5 +           13 +            18               6+              5 +                4+
InvestmentlGDP                                4-            11 +            15               5-              8 +                4-
Real exchange rate                            7+            16+             23              10+t             9+                 7+
Manufacturing exports growth                  7+            14+             21               9 +             5-                 4+
Import growth                                 6 +           14 + t          20               6 +             5+*                7+t
Resource balance/GDP                          5+            11+             16               8+              8+                 2+
Inflation                                     7 +           13 +            20               9 +             6 +                4 -
External debt/exports                         5+            14+             19               7-              9+                 5+
Debt service/exports                          5-             9 +            14               3 -             5 +                4 +
Share showing improvementa                   0.63           0.75            0.71            0.64             0.67              0.65
10 intensive recipients                      0.78           0.69            0.71            0.71            0.71               0.78
All 40 recipients                            0.58           0.70            0.64            0.62             0.63              0.54
The change in means for the recipients between the two periods relative to the change for nonrecipients is significant at a 5 percent confidence interval.
t The change in means for the recipients between the two periods relative to the change for nonrecipients is significant at a 10 percent confidence interval.
Note: The numbers in the table show for each indicator the number of recipients in each classification that improved in the period after the trade adjustment loan
compared with the period before the loan relative to the change over the same periods for nonrecipient comparators. The year of receipt of the first loan is excluded from
the comparison in the lower half of the table. The plus and minus signs indicate an improvement or a worsening of the average value of an indicator for recipients compared
with the change in average value for nonrecipients.
a. The share of the product of the number of variables and the number of countries showing improvement in the total.



Table 4-6. Average Annual Percentage Growth Rates of Export Volume and GDP for Selected Country Groups, 1980-88
(unweighted averages)
Percentage change
Indicator and country                                                                                           1984-86/     1985-871
group                    1980     1981      1982      1983      1984     1985     1986      1987     1988a      1980-82       1981-83
Merchandise exports
87 developing countries   5.4      2.4       2.6       0.1       6.6      5.2      3.7       6.0       4.4        49.0*         19.2t
10 intensive trade loan
recipients             12.1      7.5        7.8     -4.9      11.5      3.5       9.2      7.5       3.5       - 11.7t        94.2
26 trade loan recipients  9.1      4.8        2.2     -2.9       7.3      3.6      8.3       7.8       5.1         19.3t       380.Sf
40 trade loan recipients  7.6      4.7      -0.4      -1.2       6.8      5.0      6.8       5.7       4.1         56.3t       464.5t
47 nonrecipients          3.5      0.3        5.1       1.2      6.5      5.4       1.0      6.3       4.8         44.9         92.4
Manufacturing exportsb
87 developing countries  18.4      9.7        1.2     11.2       9.8     10.6      7.2       5.5      10.7        _5.8*          5.4t
10 intensive trade loan
recipients             26.8     20.2      -3.8       15.6     11.9      9.5      10.2     13.7      17.2       -26.9*          4.5t
26 trade loan recipients    18.7   7.3       0.9       6.7       9.0     11.5      5.0      12.7      12.2        -5.2*         92.6t
40 trade loan recipients    25.6   6.5        0.6      10.3      7.4     14.1      11.6      9.9      13.7          1.2t       104.9t
47nonrecipients          11.9     12.6        1.7      12.1     11.8      7.4      3.5       1.7       7.9       -13.4        -52.2



GDP
87 developing countries      3.6        3.4        1.8        1.2       2.4        3.1        3.2       2.3        3.3          -0.9t          34.8t
10 intensive trade loan
recipients                 0.2        2.4        1.1        0.2       2.1        2.7       3.8        4.2        4.1         128.8t         188.7t
26 trade loan recipients     3.4        2.2        0.7        0.6       2.7        3.3       4.2        3.7        3.8          60.2t         214.1t
40 trade loan recipients     2.7        2.8        0.3        0.4       2.2        3.5       3.9        3.2        3.6          63.7t          198.2t
47 nonrecipients             4.4        3.8        3.1        1.9       2.6        2.7        2.6       1.6        3.0        -29.2           -20.5
* Differences in means between the recipients and nonrecipients of trade adjustment loans were significant at the 5 percent confidence interval.
t Differences in means between the recipients and nonrecipients of trade adjustment loans were significant at the 1 percent confidence interval.
a. Preliminary estimates.
b. The definition of manufactures is from the Foreign Trade Statistics, International Economics Department, World Bank; it includes line items of S + 6 + 7 + 8 -68 in
the Standard International Trade Classification (sITc).
Source: World Bank data.
(ON



64 ISSUES OF POLICY REFORM
Figure 4-2. Export Growth for Forty Countries with Trade
Adjustment Loans
(percent)
1965-81 30
Korea
20-
Togo       :Brazil           Tke
10-        Too.         .T</ rurkey
*Thailand
* .   . .. . *.  * . Mexico
Tanzania            * . Pakistan
Mauritius
0              -
Ghana
-10*                           I             I
-10            0            10           20           30
1982-88
Source: World Bank data.
in each country, will continue to be essential, as will measures to enhance
productivity. Exporters also need access to a growing range of domest-
ically produced inputs at world prices and of world quality. Institutional
and marketing support to export activities are also especially strong
among successful exporters, while the opposite is true among the poor
performers.
Some countries have introduced export incentives while maintaining
protection in the import regime. But unless protection is reduced, the
incentive to shift resources from production for a captive domestic market
to a tough and competitive international market is likely to remain lim-
ited. And as long as import controls remain in place, there will be a need
to introduce such things as duty drawback schemes, export processing
zones, and bonded warehouses to ensure that exporters receive special
access to imported inputs at world prices. In addition, the successful East
Asian exporters have paid attention to easy access to foreign exchange
for exporters, preshipment credit for working capital, labor costs, train-
ing, education, infrastructure, technology, and marketing (for a discus-



Trade Policy Reform 65
Figure 4-3. Change in Growth Rate of GDP and Imports for
Countries with Trade Adjustment Loans
Imports
30 |                            Turkey
*   * .Chile
20 -                   Tunisia.
. . Brazil
1  *Colombia
Mexico .  '
Panama.          . Senegal
-10-                        *     . Burundi
CUte d'lvoire.
-20 -
Kenya
-30
-30      -20      -10        0        10       20       30
GDP
Note: The figure shows the percentage point improvement in the growth rate of GDP
and imports three years after the first trade loan compared with three years before.
Source: World Bank data.
sion, see Keesing 1988). Their exports have also benefited from a fa-
vorable regulatory environment, support for enterprise development, and
a forward-looking industrial policy. Furthermore, the East Asian coun-
tries have dealt successfully with protection abroad to maintain their
prospects for market penetration.
Does Policy Matter?
In the first half of the 1980s, import compression was the dominant force
behind a negative relation between changes in the trade balance and
changes in GDP. Figure 4-3 shows the strong positive link between import
growth and GDP growth. Some observers argue on this basis that the
changes in the availability of external financing (affecting imports) were
a powerful explanation of changes in GDP in this period. To be sure,
starting from a situation in which domestic savings are not easily con-
verted into foreign exchange to purchase imports, greater external fi-



66 ISSUES OF POLICY REFORM
nancing and the relaxation of import controls, by permitting increased
importation of raw materials and intermediate and capital goods, can
contribute to higher GDP. Adjustment lending has facilitated additional
financing for imports. It has also supported import liberalization, which
together with real exchange rate adjustments was expected to boost GDP
more than would the increased importation of inputs alone.
The possible effect of changes in commercial policy and in the real
exchange rate is considered below in a framework that incorporates the
effect of external financing and the effect of changes in the terms of trade.
The short-term impact of increased financing on the growth of GDP iS
expected to be positive. An improvement in terms of trade is also expected
to increase GDP growth. A depreciation of the real exchange rate would
have a positive effect on GDP growth if the increase in tradable production
more than offsets the decline in nontradable production. Commercial
policy reform that reduces antiexport bias would increase GDP growth if
the positive effect on exportables is larger than the negative effect on
importables.
To illustrate these propositions, we consider changes between a period
before reforms and after reforms as below:
(4-1)  GDPi = Po + P1 Fi + P2 TOTi +  3 RERRi + [4 Pi + Ei
where
Po = constant
1, 12, 03, 04 = coefficients of the respective variables
F = financing
TOT = terms of trade
RER = real exchange rate
P = index for commercial policy reform
e= the error term, assumed to be uncorrelated with the
independent variables, and with zero mean and con-
stant variance
i= each of the countries
*= rate of change.
The four variables on the right-hand side are assumed to be related to
GDP in a log linear fashion. Estimations were made for the forty countries
with trade adjustment loans as opposed to forty-seven nonrecipients, with
a dummy variable indicating the presence of a loan. Estimations were
also made for a group of thirty-eight reform countries that excluded the
four trade loan recipients with no progress or a reversal in commercial
policy (table 4-2) plus Bolivia and Haiti (which carried out trade reforms
without adjustment loans) as opposed to the remaining forty-nine non-
reformers; a dummy variable indicated the presence of reforms. Finally,
for the twenty-four countries with detailed implementation data, a vari-



Trade Policy Reform 67
able was used to indicate the extent of policy change. This variable was,
in turn, an index of policy change-obtained as the product of the indices
of reform proposals-ranked 3, 2, 1-and the degree of fulfillment of
what was proposed-ranked 80, 60, 40, 20, and 0 percent (based on
Halevi 1989).
Table 4-7 provides the results of ordinary least squares estimations for
the eighty-eight developing countries. Changes are considered during
three years after the loan compared with three years before. Changes in
financing and terms of trade generally have the expected effects. (Alter-
natively, changes in imports-as a proxy for financing, not shown in
table 4-7-are found to be strongly and positively associated with
changes in GDP growth.) The contribution to GDP growth of policy reform
is positive and strong when reformers are compared with nonreformers,
and when the extent of policy implementation is accounted for. The
coefficient of P in the analysis of the twenty-four countries with data on
implementation is extremely robust. The impact of real exchange rate
depreciation (defined as a decrease) is positive and significant. In the
absence of either RER or P, the other variable assumes much greater
significance (not shown), presumably because of the high correlation
between the two variables.
Conclusion
Overall, implementation of trade policy reform has been moderately sig-
nificant in the sample of developing countries. Reforms in exchange rate
policy and a reduction in impediments to export have occurred, as well
as a reduction in impediments to the import of inputs needed by ex-
porters. Although quantitative restrictions have been replaced by tariffs
in many countries, success in lowering quantitative restrictions has been
more modest in the face of foreign exchange constraints, except in se-
lected cases (for instance, Chile, Korea, Mauritius, and Mexico). In some
cases there has been a reduction in effective protection for importables
and in antiexport bias (for example, Mexico, Morocco, and Philippines).
Domestic reforms, however, have lagged even in some of the major trade
reformers (for example, Mexico), and institutionalization of reforms and
reductions in protection levels have been limited.
Given the strong emphasis on trade policy under adjustment lending,
one might expect greater reforms of the trade regimes than actually oc-
curred during this period. In particular, four factors have constrained
reform: macroeconomic instability, inadequate conviction concerning the
benefits of reform and vested interests against reform, weak capacity for
implementation, and conflicts in design. Institutional reform has been
found to be particularly slow, while price reforms have not always been
sustained. These issues are important because sustainable price changes



Table 4-7. GDP Growth, External Factors, and Policy Reform
Number of
Equation                            Constant          F           T6T          RER           P                     F statistic     observations
Three years after compared with three years beforea
Dummy: 1 = 40 recipients                0.33        0.002         0.08        -0.05         0.32        0.18           3.3              66
0 = 47 nonrecipients          (0.42)       (0.32)       (2.55)       (-2.54)       (0.29)
°°    Dummy: 1 = 38 reformersb              -0.29         0.002         0.09        -0.04          1.78       0.22           4.2              66
0 = 49 nonreformers'        (-0.40)        (0.33)        (2.85)      (-2.43)       (1.67)
Reform: Index of policy change        -3.09         0.02          0.09        -0.08         3.38        0.60           7.0              24
(-2.09)       (1.70)        (1.92)      (-2.29)       (2.58)
Note: t statistics are in parentheses. The dependent variable is GOP. Changes and changes in growth between the periods give similar results.
a. The reference year 1984 is for nonrecipients.
b. Excluding Guyana, Yugoslavia, Zambia, and Zimbahwe from the group of forty trade loan recipients, but including the nonrecipient reformers Bolivia and Haiti.
c. Excluding Bolivia and Haiti from the group of forty-seven nonrecipients but including the four countries listed in footnote b.



Trade Policy Reform 69
and effective institutional support are vital to achieving meaningful sup-
ply responses.
On nine performance indicators, the strongest improvement of trade
loan recipients over nonrecipients in the postloan period relative to the
preloan period was in the growth in manufacturing exports and in im-
ports. Less progress was made with respect to debt indicators. For the
short period under review, the overall positive evidence is modest. In
general, the more significant improvements concern middle-income coun-
tries. The evidence is also more favorable when early loan recipients are
considered (especially when only the ten intensive trade loan recipients
are considered) than when all forty recipients are included in the com-
parisons.
Regression analysis found the growth in output to be affected by prox-
ies for external factors. At the same time, policy reform was found to
have a positive impact on growth performance (see also Edwards 1989).
The evidence of this policy impact is mild when all trade loan recipients
are compared with nonrecipients but stronger when reformers are com-
pared with nonreformers. Country studies corroborate this finding as
well. The evidence supports the need for continued and stronger efforts
to reform trade regimes.
Notes
I gratefully acknowledge the collaborative assistance of Bela Balassa, Mansoor
Dailami, Jaime de Melo, Sebastian Edwards, Nadav Halevi, Ramon Lopez, Fran-
cis Ng, and Julie Stanton. The work benefited from comments from Ajay Chhib-
ber, Patrick Conway, Faezeh Foroutan, Gene Grossman, Arnold Harberger, Ann
Harrison, Thomas Hutcheson, Kazi Matin, John Nash, Arvind Panagariya, Dani
Rodrik, James Tybout, John Williamson, Alan Winters, and Alexander Yeats.
Rebecca Sugui prepared the paper.
1. The analysis in this chapter and several others is based on a review of eighty-
one adjustment loans with a significant component for trade reform that were
approved between 1979 and 1987. Of these eighty-one loans, forty-seven were
structural adjustment loans (SALS), thirty-two were sectoral adjustment loans (SE-
CALS), and two were program loans. Many included technical assistance com-
ponents or were accompanied by technical assistance loans in support of trade
reforms. Of the forty countries that received loans, detailed implementation data
were available for only twenty-four in which sufficient time had elapsed since
their first trade adjustment loans. Most of these twenty-four countries had re-
ceived a trade loan before 1986, although a slightly different group of twenty-
six countries constitutes the pre-1986 recipients. Among the forty countries, ten
"intensive" adjusters received three or more trade adjustment loans.
Eighty-eight developing countries constitute the full sample. These are the
ninety-five countries defined by the World Development Report 1989 as low-
income and middle-income countries, minus ten countries (Afghanistan, Bhutan,
Iran, Iraq, Kampuchea, Lao People's Democratic Republic, Lebanon, Libya, Ro-
mania, Viet Nam) for which data were not available, plus three (Gambia, Guinea-
Bissau, and Guyana) that had received adjustment loans but were not already



70 ISSUES OF POLICY REFORM
included in the World Development Report definition because of their small
populations. Forty-eight countries in the sample were middle-income countries
and forty were low-income countries (GNP per capita of $480 or less in 1987).
2. Export restrictions have included prohibitions based on economic or safety
grounds, restrictive licensing, export quotas, export taxes, and limitations on the
retention of foreign exchange.
3. Quantitative restrictions have included import prohibitions, quotas, and
restrictive licensing of various sorts. Other restrictions include licensing and con-
trols for foreign exchange, requirements for advance deposits for imports, and
restricted import channels (as in the case of a state trading monopoly).
4. In addition to customs duties, tariffs include customs surcharges, surtaxes,
stamp taxes, and taxes on foreign exchange.
5. The judgments were based on recommendations for loans, country mem-
oranda, country briefs, audit reports, mission reports, background work for
World Bank (1988), IMF reports, and the Ford Foundation project on trade policy
and the developing world. Sufficiently large differences were detected to permit
such a broad classification.
6. When "secondary trade restrictive intent" is included, the figure rises to
27.2 percent; estimates for imports "affected," rather than covered, are higher
still-48 percent.
7. These proposals are grounded in the conceptual and empirical work of
many trade policy analysts; see, for example, Balassa (1988); Bhagwati (1978);
Corden (1974); Krueger (1978); and Little, Scitovsky, and Scott (1970).
8. Although the achievement of a real depreciation clearly depended on ma-
croeconomic adjustments (fiscal, monetary, and wage), this chapter focuses only
on exchange rate policy identified under trade adjustment lending.
9. When quantitative restrictions are binding before and after the depreciation,
however, the depreciation increases the price of exportables relative to import-
ables.
10. For most indicators a positive change is an improvement. For resource
balance/GDP, external debt/exports, and debt service/exports, a positive change
is a worsening and is indicated in the table by a minus. For the real exchange
rate, a greater real depreciation for recipients between periods than that for com-
parators is an improvement.
11. In particular, sample selectivity bias is likely to come into play here in
that changes in performance attribute-d to the receipt of a loan may reflect con-
ditions that systematically led to the receipt of the loan.
References
Balassa, Bela. 1988. "Quantitative Appraisal of Adjustment Lending." PPR Work-
ing Paper 79. Policy, Planning, and Research, World Bank, Washington, D.C.
Bhagwati, Jadgish. 1978. Foreign Trade Regimes and Economic Development.
Cambridge, Mass.: National Bureau of Economic Research.
Corden, W. M. 1974. Trade Policy and Economic Welfare. Oxford: Clarendon
Press.
Edwards, Sebastian. 1989. "Openness, Outward Orientation, Trade Liberali-
zation, and Economic Performance in Developing Countries." Trade Policy
Division, Country Economics Department, World Bank, Washington, D.C.



Trade Policy Reform 71
Erzan, R., H. Kuwahara, S. Marchese, and R. Vossenaar. 1986. "The Profile of
Protection in Developing Countries." UNCTAD Discussion Paper 21. Geneva:
U.N. Conference on Trade and Development.
Finger, Joseph M., and Samuel Laird. 1987. "Protection in Developed and De-
veloping Countries-An Overview." Journal of World Trade Law 2 (6) De-
cember.
GAIT. 1980. "The Tokyo Round of Multilateral Trade Negotiations-Il, Sup-
plementary Report." Geneva: General Agreement on Tariffs and Trade.
Halevi, Nadav. 1989. "Trade Liberalization in Adjustment Lending." Trade Pol-
icy Division, Country Economics Department, World Bank, Washington, D.C.
Keesing, Donald B. 1988. "Adjustment and Export Development." Trade Policy
Division, Country Economics Department, World Bank, Washington, D.C.
Khan, Mohsin S. 1988. "The Macroeconomic Effects of Fund-Supported Ad-
justment Programs: An Empirical Assessment." International Monetary Fund,
Washington, D.C.
Krueger, A. 0. 1978. Foreign Trade Regimes and Economic Development: Lib-
eralization Attempts and Consequences. Cambridge, Mass.: Ballinger.
Laird, Samuel, and Alexander Yeats. 1988. "Trends in Nontariff Barriers of
Developed Countries, 1966-86." PPR Working Paper 137. Policy, Planning,
and Research, World Bank, Washington, D.C.
Little, I. M. D., Tibor Scitovsky, and M. F. Scott. 1970. Industry and Trade in
Some Developing Countries: A Comparative Study. New York and London:
Oxford University Press.
Papageorgiou, Demetrios, Michael Michaely, and Armeane Choksi, eds. Forth-
coming. Liberalizing Foreign Trade. Vol. 7, Lessons of Experience in the De-
veloping World. Oxford, U.K.: Blackwell.
Rodrik, Dani. 1988. "Liberalization, Sustainability, and the Design of Structural
Adjustment Programs." Trade Policy Division, Country Economics Depart-
ment, World Bank, Washington, D.C.
Williamson, John. 1985. The Exchange Rate System. Rev. ed. Washington, D.C.:
Institute for International Economics.
World Bank. 1988. Adjustment Lending: An Evaluation of Ten Years of Ex-
perience. Policy and Research Series 1. Washington, D.C.
. 1989. World Development Report 1989. New York: Oxford University
Press.



72 ISSUES OF POLICY REFORM
Comments
John Williamson
MY PRIMARY DUTY here is presumably to offer a verdict on whether the
World Bank has been pursuing the right objectives in the trade conditions
attached to adjustment loans and whether it has been achieving those
objectives. Accordingly, I shall discuss first the objectives pursued by the
World Bank and subsequently Thomas's attempt in this chapter to mea-
sure success.
The World Bank has sought through its trade adjustment loans to
persuade developing countries to open up their economies and exploit
rather than fight their comparative advantage. Had adjustment lending
started twenty rather than ten years ago, I doubt whether opening up to
the world economy would have been a central focus. The profession was
converted by Little, Scitovsky, and Scott (1970), followed by the metic-
ulous assembly of empirical evidence at the World Bank under the in-
tellectual leadership of Bela Balassa, and reinforced by the dramatic and
visible success of outwardly oriented policies in the East Asian newly
industrialized countries. The benefits of outward orientation are now a
staple component of Washington conventional wisdom, a belief that I
share. Chapter 4 provides some modest additional support for this po-
sition, especially in the positive coefficient on the policy variable, al-
though I doubt if it will convert many of the skeptics in the World Bank's
client countries.'
A key ingredient of an outward-orientation policy is a "competitive"
or "realistic" exchange rate. Figure 4-1 shows conventional International
Monetary Fund (IMF) measures of real effective exchange rates as the
relative prices of different nations' output, and this measure is used to
generate the estimates shown in table 4-7. Thomas takes for granted that
depreciation of the real exchange rate in that sense will increase the
relative price of tradables in terms of nontradables, but this is a hypothesis
that deserves testing. The empirical results would have been more inter-
esting had they been based on the relative price of nontradables, which
measures the supply-side incentive to produce exportables and so is a
more interesting concept-although one that is more difficult to mea-
sure-than the real exchange rate in most developing countries.



Trade Policy Reform 73
My major concern with the treatment of exchange rates in this chapter
is that exchange rate policy is judged by how much real depreciation was
achieved rather than by whether misalignments were avoided. The prob-
lem is acknowledged, but the procedure is nonetheless misleading in prin-
ciple: for example, it means that encouraging Korea to further devalue
an already undervalued won is counted as a success. Any adequate eval-
uation of the World Bank's success in this field requires judgments on
"sproper" levels of exchange rates analogous to my fundamental equilib-
rium exchange rates (Williamson 1985). In the context of a developing
country, however, one might want to modify the definition to refer to a
rate that would be expected to limit current imbalances to a size that
could be sustainably financed while promoting export growth adequate
to buy the imports needed to exploit the economy's full growth potential.
I have the impression that, when the World Bank and the IMF have
differed over their advice on exchange rates, the reason is that the World
Bank has been basing its advice on a concept such as this while the IMF
has employed the narrower criterion of financial viability. The World
Bank was right to use the broader criterion and should not allow charges
that the concept contains an element of subjectivity to prevent it from
making its use explicit. The failure to measure the extent of overvaluation
means, for example, that Thomas cannot emphasize the harm being done
to C6te d'Ivoire and Senegal by the obstinate insistence on treating the
fixed parity of the CFA franc as sacrosanct.
The other key element of an open economy is a commercial policy that
roughly equalizes the domestic resource cost of earning or saving a unit
of foreign exchange in all the various ways possible. Balassa's (1975)
concept of outward orientation refers to the similarity of the domestic
resource costs of exporting and of import substitution. The dispersion
of effective protection rates measures variations in the domestic resource
cost of saving foreign exchange by import substitution, and an analogous
measure could obviously be constructed on the export side.
Like the fundamental equilibrium exchange rate, the dispersion of do-
mestic resource costs is a sophisticated concept for which accurate mea-
sures are not available. I nevertheless hold the (quixotic?) view that a
crude measure of a relevant concept is preferable to an accurate measure
of a fundamentally irrelevant concept. It is rather disappointing to dis-
cover that the vast amount of empirical study undertaken in the World
Bank could not be exploited to yield some rough estimates of how much
progress in equalizing domestic resource costs has been accomplished by
trade adjustment loans.
Reforms of most of the measures listed under "overall import policy"
in table 4-1 seem likely to have helped equalize domestic resource costs.
I have two reservations, however, about World Bank advice on com-
mercial policy.



74 ISSUES OF POLICY REFORM
The first concerns the advice to reduce or preferably eliminate taxation
of agricultural export crops. Every economist knows that, for a country
with monopoly power in world trade, equalization of domestic resource
costs would require an "optimum tariff," but the relevance of this prin-
ciple for developing countries is usually denied on the grounds that they
do not possess any significant monopoly power. Although that is true
for individual countries, it is not true for developing countries collectively
with respect to a few important products produced only in the tropics
(most notably beverages). A common export tax on those commodities,
by enabling low-income countries to redistribute income toward them-
selves at the expense of the rich, seems to me worthy of encouragement
rather than the reverse (Williamson 1987).
The second reservation concerns the timing of import liberalization.
As Thomas admits in passing, the liberalization of competitive imports
can have a negative effect on domestic output, which can be important
when the mobility of resources between sectors is low, as is often the
case in the short run. This implies a low domestic resource cost from
continuing to restrict imports until new investment is needed. Even if one
cannot hope to synchronize tariff reductions with investment needs in
any sophisticated way, a policy of announcing a tariff reduction well in
advance (to discourage new investment in inefficient import-substituting
industries) and then timing it to coincide with an easing of the payments
position may be preferable to the rigid timetable of import liberalization
seemingly urged (if not always enforced) by the World Bank. The former
strategy is, after all, that adopted by Europe during its postwar recovery.
The measures of the success of the World Bank's trade adjustment
loans are presented in tables 4-5 to 4-7. Table 4-5 does not present par-
ticularly persuasive evidence, partly because the margin of superior per-
formance on the part of the countries receiving trade adjustment loans
is less than overwhelming and partly because some of the superiority
could be due to the import growth permitted by increased finance rather
than to the policy changes that conditioned the loans. Table 4-7 provides
some assurance that import growth from increased finance is not the
entire story, since the policy variable is significant (and implies increased
growth of as much as 2 percent a year) even after allowing separately
for import growth.2 And table 4-6 indicates that export growth has been
stronger, to a marginally significant extent, in the countries with trade
loans. Thus the evidence justifies a verdict of at least limited success.
Nevertheless, I wonder whether a better research strategy might not
be possible. Rather than ask whether countries with trade loans had better
performance, one could break the question down into two. The first
would ask whether smaller exchange rate misalignments and lower do-
mestic resource cost dispersion achieved superior performance, which
would provide a direct test of whether those are indeed the policies the



Trade Policy Reform 75
World Bank should have been promoting. The second would then ask
whether trade loans succeeded in reducing misalignments and the dis-
persion of domestic resource costs, which would provide a direct test of
whether the World Bank had succeeded in promoting those policies. If
both questions were answered in the affirmative, one would have a much
stronger basis than that provided in the chapter for endorsing the World
Bank's trade adjustment loan policy. This alternative approach would
require at least crude measures of exchange rate misalignments and the
dispersion of domestic resource costs, an admittedly difficult task. But I
believe it would be worthwhile to make the effort.
Notes
1. A recent example that came across my desk argued as follows: "something
should be done urgently to further protect our local industries. The area of policy
focus to be addressed now is a discriminatory export policy with the sole aim
of conserving the country's food and raw materials. If protectionism through
import duty variations is informed by the need to protect local industries, then
the government should not hesitate to balance the present tariff structure with
the introduction of export quotas for food and raw materials.... Export quota
is the practice in most market-oriented economies . . ." (Monthly Business and
Economic Digest, United Bank for Africa, Lagos, January 1989, p. 1).
2. Average import propensities are typically around 0.2, so a finding that 5
percent import growth buys 1 percent output growth implies that a dollar of
imports buys only a dollar of extra output (and hence $2 of extra supply of
goods). It is perhaps surprising that extra imports buy as little growth as this,
given that most of these countries were foreign-exchange constrained over the
estimation period.
References
Balassa, Bela. 1975. "Reforming the System of Incentives in Developing Coun-
tries." World Development.
Little, I. M. D., Tibor Scitovsky, and M. F. Scott. 1970. Industry and Trade in
Some Developing Countries: A Comparative Study. New York and London:
Oxford University Press.
Williamson, John. 1985. The Exchange Rate System. Rev. ed. Washington, D.C.:
Institute for International Economics.
-. 1987. "Exchange Rate Policy for Developing Countries." Journal of
Foreign Exchange and International Finance 1 (1) January:39-46.



S Financial Sector Reform
Alan Gelb
Patrick Honohan
THE RECENT UPSURGE of concern with financial policy in developing coun-
tries arises primarily from three characteristics of their financial systems.
Many financial institutions are extremely unsound; control over interest
rates and the direction of credit is often excessive and amounts to repres-
sion of the financial systems; and deposit money banks dominate the
financial systems while the nonbank financial sector has been neglected.
To help developing countries address these problems, the World Bank
introduced financial sector adjustment loans. By mid-1988, four such
loans had been made: one each to Argentina (1988) and Ecuador (1988)
and two to Turkey (1986 and 1988), with an additional trade and fi-
nancial sector adjustment loan to Jamaica (1987; all fiscal years). Fi-
nancial reform has also been a major component of a few structural
adjustment loans, notably to Chile and to the Philippines. Changes in
financial policies have also been incorporated in the World Bank's fi-
nancial intermediation loans (which use financial institutions in borrow-
ing countries to on-lend World Bank funds to final borrowers); these
have a longer history, but the scope of policy reform has been more
limited. Financial sector loans are now in preparation for Ghana, Hun-
gary, Kenya, Nepal, and Nigeria, and others are being considered. The
empirical evidence on the conception, implementation, and effects of such
loans is still limited, however.
The Need for Reform
In perhaps the majority of developing countries, financial institutions
covering a large part of the market are technically insolvent. In a smaller
number of countries, many institutions have lost their capital several
times over, a phenomenon observed in both publicly owned and private
institutions, but especially in the former.
Soundness, Prudential Regulation, and Supervision
As with most bank failures in industrial countries, the poor quality of
bank management-including fraud and insider abuse-has been evident
76



Financial Sector Reform 77
in many failures in developing countries.1 Management's problems have
included undue concentration of risks in certain sectors or to certain
individuals, lending to uncreditworthy clients at the behest of directors
or the government, and close ownership links between banks and their
borrowing clients (including public banks and their public clients).
Insolvency has also resulted from factors external to the financial sys-
tem.2 Unforeseen macroeconomic shocks that upset the calculations on
which loans were based have left borrowers unable to service the loans.
Large devaluations have sometimes created difficulties, particularly
where many borrowers have been in import-related businesses or have
incurred dollar debts or when a bank has held a foreign exchange ex-
posure. Deflationary policies have generally resulted in a deterioration
of the quality of bank loan portfolios. Trade liberalization measures have
weakened the ability of those previously protected by trade barriers to
repay their loans. Thus bank failures may accompany some aspects of
desirable macroeconomic adjustment programs.
Although not all of these problems could have been forestalled by
adequate supervision and prudential regulation, including adequate ac-
counting and auditing standards, some could have been, and the cost of
dealing with the problem promptly would have been far lower. Fur-
thermore, a weak regulatory and supervisory framework is a barrier to
effective financial liberalization, because liberalization places different,
and in many respects heavier, demands on prudential regulation and
supervision than does a directed credit system (Dooley and Mathieson
1987; Diaz Alejandro 1985). The skills required of banking institutions
in liberalized financial markets are different from those needed where
interest rates and credit flows are heavily controlled.
In most cases of a bank's collapse, the state has ultimately indemnified
the losses to depositors because of the perceived importance of preserving
confidence and avoiding a flight from deposits. But the consequences of
bank insolvency go beyond the adverse impact on government budgets.
Distressed banks and borrowers can impede the access to credit of good
borrowers and sound investment projects. And, while bank difficulties
go unrecognized, management continues to lend to nonperforming bor-
rowers in the hope of some reversal of their fortunes. In a scramble for
liquidity to support this unsustainable behavior, banks may bid up in-
terest rates to levels well above the likely return on investments. New
lending projects are not entertained by banks in these circumstances, and
because of the liquidity squeeze not even worthwhile investment projects
can raise finance outside the banking system at a reasonable cost. The
repositioning of the economy's productive capacity to a new set of relative
prices or to an export-oriented policy is thus inhibited. Meanwhile, the
volume of "paper assets" and the government's potential liability mush-
room.3



78 ISSUES OF POLICY REFORM
Correcting the insolvency problems of the banking system is thus an
urgent task. Although the character of reform is more that of crisis man-
agement than of long-term development, the task must be accomplished
with a view to the longer run if problems are not to recur. Certain "in-
frastructural" components are needed to restructure a banking system.
For example, accounting and auditing standards must conform to gen-
erally accepted principles. A system for classifying loans by quality is
necessary to allow for adjustments for the inadequacy of capital and early
warning of problems. Adequate systems of prudential regulation and
supervision must be installed to ensure the health of the restructured
system. Implementing these reforms may require legal and institutional
changes in the relationship between banks and their regulators and be-
tween lenders and borrowers.
Financial Sector Repression
Excessive regulation of interest rates, extensive direction of credit, and
unduly heavy taxation of financial intermediation severely inhibit the
effectiveness of financial systems in many developing countries. Such
repressed financial systems may not be able to offer savers an asset that
can be expected to retain its real value over time. In that case, financial
intermediaries may be little used, and savers may rely instead on relatively
unproductive assets or capital flight.4
The disposition of the funds that are made available to financial in-
stitutions may also be constrained. Intermediaries may not be able to
expand credit to promising projects or to avoid lending to sectors and
projects in which the risk of failure outweighs the likely returns. In many
developing countries more than half of bank credit is directed. The pro-
ductivity of aggregate investment in an economy may thus be seriously
impaired, thereby wasting resources and slowing growth. By withholding
responsibility from the financial institutions, directed credit also tends
to weaken the process of loan recovery and can wreak havoc with the
financial viability of intermediaries.
From certain perspectives, central direction of credit allocation might
result in a good distribution of credit even though it does not take ad-
vantage of the informational gains offered by a decentralization of credit
decisions (see Cho 1986). In the absence of such controls, however, and
given a reasonable degree of macroeconomic stability, the financial sector
could be performing the important role of assessing the relative potential
of different projects. But because of the prevalence of credit direction,
skills in credit assessment may not be developed anywhere in the econ-
omy.
Some regulations may have been instituted in order to avoid the risks
of unduly vigorous competition or to ensure that favored borrowers were



Financial Sector Reform 79
adequately served. But it is now quite generally recognized that these
regulations have often been excessive and ultimately resulted in less
credit, much of it misdirected, and in a new set of prudential risks. These
regulations can prevent adequate diversification of bank portfolios and
severely limit the spreads that are obtainable.
In most cases, financial sector repression has a pronounced fiscal di-
mension as well. The financial system has traditionally been an admin-
istratively convenient source of tax revenue. A zero or low rate of re-
muneration on high reserve requirements (often more than one-fifth of
deposits and occasionally approaching two-thirds) has also contributed
to the fiscal burden on financial intermediation and to the widening of
the spread between borrowing and lending rates in many developing
countries. The total tax burden levied on financial systems in developing
countries is often as high as 2 to 3 percent of GDP and sometimes as high
as 7 or 8 percent.
The Nonbank Financial Sector
In most developing countries, the financial system is dominated by banks;
nonbank financial intermediation has been comparatively neglected. If
not too heavily regulated, banks can perform many of the most important
functions within a financial system: they can provide a safe and liquid
repository for savings, operate an efficient payments mechanism, and
provide the necessary financing for trade. When it comes to other finan-
cial services, however, such as the financing of long-term industrial de-
velopment (by debt or equity) or the provision of high-yielding long-term
savings media, the performance of banks has often been questioned.5 It
is widely believed that other types of institutions are needed to complete
the picture and ensure the maximum contribution of the financial sector
to economic development-as well as to stimulate the banks by com-
peting with them.
The World Bank's main focus in this area has been on the development
of capital and money markets. Capital markets give investors access to
long-term funds and equity. Long-term loans are suitable for financing
investments with long gestation periods and returns that will materialize
only over a long period. Equity capital is desirable for financing risky
investments; it allows investors to go through a bad period without fear
of bankruptcy. From the saver's point of view, equity and long-term
investments can offer a higher expected return and maturities matched
to the saver's needs. An effective capital market can thus result in greater
resource mobilization and more productive investment. But banks in
most countries are reluctant to commit themselves to long-term lending
or equity investment (or are forbidden by law to do so) since their own
liabilities are fixed and generally short term. There is also concern that



80 ISSUES OF POLICY REFORM
industrial equity in many developing countries is insufficient to provide
a robust basis for long-term growth.6
Money markets represent the short-term end of capital markets. Their
development allows large firms and institutions to satisfy their needs for
short-term borrowing and lending without the intermediation of banks,
thereby improving efficiency and cutting costs.
The emergence of a successful capital market requires not only rea-
sonable macroeconomic stability but also an adequate regulatory frame-
work that will prevent fraud and conflicts of interest and ensure adequate
disclosure. Without this framework, savers will distrust the market be-
cause they are individually unable to assess the probity and creditwor-
thiness of borrowers. The establishment of an adequate regulatory frame-
work is thus primarily an institution-building activity.
Financial Intermediation Loans
The World Bank's involvement in financial sectors has not been confined
to adjustment lending. Of quantitatively greater importance have been
operations involving the use of domestic financial institutions to on-lend
World Bank funds to final borrowers. The first loans of this type were
channeled through specialized development finance institutions (DFIS),
some of which were established specifically or primarily for this purpose,
often with World Bank involvement. More than 150 DFIS around the
world have been used by the World Bank for lending in industrial sectors
alone. Over time a larger share of loans has been channeled through the
commercial banking sector via such "apex" institutions as the central
bank.
This lending responded to a deficiency in the provision of long-term
finance and was seen as the only way to reach smaller, private borrowers
in the industrial and agricultural sectors. During the past few years in-
dustrial lending through financial intermediaries has averaged about
US$2 billion annually, and this type of lending is likely to increase.
Unfortunately, many of the loans made by the DFIS did not perform
satisfactorily, and by the start of the 1980s the parlous financial state of
many of them had become apparent. As many as one-third had severe
portfolio problems because of a variety of elements in the macro and
sectoral environments and poor lending procedures. A review in 1985
concluded that about half the DFIS were in reasonable condition but that
only about ten to fifteen were sufficiently profitable to turn to the market
for financing (World Bank 1985).
The expansion of World Bank lending activities to include a wider
range of intermediaries (notably the commercial banks) was an attempt
to reach a wider range of clients. But it also sidestepped the serious
portfolio problems in many of the development banks. The doubts about



Financial Sector Reform 81
the soundness of many banking systems raise questions about the allo-
cation of resources intermediated by the commercial banks as well as
about the ability of these banks to assess credit risk.
To the extent that World Bank lending for financial intermediation
has included conditions related to general financial policy, the conditions
have been confined largely to reforms to raise interest rates on subsidized
lines of credit. This focus has often reflected concern that other, heavily
subsidized lines of credit would undercut demand for World Bank-fi-
nanced credit lines and prevent the loans from being disbursed. As the
prevalence of repeat loans to intermediaries with poor loan recovery
records suggests, the solvency of intermediaries, their managerial capa-
bilities, and their credit processes have often received little attention,
although these aspects have now begun to be addressed more seriously.
The Content of Adjustment Policy Packages
Reform of a country's financial policy will often be directed toward re-
storing the solvency of the banks, freeing up the financial system, and
developing a capital market.
The Prudential Framework and Bank Restructuring
Three elements of policy are typically required to restore financial systems
to a solvent state. First are measures to determine the state of bank port-
folios and the necessary provisions and write-offs. These measures in-
clude the strengthening of accounting and auditing systems and skills,
implementation of external audits, and classification of the portfolio ac-
cording to loan quality. Legal changes to facilitate the intervention of
regulators may also be needed. Second are measures to implement pro-
visioning rules for less severely affected banks and to require weaker
banks to clean up their loan portfolios and install new management. Some
banks will need to recapitalize and reduce costs, merge (if a satisfactory
business plan can be developed), or be liquidated. Third are measures to
improve bank supervision and regulation for the system as a whole to
ensure timely provisioning against losses and provide information for
both bank management and regulators.
This third measure will usually require an overhaul of the accounting
and auditing rules and systems and of the rules for portfolio classification
and the treatment of unpaid interest. Measures to permit a credible, flex-
ible response by regulators may also be needed. For example, the intro-
duction of "cease and desist" orders would permit a graduated tightening
of conditions, or the creation of a deposit insurance corporation would
insure depositors and free the central bank from the need to intervene



82 ISSUES OF POLICY REFORM
in failing banks, which is extraneous to its main policy function. Changes
may also be needed in the laws affecting loan recovery.
These three measures, prosaic as they may seem, constitute a vital part
of the restoration process. They may almost amount to a cultural rev-
olution in the way the financial system and the corporate sector interact.
Adequate accounting and auditing constitute the information infrastruc-
ture needed for financial markets to operate well, in the sense of enabling
claims on real assets to be properly valued. In most developing countries,
this infrastructure is sadly lacking.
Why not simply allow insolvent banks to fail? First, if depositors do
not have confidence in the banking system, resource mobilization will
be severely curtailed and gains in efficiency and convenience will be di-
minished accordingly. Second, a wave of nervousness among depositors
may lead to a bank run. Runs that force the liquidation of bank assets
can severely disrupt the economy and significantly set back growth. Even
if a bank run is forestalled by bank closure, this in itself impedes the
operation of the payments system and disrupts economic activity.
In the case of insolvent state-owned banks, some people have argued
that immediate action to reveal the poor quality of the loan portfolio
and to clean up the balance sheet may be unnecessary and imprudent.
Unnecessary because the threat of a bank run may not be great in the
case of state-owned banks and because changing the management may
be sufficient to avoid a recurrence of the poor lending policies. Imprudent
because announcement of the insolvency may damage business confi-
dence, diminish the chances of good recovery on doubtful loans, and
present the government with a budgetary commitment it can ill afford
to absorb in the short run. The choice between the orthodox and dis-
senting views here hinges on the likely success of an attempt to cover up
the insolvency and on the likely effectiveness of a bank management
encumbered with a balance sheet so bad as to make profitability unat-
tainable for years.
Most governments, however, consider it important to preserve the con-
fidence of depositors and have arranged for at least some depositors to
be indemnified from loss when banks fail. In most cases indemnification
has been after the fact, but in several cases an explicit deposit insurance
scheme has been proposed in preference to an implicit guarantee by the
state.
Not everyone agrees on the desirability of formal deposit insurance
schemes, while others argue about what form they should take and
whether they should be compulsory. Those opposed claim that deposit
insurance schemes eliminate the incentive for depositors to select a fi-
nancial institution according to criteria for prudential management and
make them indifferent even to reckless management behavior. An in-
surance fund is rarely sufficient to meet major calls upon it, and efforts



Financial Sector Reform 83
to limit insurance to small deposits are routinely sidestepped by such
practices as deposit splitting. Alternative approaches that might restore
incentives, such as relating the insurance premium to the regulator's per-
ception of the risk of the portfolio, are not easily effected, especially in
developing countries. The main response to this line of criticism is that,
since ex post indemnification of depositors has been widespread, moral
hazard already exists and that explicit insurance is preferable to implicit,
not least because it provides a mechanism for regulation.
A related controversial issue is the extent to which a financial system
can be made self-regulating by mandating extensive disclosure of infor-
mation and relying on depositors to police the banks. For various reasons,
such an approach is not generally thought to be robust. It does, however,
form the philosophical basis of the most comprehensive banking reform
undertaken in a developing country, that of Chile, although the Chilean
system also mandates an active role for the bank superintendency.
Where establishment of a deposit insurance scheme has formed part
of an adjustment program, the insurance agency has had functions be-
yond paying out to depositors of failed banks.7 In such cases, the agency
can also intervene in banks identified as in serious difficulties; this allows
the central bank, which might otherwise be the only agency equipped to
undertake the work, to concentrate on policy as it affects the continuing
banks.
The task of the intervening agency, whatever its nature, is to decide
whether the failed banks should be liquidated or restarted with an in-
jection of capital. A restart can be contemplated only in the context of
a carefully worked-out business plan probably involving considerable
reforms to the internal procedures and policies of the bank. New man-
agement is essential; in the initial, transition phase managers might come
from the intervening agency, but they should be replaced by outsiders as
soon as possible. The next stage is to identify nonperforming parts of
the portfolio and separate them from the remainder of the banks' assets.
If this job is left to bank management, a successful restart of the banks'
operations could be jeopardized. Invariably, a substantial reduction in
staff and branch expenses will be required. It may be quite some time
before the rehabilitated banks are in a position to operate independently
(in the case of publicly owned banks) or be bought by private interests.
The financial restructuring should not include anything for the former
shareholders: it is banks (or their depositors) that are being bailed out,
not the bankers.8
Whether liquidation or rehabilitation is chosen, the protection of de-
positors is costly. Even in countries that already have a deposit insurance
scheme, the fund built up with insurance premiums is rarely sufficient
to meet the deficits that emerge in a major crisis. In most cases, the cost
falls directly or indirectly on the government. A cost borne by the central



84 ISSUES OF POLICY REFORM
bank eventually tends to fall back on the budget and must be managed
and spread over time to prevent an inflationary expansion of domestic
credit. Part of the counterpart funds generated by a policy-based loan
disbursed against general imports can be earmarked for capitalization of
the insurance scheme, as was done in Turkey under its second financial
sector adjustment loan.
To avoid a recurrence of bank failures, prudential regulation and su-
pervision need to be strengthened. Adjustment programs typically rec-
ognize this by proposing an increase in the number and quality of bank
inspectors to review bank management and portfolios and an improve-
ment in the procedures for off-site review and analysis by bank regulators
of returns made by banks. Many problems could have been avoided by
strict adherence to formal procedures for the diversification and classi-
fication of loans, nonaccrual of interest, and provisioning of nonper-
forming loans.
Freeing Up the Financial System
The policy agenda for dismantling controls on interest rates and credit
allocation includes some or all of the following features: removing all
interest rate controls on deposits, withdrawing interest subsidies and
cross-subsidies, reducing the complexity of sectoral credit targets and
their deviation from uncontrolled levels, removing credit ceilings in favor
of broad instruments of monetary control, allowing commercial banks
to engage in underwriting or to invest in private corporate securities,
limiting entry barriers to those required by prudence (which can some-
times include an increase in capital),9 reducing fiscal or quasi-fiscal bur-
dens on the financial sector (such as those caused by low remuneration
of heavy reserve requirements), and liberalizing foreign exchange con-
trols.
All these measures have the general objective of using broad market
signals wherever possible to improve the effectiveness of the financial
sector in mobilizing and allocating resources. In practice, and for a variety
of reasons, however, most countries choose to retain some degree of
control over financial resource allocation and some degree of subsidy in
their financial systems. World Bank programs often focus on reducing
the most egregious deviations from a market configuration, such as the
prevalence of credit at negative real interest rates or at rates below those
available on deposits (which encourages the substitution of borrowed for
one's own funds). As noted below, the timing of liberalization and its
extent are sometimes debatable, especially when financial reforms inter-
act with other policy reforms or macroeconomic imbalances."0 Still dis-
puted is the extent to which one can rely on market signals to meet
development needs for credit in otherwise suboptimal environments (al-



Financial Sector Reform 8S
though it may be argued that the allocation of resources to desirable uses
is better dealt with by fiscal means than by credit allocation).
There is no uniform approach to an ideal regime of taxation of financial
intermediation. Such taxes add to the wedge between borrowing and
lending rates, thereby contributing to inefficiency in the allocation of
resources. Where capital shortage is a major constraint to development,
distortions of this type may be more severe than in capital-rich countries.
For this reason, a reduction in financial sector taxation is often a priority
in financial reform. Excessive zeal in eliminating fiscal burdens on the
financial sector, however, could result in undue taxation elsewhere in the
economy. In inflation-prone countries, the tax should not be linked to
nominal interest receipts because, if inflation and nominal interest rates
rise sharply, the tax could grow quickly in importance and the spread
of interest rates could severely widen.1"
The taxation, explicit and implicit, of financial intermediation could
be used to offset the adverse effects of cartelization and monopoly power
in the banking system. The authorities, faced with a situation in which
banks are able to unduly depress deposit rates and increase lending rates,
can make skillful use of unremunerated reserve requirements to offset
these distortions."2 Although the ideal long-run solution is to encourage
greater competition through liberal (although prudent) entry policies, the
results of such a program may be slow, and to the extent that economies
of scale exist in financial intermediation (itself a much-debated issue),
some degree of monopoly power may remain even when entry is free.
Development of Capital Markets
The development of a capital market typically involves four main areas
of policy. First, legislation is needed to establish an adequate regulatory
framework for both primary and secondary markets so that a variety of
securities can be sold. The framework should include adequate require-
ments for systematic accounting and disclosure of accounts and protec-
tion of minority interests.13 Second, taxation reform is needed to reduce
disparities in the effective rate of taxation of different types of security
in order to reduce fiscal disincentives, especially to the use of equity and
long-term debt.14 Third, technical assistance and training are needed for
the establishment of necessary institutions and intermediaries. Fourth,
an appropriate framework is needed for providing adequate liquidity to
the market. This might, for example, include increased authority for
banks to participate in securities markets (a measure that poses problems
of both prudential regulation and potential conflicts of interest, how-
ever).'5



86 ISSUES OF POLICY REFORM
Problems and Issues of Financial Reform
Adjustment of the financial sector may be unsuccessful or even counter-
productive if not supported by appropriate macroeconomic policies. De-
spite the close connection between the macroeconomic situation and the
health of the financial sector, financial policy loans have been made to
countries in which the macroeconomic situation was less than satisfac-
tory."6 Although a review of experience suggests that countries have gen-
erally complied with the specific conditions for financial policy contained
in such loans, some analysts argue that it is unrealistic to expect a fa-
vorable result when macroeconomic stability is not being achieved.17
They believe that financial sector reforms should be deferred until a more
stable macroeconomic environment can be ensured.
Interaction with Other Adjustment Programs
Indeed, the phasing of financial sector reforms in relation to other reforms
has become an important policy issue. Macroeconomic and trade policy
adjustment measures can have undesirable side effects for the financial
sector, notably by worsening the condition of bank portfolios."8 Con-
versely, distortions caused by unduly rigid financial regulation can be
much greater in the absence of macroeconomic balance. High inflation,
for example, can cause a fixed nominal interest rate to diverge even more
from the market-clearing level and result in a deterioration of resource
allocation and a worsening of credit rationing.19
TIhis interactive nature of reforms raises difficult issues of the phasing
and sequencing of various reforms. Some considerations argue for delay
in the liberalization of financial markets until other aspects of reform
have been implemented. Because of the increased speed and magnitude
of the response of a liberalized financial sector to changes in the policy
environment, some firms and financial institutions that might have sur-
vived under a more controlled regime will fail if the system is subjected
to a significant shock to macroeconomic, trade, or fiscal policy.20
Abrupt liberalization of the financial sector can itself generate unde-
sirable side effects. For example, if all interest rates are deregulated at
once, the (generally) shorter maturity of deposits than of loans will imply
a severe squeeze on the profitability of banks.21 Unless the low interest
rates on loans can be renegotiated, permanent increases in interest rates
may require some form of state subvention if depositors are to be pro-
tected. Similarly, reforms that seek to ease the tax burden on the financial
system will generally result in pressure on the budget. This in turn may
manifest itself in a distorting tax imposed elsewhere in the economy, in
inflation, or in the explosive issue of public debt. An increase in public



Financial Sector Reform 87
debt further raises interest rates and contributes to financial instability,
as Turkey's experience has shown.
Although the consensus is that full-blown financial sector liberalization
is best introduced at a time of macroeconomic and policy stability, it
must be recognized that every delay in eliminating distortions represents
a real loss to the economy (via the adverse effect on resource allocation).
Delay may also be accompanied by a buildup of off-balance-sheet gov-
ernment liabilities. Furthermore, some reform steps can be taken at an
early stage without the same risk entailed in other reforms. This applies
particularly to the correction of serious problems of insolvency and to
institution-building. The interest rate structure can be rationalized by
eliminating preferential and subsidized rates even if it is not yet fully
liberalized. Measures to promote competition, such as announcing (with
adequate lead time) an opening of the system to foreign banks, might be
needed to induce reform in an oligopolistic domestic financial system.
Concern about the fragility of the financial system should not be used
as an excuse to delay reform indefinitely. Rather, it should spur reform
efforts that are sufficiently thoroughgoing to result in robust institutions.
In practice, the timing of financial reforms will probably follow "win-
dows of opportunity" (as, for example, in Argentina), and the shocks
that might arise in the course of other likely reforms will need to be
factored in.
Dissenting Voices
As is the case with most issues of economic policy, some people believe
that conventional advice concerning financial sector reform neglects im-
portant institutional considerations. Some of these arguments are pre-
sented here.
Some commentators argue that the higher interest rates that result from
liberalization are likely to reduce growth and increase inflation, at least
in the short run.22 This is an empirical issue that has not yet been fully
resolved, although the evidence is currently against this dissenting view,
at least in the long-run context. The experience of the Southern Cone
countries in Latin America, especially Chile, in the late 1970s is often
mentioned in this connection.23 Although it is true that real interest rates
soared in Chile, with real bank lending rates averaging 74 percent in
1975-78, isolating the role of interest rate liberalization in a period of
extensive policy changes is difficult. Not only were there tariff reforms
and an opening of the capital account during the same period as liber-
alization of the financial sector, but inadequate supervision of financial
institutions and an unsustainable exchange rate policy also influenced
events. Despite the rapid growth experienced by Chile in the years im-
mediately following the reforms, the financial instability that ensued re-



88 ISSUES OF POLICY REFORM
mains a specter that is hard to exorcise and that haunts discussions of
financial liberalization in other countries.
Another area of disagreement relates to the potential conflict between
short- and long-term objectives arising from the use of monetary policy
to combat inflation. Structural changes that substitute market-based
monetary policy instruments for bank-by-bank credit ceilings pose po-
tential short-term risks for monetary control. Although it should in prin-
ciple be possible to neutralize any undesired side effects of the change
on macroeconomic conditions, lack of experience in the operation of the
new (mostly indirect) system could cause problems. Some policymakers
wish to retain certain features of the unreformed system (such as credit
ceilings) despite their structural inefficiency because they perceive that
the features make monetary control easier. This perception may make
reform programs less ambitious than they might otherwise be.
A closely related argument is based on the controversial theory that,
in a liberalized financial system, monetary restraint might be compara-
tively ineffective (or even counterproductive) in restoring balance in cur-
rent account payments and the macroeconomy in general. According to
this neostructuralist view, monetary restraint might choke off production
as much as, or even more than, it does demand, whereas in a system of
directed credit, monetary restraint could be confined mainly to less pro-
ductive sectors or to consumer spending. This line of argument is par-
ticularly relevant to the short run, and it is weakened if fiscal policy
instruments more appropriate to achieving macroeconomic balance are
available.
More generally, it is sometimes held that the financial system will not
on its own channel funds to the most socially, as opposed to privately,
productive sectors, so that some degree of directed credit is needed. In
particular, it is argued, a "sound and totally liberalized" financial system
may not always be the most desirable for development. Some situations
lend themselves to second-best considerations, for example, where a poor
legal framework in a volatile economy leads the formal sector to consider
only short-term, heavily collateralized lending and inhibits equity mar-
kets.24
A conventional response to this point is that the divergence between
social and private valuations is best met with explicit tax or subsidy
measures because using directed credit can conceal a very large implicit
subsidy that may well be far in excess of the difference between social
and private valuations. In addition, the use of subsidized credit frequently
involves high leakages because of fungibility. This is only a partial re-
sponse, however. To the extent that the discrepancy being identified is
the result of an institutional failure, such as the absence of certain fi-
nancial instruments, it may well be that the best cure lies partly in policy



Financial Sector Reform 89
interventions in the financial sector. This does not necessarily argue, how-
ever, for programs and institutions of the types common at present.
Frequently discussion and disagreement center on the institutional ar-
rangements for adequately implementing monetary policy and prudential
regulations and for dealing with failed banks. Should these functions be
united under one institution-say, the central bank-or is a separation
of functions desirable? Also, should insurance intermediaries and secu-
rities markets have their own regulators, or does the need for coordi-
nation argue for a single regulatory body?
Country experience on this issue varies, given the varying degree of
complexity of financial systems and the resources that are available to
implement financial policy. Many countries accept the importance of a
degree of independence of the monetary policy authority from other as-
pects of government, although the degree of independence varies.25 The
responsibility for bank supervision is given sometimes to the central bank,
sometimes to a government ministry, and sometimes to a separate su-
perintendency of banks. A separate superintendency can have merit to
the extent that it is needed to ensure that sufficient attention is devoted
to prudential supervision in the face of competing demands on the at-
tention of senior central bank or ministry officials. But whatever insti-
tutional arrangement is preferred, it is essential that the bank supervision
function be adequately funded and that there be an adequate two-way
flow of information between the supervisors and those responsible for
monetary policy.
Another debate concerns universal banking. Although the introduction
of universal banking has been recommended in some World Bank pro-
grams (an early example was in the Philippines), some analysts believe
that allowing banks to become involved in corporate securities weakens
prudential safeguards and obstructs the development of autonomous cap-
ital markets. This conflict mirrors the longstanding debate between ad-
herents of the traditional Japanese and German approach to banking and
those supporting the approach of the United Kingdom and the United
States. By now, however, it seems clear that attempts to retain rigid
barriers between segments of the financial sector can have, at best, a
finite lifetime.
All in all, these and other dissenting voices call attention to the need
to think through reforms carefully and to take account of the specific
institutional features of each country.
Winners and Losers
Achieving consensus on financial sector reforms is usually difficult be-
cause of the powerful special interests that are affected.26 A highly reg-
ulated banking system is typically a cartelized and potentially profitable



90 ISSUES OF POLICY REFORM
one. Opening the system up to competition will erode high profits and
allow new institutions and markets to flourish at the expense of the old.
In anticipation of such effects, therefore, there will be strong lobbying
against liberalization.
There are losers as well from the establishment of tighter prudential
regulation. Uncreditworthy borrowers who received loans from badly
managed banks will no longer have access to credit. That they have
received loans in the past, however, suggests that they will form a pow-
erful lobby against reform. In most developing countries, extensive ties
exist between major banks and borrowers; moving them to an "arms-
length" relationship attacks the basis of the economic power structure
and is therefore a slow and difficult process.
It is often asserted that financial liberalization will improve income
distribution and reduce the concentration of economic power. There is
probably something to this assertion, although the evidence is not con-
clusive. Credit allocation is normally more concentrated than are deposit
holdings and tends to follow the distribution of assets available for col-
lateral.27 Many examples suggest that subsidized credit is in practice
available mostly to the larger or more prosperous segments of the targeted
sectors. Small savers benefit from the liberalization of deposit rates. The
very poor have little or no access to formal credit, however; their interests
lie in the improvement of economic growth and employment that may
be expected to flow from reform. Reform may also end the artificial
stimulus to capital-intensive production that severely negative interest
rates provide.
Drastic changes in interest rates can have very large distributional ef-
fects, however, to the extent that they affect long-term loans. This effect
is especially important with respect to housing credit, which is heavily
subsidized in many countries. Middle-class borrowers may have entered
into long-term commitments wholly unaware that interest rates would
be much higher without the subsidies. Even though these borrowers have
gained substantially from the subsidy, they might be unable to meet their
commitments at the unsubsidized rate. In practice, both equity and po-
litical considerations may call for partial compensation of long-term bor-
rowers who are severely affected by a change in the interest rate regime.
But such compensation schemes must be strictly circumscribed if the
benefits of the reform are not to be dissipated. Certainly the new interest
rates must apply to all new borrowing, and the scheme of compensation
should be simple, open, and preferably funded out of fiscal resources.
Given the sensitivity of the issues addressed, governments and the
World Bank sometimes prefer not to state explicitly in loan documents
some of the steps that may need to be taken, but let them be implied by
other requirements. As with other policy-based loans, full implementa-
tion of a financial sector adjustment loan is most likely when the gov-



Financial Sector Reform 91
ernment is firmly committed to reform and may even have already taken
some preparatory steps prior to the loan.
The Success of Financial Adjustment Policies
The ultimate objective of financial sector reform is to achieve a higher
rate of sustained economic growth. Attempts to measure the success of
reform in these terms, however, are compromised by the multitude of
other factors that contribute to economic growth. The effectiveness of
financial sector reforms also needs to be examined in terms of inter-
mediate objectives. For example, has the liberalization of interest rates
contributed to a higher savings rate, to a higher rate of monetization, or
to reduced capital outflow? Has the removal of sectoral credit controls
led to more productive investment choices? Has the removal of bank-
by-bank credit restrictions reduced bank margins? Have the rehabilita-
tion and improved supervision of banks resulted in a sound and energetic
banking system? Has the elimination of directed credit resulted in reduced
loan delinquencies? Have structural improvements in capital markets led
to increased recourse to the capital market?
In most cases, the financial adjustment programs supported by the
World Bank have not been in effect long enough to provide definitive
answers to these questions. On the basis of the cases that have been
reviewed, it seems that countries have complied with most of the con-
ditions of the programs. But financial sector reforms, being part of quick-
disbursing, policy-based operations, have sometimes been introduced in
problematic macroeconomic circumstances, and in some cases the poor
macroeconomic environment has continued throughout the period of
loan implementation.28
More generally, the degree to which an efficient financial system can
contribute to the process of economic growth has not been precisely
established. Certainly, a well-performing financial system is no remedy
for policy weakness elsewhere in the economy. Nevertheless, when things
go seriously wrong in the financial system, such as hyperinflation or
widespread bank failures, the impact on the economy can be sudden,
dramatic, and long-lived. Recent research on the depression in the United
States in the 1930s attaches considerable importance to the impact of
bank failures on the availability of credit (independent of the deflationary
effect of monetary contraction). In some episodes of financial insolvency
in developing countries, cause and effect have been difficult to disentangle
because the insolvency of the banks became evident only at the time of
recession.
Some indirect evidence of the influence of financial reform on economic
growth may be obtained by an examination of countries whose financial
sectors vary from very repressed to close to the ideal of liberalization



92 ISSUES OF POLICY REFORM
programs. The relative contribution of liberalized and repressed financial
systems to economic growth has also been considered in a number of
statistical studies (Fry 1988; Gonzales Arrieta 1988). In a review of
twenty-one countries during 1971-80 (IMF 1983), the six countries with
positive real interest rates grew more rapidly than the ten with moderately
negative rates, which in turn had average growth rates well above those
of the five with severely negative rates. A recent study of this issue (Gelb
1989) using 1965-84 data for thirty-four countries also finds a positive
impact of interest rates on growth.29
Evidence on the impact of distortions caused by directed credit is even
less easily quantified, essentially because of fungibility, which makes it
difficult to assess what would have happened in the absence of directed
credit. Few, if any, studies conclusively attribute favorable results to pro-
grams of directed credit. In many countries, institutions implementing
directed credit programs have faced serious problems with the quality of
loans, which suggests poor credit allocation. To the extent that directed
and subsidized credit diverts funds from still more worthwhile projects,
it may cause social losses even if it is profitable for the financial inter-
mediary.30
The Form of Financial Sector Loans
A policy-based loan has the effect of financing part of the borrowing
country's international payments deficit, as well as supporting the budget
through the provision of counterpart funds.3' A typical loan for financial
policy reform might be disbursed against general imports over a period
of two years in two tranches. Counterpart domestic funds are generated
as importers turn in domestic currency for foreign exchange; what hap-
pens then depends on how the government uses these payments.
The major fiscal costs of financial sector reforms are of two types. One
results from a financial liberalization that substantially removes the fi-
nancial system from the tax base. In a repressed financial system, high,
unremunerated reserve requirements may yield the government the an-
nual equivalent of 2 to 3 percent of GDP, and lowered debt-servicing
charges from repressed interest rates on public debt may amount to a
further percentage point or two. Liberalization may therefore have an
appreciable fiscal cost, especially in terms of present value. There may
also be some fiscal cost in compensating for the elimination of interest
cross-subsidization.32 A second fiscal cost attends financial restructuring
as depositors and debtors are bailed out. This cost can represent as much
as 20 to 30 percent of one year's GDP, as the experience of Chile and the
Philippines suggests, but it can be spread out over time.33 For example,
the government can provide intermediaries with interest-bearing paper



Financial Sector Reform 93
in exchange for the bad portfolio to enable intermediaries to continue
to serve depositors.
Compared with these two components, the fiscal costs of institutional
changes, such as introducing external auditing, are naturally small. The
major cost of such actions can be political, inasmuch as they affect pow-
erful interest groups. Financial sector reforms of the type being discussed
do not typically have important direct foreign exchange costs.34
The amount of counterpart funds generated by World Bank loans for
financial reforms is not usually closely related to the fiscal costs of those
reforms nor to the time profile of such costs. Depending on the size of
the loan and the nature of the policy reforms, such loans may cover only
a part of the fiscal costs (Chile, Ghana, Philippines) or more than the
fiscal costs (Turkey). The loans will generally far exceed any direct foreign
exchange costs of the reforms.
Inasmuch as balance of payments considerations have substantial
weight in determining loan amounts, it seems that, by their nature, fi-
nancial reforms do not provide a good yardstick for judging the appro-
priate size of policy-based loans. Should loans therefore be designed to
establish a tighter link between disbursements and fiscal costs? For fi-
nancial sector reforms, this would mean severely reducing the size of
tranches linked to the reform of accounting and auditing practices; link-
ing disbursements to market liberalization (but only to ease adjustment
to a better fiscal system; there is little point in reducing taxes on financial
transactions only to raise them in other distorting ways, such as on ex-
ports); and establishing a link between disbursement and the fiscal costs
incurred in the process of financial restructuring.35 This would have im-
plications for the speed of disbursement and would require major changes
in the type of loan. Major disbursements would be linked to the adjust-
ment costs of fiscal reforms and of restructuring only, with small technical
assistance components for reforming the financial infrastructure. It would
perhaps be necessary to supplement this by an explicit balance of pay-
ments component, with its own appropriate conditionality.
The disbursement of financial intermediation loans, by contrast, de-
pends on the speed of disbursement of the corresponding credit lines to
final borrowers, even when such loans include policy reform as a con-
dition.36 As noted, the policy content of financial intermediation loans
has typically been limited to the raising of subsidized interest rates. Rais-
ing these interest rates is often necessary so that lines of credit disbursed
on terms acceptable to the World Bank will not be undercut by cheaper
domestic lines. This introduces a link between policy reform and dis-
bursement, but it is implicit rather than explicit.
It may not be realistic to expect the "real" economy to be reformed
prior to undertaking the reform of a financial sector, but financial policy
loans must be part of a viable package. They must be designed in the



94 ISSUES OF POLICY REFORM
context of an adequate macroeconomic policy framework if they are not
to misfire or result in a deferral of needed macroeconomic adjustment.
Adherence to specified macroeconomic conditionality should be required
for the disbursement of financial adjustment loans to ensure that the
recipients of such large and quick-disbursing loans do not defer or slow
the implementation of other measures needed to ensure lasting macro-
economic adjustment.
In cases where the financial infrastructure is deficient, is it realistic to
expect rapid results from reforms? Banks may never have had to develop
the credit processes appropriate for a market economy, and deep insti-
tutional reforms take time. How soon can we expect a change in the
banking culture of a country? It takes at least five years to train a bank
examiner. Legal reforms also take time, especially in countries with a
division of executive and legislative powers. The experience of Chile and
the Philippines shows how slow and difficult it can be to dissolve own-
ership links between banks and firms. In these circumstances, quick-
disbursing policy-based operations may not be appropriate. A more ef-
fective approach might be to combine smaller technical assistance pack-
ages that disburse over periods of about five years with complementary
larger, quick-disbursing loans provided at the time of financial restruc-
turing.
In planning such policy reform operations, the adverse consequences
for bank portfolios of adjustment measures to redirect flows of factors
and goods (notably exchange rate adjustments and trade liberalization)
are not sufficiently taken into account. They should be anticipated and
measures taken to prepare for them. For example, bank regulation and
supervision could be addressed at an early stage, rather than waiting until
problems multiply.
Other questions need to be answered. Given the poor financial con-
dition of banking systems in many countries (and the adverse implications
for resource allocation that follow), should the volume of financial in-
termediation loans increase rapidly? Is there a danger of further weak-
ening the system by piling still more debt onto undercapitalized borrow-
ers? Given the poor history of directed credit, at least in terms of
repayments, is debt the right instrument to foster development? Can eq-
uity or pseudo-equity contracts be used instead to strengthen capital
structures? How would such a move be compatible with the lending
policies of the World Bank?
What kind of financial system are we trying to achieve with restruc-
turing? In particular, what role should development banks play in the
more liberalized systems of the future? If these questions are not consid-
ered carefully, the restructured banks may face a long-run earnings prob-
lem that leads them quickly back to a distressed condition.37



Financial Sector Reform 95
Conclusion
The typical package for financial sector reform involves policy changes
to increase the power of centralized decisionmaking in some areas and
to reduce it in others. In the area of prudential regulation and supervision,
reforms seek strengthened information systems, stronger and more de-
tailed regulations, and closer central supervision. At the level of the in-
termediaries, reforms seek improved procedures, some of which (such as
credit policies, loan review, and management information systems) are
natural complements to improvements at the central level. In regard to
the relative cost and availability of credit, the typical reform program
calls for a reduction in government control and tries to broaden the range
of options for finance.
The internal consistency of policy packages of this type follows from
certain assumptions underlying the market-economy model, in particular
the assumption that the qualities of goods and services will be reasonably
apparent to participants in the market so they can formulate appropriate
plans for production and spending. Liberalizing the financial market
without also ensuring a correspondence between financial contracts and
the values and productivity of the real assets underlying them does not
help strengthen market mechanisms. It may even undermine them, as
indicated by the experience of more than one country.
It is also possible, of course, for the reform pendulum to swing too
far-and in several directions. If carried too far, prudential regulation
can strangle financial innovation. A drive focused exclusively on priva-
tization can cause much-needed public sector reforms to be neglected. In
general, however, given the starting positions of the financial systems in
most developing countries, financial reforms in the indicated directions
are vital for strengthening the role of the market in allocating resources,
both internally and internationally. They will also be needed if developing
countries are to take advantage of the internationalization of finance that
is currently under way.
How should financial sector operations evolve? Many needed financial
reforms are institutional and infrastructural and require the acquisition
of scarce skills. They take time to become effective, and it is not clear
that a quick-disbursing policy-based operation is the best medium for
effecting them. It might be better to formulate medium-term, lower-in-
tensity operations for certain components of financial sector reform.
Given the close links between financial policies and performance and the
macroeconomic situation, the macroeconomy needs to be taken into ac-
count when planning reforms, especially when they have fiscal impli-
cations. The implications for the financial system of other reforms, such
as reform of the trade regime, need to be considered at an earlier stage
than they have been. Disbursement of financial reform loans could be



96 ISSUES OF POLICY REFORM
tied in more closely to the costs of the reforms, which are primarily those
related to restructuring and fiscal adjustment. And, as with other reforms,
if financial reforms are not embraced intellectually by the government
of the adjusting country, they will be adopted reluctantly and will surely
fail.
Notes
1. For an assessment of the causes of bank failure in the United States, see
U.S. Government (1988).
2. It is difficult to assign relative weights to internal and external causes of
failures among financial institutions. Macroeconomic disturbances often hasten
the demise of weak financial institutions and may be mistaken for the underlying
causes. Conversely, an institution weakened by external shocks may be partic-
ularly prone to experience a deterioration in the quality of management (see De
Juan 1987).
3. For a discussion of the consequences of widespread insolvency, see Hinds
(1988).
4. There is some evidence that countries with repressed financial systems grow
more slowly and that their output/capital ratios are low (see Gelb 1989).
5. Among the financial instruments that the World Bank has frequently sought
to expand in developing countries are long-term loans. But how important is the
provision of long-term loans? If they were vital, the long-term resources made
available from World Bank funds, for example, could be expected to command
much higher interest rates than do short-term loans. Top-grade borrowers feel
reasonably confident of being able to roll over short-term bank loans and so are
unwilling to pay much of a premium for long-term funds. It seems likely, instead,
that unsatisfied demand for long-term loans has come mainly from less credit-
worthy clients. Willingness to pay a maturity premium thus could signal low
creditworthiness, which could make a competitive banking system reluctant to
make long-term loans to a broader spectrum of clients, even at a premium. The
advantages to a bank of being able to review and recall loans at regular intervals
should not be minimized. With early evidence of a borrower's deteriorating po-
sition, prompt action can possibly turn the position around, whereas the holder
of a long-term claim may have no comparable sanction.
6. Corporations in developing countries often have higher debt/equity ratios
than those in developed countries, apparently because of low retention of profits.
7. In the case of a structural adjustment loan to Chile, for example, a separate
superintendency of banks already existed, but there was no formal deposit in-
surance system.
8. In some cases, debt-relief programs may also assist borrowers. In Chile loans
were reprogrammed to reduce interest rates and extend terms, and preferential
exchange rates were introduced after devaluation. Such debt-relief programs for
borrowers may be introduced to avoid a massive sale of assets, which would
further reduce their prices and add to the burden of bailing out the system.
9. Reducing entry barriers can strengthen a banking system when it has been
dominated by industrial-financial groups that have relied too much on within-
group lending, as in Chile, the Philippines, and Turkey. In 1980 the Philippines
offered universal status to banks with sufficiently diversified ownership and en-
couraged the entry of foreign shareholding partners. The success of this measure
can be debated; ownership of many Philippine banks remains dominated by



Financial Sector Reform 97
family groups, although the leavening of outside shareholders may have had
beneficial effects. The procedures for the restructuring or liquidation of insolvent
banks can be viewed as constituting a parallel reduction in the barriers to exit.
10. Some theorists have observed that partial liberalization of a formal bank-
ing system may even worsen the supply of credit to the productive sector to the
extent that it diverts deposits from a curb market that was more efficient in
channeling deposit resources to that sector. This argument seems to be based
mainly on experience in Korea, where curb markets are extensive and, unlike
formal sector banks, are not subject to reserve requirements. It is not evident
that curb markets are generally more efficient than the formal banking system,
which might give this point broader application; furthermore, the use of unre-
gulated intermediaries increases the risks of crisis (see, for example, Sheng 1988).
11. This happened, for example, in the Philippines where, without any change
in tax rates, the burden of the tax on gross receipts of financial institutions
approximately tripled in 1984. The remuneration of large reserve requirements
at low fixed rates has a similar impact as market rates increase.
12. In Ghana, for instance, banks were able to command a very high spread
between borrowing and lending rates because of the interaction of credit ceilings
and an inflow of funds from abroad.
13. In Turkey enactment of a Capital Market Law (in 1981) and establishment
of a Capital Market Board (1982) were among the conditions of structural ad-
justment lending. Unfortunately, macroeconomic conditions did not favor the
growth of the market, and its role in financing private investment remained
minor.
14. In Chile, Jamaica, and Malaysia, for example, bank deposits have been
exempt from the withholding tax that applies to securities. This discriminatory
taxation discouraged the holding of corporate debt and the development of the
capital market in general.
15. There is no consensus on the optimal structure of financial intermediation
and its relationship with capital markets. Some developed countries, notably the
Federal Republic of Germany and Japan, have done well with "bank-based"
systems in which banks and business firms have close relationships; in others,
the system is "market-based," with banks and firms more at arms-length. In
some countries, banks are "universal"; in others, financial institutions are more
specialized. How the various elements of a financial system perform is probably
far more important than the precise configuration of the system.
16. Argentina and Turkey faced major macroeconomic imbalances during im-
plementation of financial sector reforms. After floods and a disastrous earthquake
in Ecuador during early 1987 the fiscal deficit expanded and inflation accelerated.
Implementation of financial sector reform in such a sharply worsened environ-
ment was clearly going to be more difficult. The major problems with the ex-
periments of the 1970s with financial liberalization in the Southern Cone coun-
tries may be attributed mainly to inappropriate macroeconomic policies
(especially exchange rate policy), although inadequate prudential supervision and
regulation were also to blame.
17. Some governments have even taken measures that exceeded the loan re-
quirements, as in Argentina and Ecuador. But there have also been cases in which
reforms have subsequently been partially reversed; the financial transactions tax
in Turkey, for example, was initially reduced from 15 to 3 percent, but was
subsequently increased. Furthermore, continuing high budgetary deficits in Tur-
key have kept real interest rates high and inhibited recovery of the financial sector
despite the sectoral adjustment measures. For a discussion of the consequences,
see Atiyas (1989).



98 ISSUES OF POLICY REFORM
18. It could be argued that the emergence of this problem in Turkey in the
early 1980s could have been anticipated far earlier in the structural adjustment
program. Perhaps more should have been done to improve auditing and bank
supervision so that the problem could at least have been monitored. By the time
actions of this type were introduced in financial sector loans, the damage to the
banks' portfolios was already serious.
19. For example, fixed rates on rural lending in Brazil resulted in large losses
on official credit lines as inflation soared following the breakdown of the cruzado.
20. In Chile, for example, liberalizing the capital account of the balance of
payments in an environment of market-determined interest rates resulted in a
surge in the external value of the currency, overborrowing, and the puncturing
of the speculative boom in real assets.
21. The problem is particularly acute in the case of housing loans. In Hungary,
for example, a very high proportion of the portfolio of the financial system is
in the form of long-term fixed-interest housing loans, which are now well below
market levels.
22. There are two separate issues here: first, whether real interest rates "over-
shoot" or rise to unduly high levels following liberalization, and second, whether
the higher interest rates actually depress economic activity and worsen inflation.
23. McKinnon (1988) compares financial sector reforms in Chile with those
in other countries, notably in Asia.
24. This point has been much discussed in the context of the celebrated Dewey
Dee affair in the Philippines, in which investors' confidence was badly jolted
when a prominent financier suddenly absconded, leaving large money market
debts behind him. It is to fill this perceived gap in the financial system that many
see the proper role of DFIS, not only in the Philippines but in many other countries
as well, notably in Sub-Saharan Africa. If the gap is an imaginary one, then both
the role and the viability of DFIS are placed in question.
25. Perhaps the greatest degree of independence is provided by adherence to
a multicountry currency zone, such as the CFA zone in Africa, where the central
banks, being international institutions, are not answerable to any one govern-
ment.
26. This is true for all policy reforms; spurred on by massive insolvency, some
financial reforms have been more easily accomplished than, for example, tariff
reforms.
27. For an interesting analysis of the impact of collateral on credit allocation
and hence of efficiency, see Feder and others (1988).
28. Further examination of the Chilean case suggests that financial sector
reform can have the beneficent effects suggested above if the macroeconomic
situation is kept under control. Real interest rates and bank spreads have mod-
erated and the real volume of deposits and bank credit has grown since the
reforms, all of which support the rapid growth of output.
29. This study identifies a causal chain from higher interest rates, through
greater mobilization of savings by the financial system, to increased efficiency
of investment, and thereby to growth. Analyses also suggest that high interest
rates on deposits encourage additional saving in developing countries (notably
in Asia), but that the effect is very small. There are also a number of clear-cut
examples of the effect of real interest rate increases on monetization; often this
seems to have been attributable to a reflow of funds previously held outside the
country.
30. Korea is sometimes mentioned as a country that experienced rapid growth
even though a selective credit policy was in operation. But rapid growth in the



Financial Sector Reform 99
uncontrolled financial curb market substantially offset the impact of this policy
on the overall distribution of credit there. A more conventional story is that of
Indonesia, which introduced a system of administrative allocation of credit in
1974 (and abandoned it in 1983). Despite the higher rate of investment in In-
donesia after 1974, the growth rate of output actually declined, which suggests
that investment had not been especially well directed.
31. For a discussion of counterpart funds and the relationship with monetary
policy, see Roemer (1988).
32. This has been a notable factor in the case of Turkey. In the Philippines,
by contrast, the higher interest rates of 1984 probably contributed to government
revenue.
33. Of course, in the absence of restructuring, a government would also incur
costs (as well as the buildup of claims due to weakening institutions) because
subsidies would be needed to maintain the liquidity of insolvent intermediaries.
34. It might be argued, however, that if depositors were to suffer losses they
would reduce their spending on imports as well as on domestic output. If so,
recapitalization of the banking system, which protects deposits, could have in-
direct foreign exchange costs.
35. The second financial sector adjustment loan to Turkey contained a com-
ponent specifically for restructuring, which linked disbursements with the cost
of system reform.
36. The terms of World Bank loans are typically far longer than the terms of
the loans offered to final users of funds, and there may be no requirement that
domestic repayments be relent again up to the term of the bank loan. In such
cases, the World Bank loan initially funds the private sector, but later it funds
the government.
37. This issue is now being faced by at least one large institution restructured
according to a World Bank program.
References
Atiyas, Izak. 1989. "The Private Sector's Response to Financial Liberalization
in Turkey: 1980-82." PPR Working Paper 147. Policy, Planning, and Research,
World Bank, Washington, D.C.
Cho, Yoon Je. 1986. "Inefficiencies from Financial Liberalization in the Absence
of Well-Functioning Equity Markets." Journal of Money, Credit and Banking
18 (2) May:191-94.
De Juan, A. 1987. "From Good Bankers to Bad Bankers: Ineffective Supervision
and Management Deterioration as Major Elements in Banking Crises." Fi-
nancial Policy Division, Country Economics Department, World Bank, Wash-
ington, D.C.
Diaz Alejandro, Carlos F. 1985. "Good-Bye Financial Repression, Hello Finan-
cial Crash." Journal of Development Economics 19 (September-October):1-
24.
Dooley, M. P., and Mathieson D. J. 1987. "Financial Liberalization and Stability
in Developing Countries." WP/87/19. International Monetary Fund, Wash-
ington, D.C.
Feder, Gershon, Tongroj Onchan, Yongyuth Chalamwong, and Chira Hongla-
darom. 1988. Land Policies and Farm Productivity in Thailand. Baltimore,
Md.: Johns Hopkins University Press.



100 ISSUES OF POLICY REFORM
Fry, Maxwell J. 1988. Money, Interest, and Banking in Economic Development.
Baltimore, Md.- Johns Hopkins University Press.
Gelb, Alan. 1989. "Financial Reforms and Economic Performance." Financial
Policy Division, Country Economics Department, World Bank, Washington,
D.C.
Gonzales Arrieta, C. M. 1988. "Interest Rates, Savings, and Growth in LDCS: An
Assessment of Recent Empirical Research." World Development 16 (5):589-
605.
Hinds, Manuel E. 1988. "Economic Effects of Financial Crisis." PPR Working
Paper 104. Policy, Planning, and Research, World Bank, Washington, D.C.
IMF. 1983. Interest Rate Policies in Developing Economies. Occasional Paper 22.
Washington, D.C.: International Monetary Fund.
McKinnon, R. 1. 1988. Financial Liberalization and Economic Development: A
Reassessment of Interest Rate Policies in Asia and Latin America. International
Centre for Economic Growth Occasional Paper. San Francisco, Calif.
Roemer, Michael. 1988. The Macroeconomics of Counterpart Funds. Discussion
Paper 262. Cambridge, Mass.: Harvard University, Institute for International
Development.
Sheng, Andrew. 1988. "Financial Adjustment in a Period of Disinflation: The
Case of Malaysia." Financial Policy Division, Country Economics Department,
World Bank, Washington, D.C.
U.S. Government, Office of Comptroller of the Currency. 1988. Bank Failure:
An Evaluation of the Factors Contributing to the Failure of National Banks.
Washington, D.C.
World Bank. 1985. "Financial Intermediation Policy Paper." Industry Depart-
ment, World Bank, Washington, D.C.
Comments
Maxwell J. Fry
ALAN GELB and Patrick Honohan provide a concise summary of the com-
ponents of World Bank adjustment programs that address the reform of
financial sectors in developing countries. The basic principles they pro-
pound appear entirely reasonable. The authors recognize, however, that
precise measurement of the effects of such reforms is impossible; financial
reforms invariably come as part of a larger package of policy measures.
Furthermore, World Bank loans specifically for financial sector adjust-



Financial Sector Reform 101
ment are in their infancy. In my view, the main danger is expecting too
much, too soon.
If, as Gelb and Honohan state (and I concur; Fry 1988), the three main
problems that plague the financial sectors of developing countries are
massive insolvency, excessive controls on interest rates, and domination
by commercial banks, a question immediately arises concerning the ap-
propriateness of using short-term adjustment lending to solve them. Al-
though excessive interest rate controls can be removed rapidly, the prob-
lems of insolvency and domination by commercial banks require more
time. Even in the case of interest rate reform, however, there is now ample
evidence that the deregulation of interest rates can produce perverse ef-
fects in the absence of at least a modicum of macroeconomic stability,
adequate prudential supervision of the financial system, and a reasonable
degree of competition.
If the financial sector is important in the development process, a short-
run adjustment plan is not sufficient to ensure its health; a program for
long-run financial development is clearly needed as well. By itself, a short-
term adjustment program may remove excessive interest rate controls
and may shore up some insolvent financial institutions, but it cannot
tackle the underlying problems of insolvency and commercial bank dom-
ination. Nor can it relieve what is perhaps the most severe constraint on
the financial sector of many developing countries: the lack of adequate
accounting and audit skills, both for prudential supervision and for in-
ternal management. To tackle these problems, a much longer time frame
is needed.
Gelb and Honohan point out that insolvency and excessive control of
interest rates and the direction of credit are linked. It also appears from
the experience of some Latin American countries that insolvency and a
laissez-faire policy are linked. The paradox is resolved, as both Long
(1983) and the present authors recognize, by understanding that more
rather than less prudential supervision is a prerequisite for successful
financial liberalization. Veneroso (1986) concluded that all cases of ex-
traordinarily high real interest rates were caused by distress borrowing
in conjunction with deposit insurance. The lack of adequate regulation
of banking practices led to undue risk taking on the part of the banks,
while deposit insurance enabled them to acquire risky assets without
provoking the withdrawal of deposits. When the level of nonperforming
assets rose, banks raised deposit rates to attract more funds to pay the
interest on existing deposits. This false demand for credit by distress
borrowers put upward pressure on interest rates, which in turn dragged
down more firms.
Although there is no evidence that banking is necessarily more costly
in developing countries than in countries of the Organisation for Eco-
nomic Co-operation and Development (OECD) (Hanson and Rocha 1986),



102 ISSUES OF POLICY REFORM
spreads tend to be substantially wider nonetheless. Other things being
equal, one would expect to find no significant difference in the spreads
between average deposit and loan rates in the two groups of countries.
Other things, however, are not equal. Specifically, developing country
banking systems are less competitive, are more heavily taxed, and suffer
higher rates of loss than banks in OECD countries. In Uruguay, for ex-
ample, because of the oligopolistic structure of the banking system, the
spread between average deposit and loan rates remained more than 15
percentage points until June 1981, despite reduced required reserve ratios
(Hanson and de Melo 1985).
Hanson and Rocha (1986) also find high operating costs for banks in
developing countries with high inflation rates and little or no competition.
Financial repression, which is intensified as inflation rises, raises banks'
cost ratios by reducing the real size of the banking system and at the
same time encouraging nonprice competition, such as the proliferation
of bank branches. Implementing selective credit policies also involves
substantial administrative costs for the banks (Morris 1985).
Almost all developing countries possess specialized financial institu-
tions in the form of development finance institutions (DFIS) or develop-
ment banks. A major problem inherent in virtually all specialized finan-
cial institutions in developing countries springs from the fact that they
are established to lend to borrowers that existing financial institutions-
the commercial banks-have avoided. By and large, commercial banks
choose not to lend when the perceived risks are too high, so specialized
financial institutions have been set up deliberately to lend to high-risk
borrowers. They are not, however, compensated with higher loan rates
for assuming higher risks. Invariably, therefore, the specialized financial
institution set up to support a problem sector of the economy itself be-
comes a problem institution.
Very few DFIS have become self-supporting, autonomous financial in-
stitutions capable of mobilizing resources entirely on commercial terms.
First, DFIs have often been required to make loans at low rates of interest,
frequently negative in real terms. Second, DFIS have financed projects that
have been evaluated as economically viable (often by the World Bank)
but which are not financially viable. About one-third of the DFIS in de-
veloping countries are now in serious financial difficulties, mainly be-
cause of their large percentage of nonperforming loans. The problem of
arrears increased substantially with the worldwide recession and dete-
riorating terms of trade for developing countries in the early 1980s
(World Bank 1985).
Uncompetitive financial markets have been another side effect of gov-
ernments' selective credit policies. In many developing countries, spe-
cialized financial institutions have been established as a result of gov-
ernment decrees, with exclusive franchises for particular financial



Financial Sector Reform 103
activities or particular sectors of the economy. Not only has specialization
produced in this manner generally raised intermediation costs and de-
stroyed any actual or potential competition, it has also failed to achieve
the goal of an adequate and efficient distribution of credit. All too often,
specialized banks have expropriated scarce resources to finance large and
inefficient investments, while small investment projects have been starved
of funds. In most developing countries, less specialization and compart-
mentalization are prerequisites for an efficient and aggressive financial
sector.
Initially, the World Bank lent to specialized financial institutions
through its financial intermediation loans to DFIS. Since 1975, however,
when it decided to increase its lending to small and medium-size enter-
prises, the World Bank has found commercial banks to be more effective
than DFIS in channeling funds to wider groups of enterprises (Levitsky
1986). Specifically, these programs have worked best when all financial
institutions have had access to World Bank funds through a rediscount
mechanism with an apex institution rather than when funds have been
earmarked for specific lending institutions.
Now Gelb and Honohan conclude that the dubiousness of many bank-
ing systems raises questions about the allocation of resources interme-
diated by the commercial banks as well. Despite a voluminous body of
literature (much of it produced in the World Bank) criticizing directed
credit programs, the World Bank continued its selective credit policy in
lending to small and medium-size enterprises. These resources might bet-
ter have been used to improve commercial bank lending in general rather
than to compound the problem of dubious loans in commercial bank
portfolios.
The World Bank's policy stance on financial sector issues has changed
dramatically over the past two decades. World Bank reports on the fi-
nancial sectors of developing countries in the late 1970s and early 1980s
tended to focus on macroeconomic and sectoral issues. At the institu-
tional level, the reports favored deregulation and sometimes advocated
the adoption of universal banking.'
During the 1980s the World Bank shifted its emphasis from institution
building, often within a nationalized framework, toward greater reliance
on market forces within a systemwide approach to finance. Since 1978
the World Bank has commissioned many comparative studies of financial
systems in developing countries with the objective of developing broader-
based lending programs. Toward the end of the 1970s and during the
1980s, therefore, the World Bank began to replace project loans with
apex and structural adjustment loans aimed at development of the fi-
nancial sector.
The first wave of World Bank financial sector reports of the late 1970s
and early 1980s stressed macroeconomic and sectoral issues. These issues



104 ISSUES OF POLICY REFORM
included price stability, abolition of interest rate ceilings, and rational-
ization or elimination of directed credit programs and discriminatory
taxation of financial intermediation, whether directly through taxes on
transactions and on loan and deposit rates or indirectly through reserve
requirements, forced investments in low-yielding government bonds, and
minimum requirements for providing subsidized credit to priority (high-
risk) sectors (see Hanson and Neal 1985). Reports gave only very brief
attention to the institutional aspects of the financial sectors being ana-
lyzed. Although these early reports often stated that bank regulations
and supervision needed improvement, that was about as far as the rec-
ommendations went.
After the disastrous results of the experiments with financial liberal-
ization in Latin America and the increasing financial fragility worldwide,
the second wave of World Bank financial sector reports from about 1983
on has devoted much more attention to microeconomic, institutional,
and regulatory issues. There is also more recognition of macroeconomic
stability, particularly fiscal responsibility, as a prerequisite for successful
financial development (Hanson and Neal 1985). However, there has been
virtually no discussion of the second-best dilemma that arises when con-
trols in some areas, such as interest rate ceilings, are dismantled but
controls in other areas, such as limitations on asset portfolios, are
strengthened. Indeed, Chile may have been following the only set of pol-
icies with any theoretical underpinnings when it abolished virtually all
controls over every financial activity. Unfortunately, the Chilean gov-
ernment was unable to persuade the private sector that its hands-off
policy also implied no bailouts, and credibility in this regard was de-
stroyed when the government stepped in to rescue a failing bank as early
as 1977 (Diaz Alejandro 1985).
Recent experience with financial reform and liberalization indicates
just how important market structure and regulation are in determining
the outcome of such programs. In particular, cartelized or highly con-
centrated banking systems have not responded as anticipated to the ab-
olition of interest rate ceilings. Bank associations simply assume respon-
sibility for establishing appropriate interest rates for their members when
governments relinquish this function. Financially repressed markets tend
to be cartelized or highly oligopolistic in the first place.
In summing up the experience with liberalization policies in the South-
ern Cone countries of South America, Blejer (1983) concluded that many
domestic markets, but particularly the financial markets, given their size,
organization, and high degree of segmentation, reacted in an oligopolistic
manner to the opening of the economy. Since not all financial agents had
the same access to international borrowing, borrowers without direct
access to international markets had to obtain credit from domestic fi-
nancial intermediaries that could borrow abroad. Under these conditions,



Financial Sector Reform 105
the domestic interest rate is determined not solely by the foreign interest
rate, the expected rate of change in the exchange rate, and any risk pre-
mium, but also by domestic market conditions, including the domestic
demand for and supply of credit, the structure of the domestic financial
system, and the state of inflationary expectations.
The only part of Gelb and Honohan's chapter with which I disagree
is its apparent belief that capital markets can be developed under ad-
justment lending programs. Although an effective capital market may
indeed result in greater mobilization of resources and more productive
investment, the World Bank has yet to demonstrate that it has the key
to developing such a market.
Not only is the developing world littered with insolvent DFIS, it has all
too few successful capital markets: markets in which interest rates are
freely determined by the interaction of supply and demand are few and
far between. To establish such markets, a logical progression in financial
development programs is crucial. The monetary authority can support
the market determination of interest rates in a few easy stages. It is im-
portant to start the process by introducing short-term financial assets
with no risk other than a small amount of interest risk. Prices must be
determined in the market through the free interplay of demand and sup-
ply. A good place to start is with the establishment of a treasury bill
market, a market for central bank negotiable certificates of deposit, and
interbank money markets. Only after experience has been gained in these
simple, low-risk, open markets does it make some sense to turn to the
development of equity markets.
It strikes me that the World Bank, in its early and ongoing concern
with equity markets in developing countries, has been guilty at times of
encouraging financial sectors to run before they can walk. Surely, equity
markets require the most diligent prudential supervision of all. Possibly
the greatest challenge is to identify what can be done in the absence of
adequate accounting and auditing skills. Given the enormous cost of
hiring multinational accounting firms to audit financial institutions on
a case-by-case basis, more concern with the training of counterparts
seems vital.2
Efforts to reform financial sectors have focused mainly on liabilities.
More attention needs to be given to increasing assets. The repatriation
of funds needs to be encouraged; policies are also needed that will help
keep money at home. Innovative methods are needed to improve the
health of financial institutions. Perhaps some version of debt/equity
swaps can be brought down to the local level. Financial adjustment loans
might best be used to help developing country governments remove the
most deleterious of the discriminatory taxes that have resulted in negative
effective protection of financial intermediation (see Fry 1989). Although
Gelb and Honohan point out that a reduction in financial sector taxation



106 ISSUES OF POLICY REFORM
is often a priority in financial reform, they say little more about this
subject. The World Bank is now providing fiscal support for such reforms
in its new financial sector adjustment lending program. Perhaps recent
World Bank initiatives in this direction will be more successful than some
past efforts.
Notes
1. For example, a report prepared jointly by the International Monetary Fund
and the World Bank recommends universal banking for the Philippines (World
Bank 1980).
2. In an earlier search for a temporary remedy to this problem, I suggested
that the Belgian external audit system warranted serious consideration.
References
Blejer, M. I. 1983. "Liberalization and Stabilization Policies in the Southern Cone
Countries: An Introduction." Journal of Interamerican Studies and World Af-
fairs 25 (4) November:431-44.
Diaz Alejandro, Carlos F. 1985. "Good-Bye Financial Repression, Hello Finan-
cial Crash." Journal of Development Economics 19 (September-October): 1-
24.
Fry, Maxwell J. 1988. Money, Interest, and Banking in Economic Development.
Baltimore, Md.: Johns Hopkins University Press.
1. 989. "Managing Deficit Finance." Economic Development Institute,
World Bank, Washington, D.C.
Hanson, James A., and Jaime de Melo. 1985. "External Shocks, Financial Re-
forms, and Stabilization Attempts in Uruguay during 1974-83." World De-
velopment 13 (8) August:917-39.
Hanson, James A., and Craig R. Neal. 1985. Interest Rate Policies in Selected
Developing Countries, 1970-1982. World Bank Staff Working Paper 753.
Washington, D.C.
Hanson, James A., and Roberto de Rezende Rocha. 1986. "High Interest Rates,
Spreads, and the Costs of Intermediation: Two Studies." Industry and Finance
Paper 18. World Bank, Washington, D.C.
Levitsky, Jacob. 1986. "World Bank Lending to Small Enterprises: A Review."
Industry and Finance Paper 16. World Bank, Washington, D.C.
Long, Millard. 1983. "Review of Financial Sector Work." Industry Department,
Financial Development Unit, World Bank, Washington, D.C.
Morris, Felipe. 1985. India's Financial System: An Overview of Its Principal
Structural Features. World Bank Staff Working Paper 739. Washington, D.C.
Veneroso, Frank. 1986. "New Patterns of Financial Instability." Industry De-
partment, World Bank, Washington, D.C.



Financial Sector Reform 107
World Bank. 1980. "The Philippines: Aspects of the Financial Sector." Wash-
ington, D.C.
. 1985. "Financial Intermediation Policy Paper." Industry Department,
World Bank, Washington, D.C.



6 Reform of Public Enterprises
John R. Nellis
THE PERFORMANCE of public enterprises in developing countries has not
generally lived up to the expectations of their creators or funders. Public
enterprises have not generated the anticipated rates of return on equity
invested. Nor, all too often, have they attained their noncommercial ob-
jectives-with regard to employment generation, technology transfers,
or regional development, for example. Nor, frequently, has performance
reached the anticipated levels of economic productivity and efficiency.
Performance varies considerably from region to region, with public en-
terprises in Asian countries showing fewer problems than those in Latin
America, which, in turn, generally perform better than enterprises in Sub-
Saharan Africa. Every country possesses a number of profitable enter-
prises (although many are exploiting a protected or monopoly position),
and some of these enterprises are sound even when measured by the
strictest economic criteria. But there are still too many money-losing,
inefficient public enterprises in developing countries.
In the case of individual enterprises, the details of poor performance
have been known for some time. Numerous attempts have been made
to resolve problems, frequently in conjunction with World Bank project
lending. Despite perceptible improvements in many firms in technology,
training, and managerial procedures, however, performance levels-
measured by the traditional criteria of profitability, or rate of return on
investment, and self-financing ratios-have all too often stagnated or
worsened.
Reviews of project lending for public enterprise reform have concluded
that many of the factors contributing to poor performance were external
to the firm and so outside the control of management: deficient ma-
croeconomic policy, uneconomic pricing regimes, lack of clearly defined
objectives for individual enterprises or subsectors of enterprises, conflicts
and shortcomings in the legal and regulatory framework, too easy access
to the banking and credit system, weak financial structures, ineffective
or interfering monitoring systems, and chaotic budgetary relations. Given
these common circumstances, even dynamic and competent management,
working with high-quality equipment, may be incapable of generating
reasonable returns.' Conversely, good financial returns are sometimes a
108



Reform of Public Enterprises 109
function of distortions and are not due to the efforts of management.2
These realizations constituted the fundamental justification for the in-
clusion of public enterprise sectors in adjustment operations. Since piece-
meal and particularistic reforms had proven to be insufficient, the logical
alternative was to attempt systemic revision of policy and procedures.
Reforms Attempted through Adjustment Programs
The first World Bank-sponsored attempt to reform public enterprises in
a systemic, cross-sectoral manner was the Senegal parapublic project,
which became effective in 1978, three years before the first Senegal struc-
tural adjustment loan (sAL). Between 1982 and 1984 a few other World
Bank-assisted reforms of public enterprises took the form of loans to
fund technical assistance outside the framework of a structural or sectoral
adjustment operation. These programs were in the Republic of Congo,
Madagascar, Mali, Mauritania, and Peru. All other cases have been an
integral part of, or closely associated with, adjustment lending. As of
March 1988, the number of adjustment-related operations totaled 101-
51 in SALS and 50 in sectoral adjustment loans (SECALS).3 Public enterprise
reforms have been an element of almost every adjustment operation,
structural or sectoral.
In keeping with perceived deficiencies in performance, there have been
many more actions in Sub-Saharan Africa than elsewhere: through 1988,
twenty-six SALS in seventeen different countries. All twenty-six had public
enterprise components.
The situation for SECALS is more complicated. Those specifically and
uniquely devoted to public enterprise reform have been approved in seven
countries: Benin, Jamaica, Mauritania, Morocco, Niger, the Philippines,
and Tunisia. Similar projects, commonly referred to as PERLS (public en-
terprise restructuring loans), are in preparation in at least six more coun-
tries. But public enterprise reforms-of policy, pricing, relations between
the state and the public enterprises, staffing, training, and the financial
and managerial structure of individual firms-have also figured in a large
percentage of the other forty-four SECALS approved between 1980 and
March 1988. In many of these loans, especially those to the agricultural
and industrial sectors, the public enterprise components have been prom-
inent.
The reform packages have some basic similarities.4 Cross-sectoral im-
provement activities have fallen under three general headings: (1) reforms
of the macroeconomic policy and financial frameworks affecting public
enterprise performance; (2) reforms of the institutional framework (the
ways in which the government guides, supervises, and evaluates public
enterprises); and (3) issues of divestiture.



110 ISSUES OF POLICY REFORM
In addition, more than half of all public enterprise components of
adjustment loans have contained elements aimed at the restructuring and
improvement of specific enterprises. By bringing some of the most im-
portant and sensitive reforms of individual enterprises into a large and
visible adjustment operation, it was hoped to increase the likelihood that
painful reforms-such as staff reductions, financial restructurings, plant
closures, and price increases-would actually be implemented. Including
enterprise-level reforms as conditions of an adjustment operation appeals
to World Bank project officers and frustrated government staff alike, but
the practice does add to the number and complexity of conditions. Im-
provements in specific enterprises have figured in more than forty op-
erations; in many of the more recent and important ones the resources
devoted to rehabilitating enterprises have been equal to or greater than
the funds spent on cross-sectoral reforms.
Reform of the Policy Framework
An adjustment operation, especially one with a public enterprise com-
ponent, inevitably raises questions of the proper role, or the redefinition
of the role, of the state. Most public enterprise-related adjustment pro-
grams, and all those approved after 1984, propose some actions-often
backed up by conditionality-related to this broad policy issue. Actions
have included (1) analyzing and clarifying which economic activities are
appropriate for state involvement; (2) classifying enterprises on the basis
of whether they should be retained, restructured, sold, or closed; and (3)
identifying specific enterprises to be sold or closed, subsectors in which
privatization will take place, and measures to facilitate or support the
sale of enterprises. Countries with reform programs range from those
which are still undertaking studies and policy reform preparatory to di-
vestiture (Morocco and Turkey, for example) to those which have moved
into the action stage of sale and liquidation (Guinea, Jamaica, Niger,
Panama, and Togo, for example).
A modest number of adjustment operations have addressed the elim-
ination of monopolies and monopsonies in public enterprises. This has
been the subject of conditionality in ten SALs and six SECALS. More rare
have been attempts to place public and private enterprises on an equal
regulatory footing, although the Niger SAL abolished tax exemptions for
public enterprises. Trade liberalization measures are much more com-
monly advocated; they are present in about half the SALs with public
enterprise components and are either directed toward or have an effect
on industrial and manufacturing public enterprises. Another general pol-
icy area is that of legal and legislative reform, which has figured in the
public enterprise adjustment operations in Turkey and throughout fran-
cophone Africa. The main tasks are to clarify the juridical framework,



Reform of Public Enterprises 111
decrease the number of legal types of enterprises, change decrees to elimi-
nate or lighten control procedures and increase managerial autonomy,
create the legal texts necessary to implement privatization, and rationalize
the usually cluttered scene with regard to supervisory agencies.
Public enterprise components of adjustment operations in Benin,
Ghana, Madagascar, Morocco, Panama, Senegal, and Turkey and in
most other countries have set pricing formulas that, in most instances,
are aimed at rationalizing the way prices are set and periodically adjusted.
In six francophone countries pricing issues have been addressed through
support of contract plans that specify the mutual obligations and rights
of the owner-the state-and the enterprise for a three- to five-year
period and usually specify pricing arrangements for the firm.' Loan con-
ditionalities often call for studies on procedures of price setting in the
natural monopoly, service-providing public enterprises; on methods and
timing for price adjustments; and on mechanisms to improve billing and
collection procedures.
Labor policies have featured prominently in the public enterprise com-
ponents of adjustment operations. Excess staffing is a common problem
in public enterprises. Substantial layoffs can cut costs considerably, and
the resulting savings in the wage bill can provide resources to motivate
the remaining work force. Reductions in staff and changes in employment
rules are conditions in twelve SALS and in a larger number of SECALS. Most
countries have approached work force reductions with caution, however,
because of concern about the welfare of affected employees and the fear
of provoking labor unrest or organized political resistance. In response
to these legitimate concerns, several recent reform programs-in Benin,
Congo, Ghana, and Mali (and several in preparation)-have created
training and redeployment funds or supported forms of severance pay.
Reforms in financial policy have also become prominent in public en-
terprise components of adjustment operations. Many governments have
tried to reduce the burden of public enterprises on their budgets by simply
declaring an end to direct subventions from the state. In many cases the
public enterprises then had to resort to the national and international
credit systems, which, in turn, began to suffer from holding the nonper-
forming debts of those enterprises. Congo and Senegal have attempted
to analyze this situation, to propose reforms, and to improve the analyses
by domestic banks or reviewing ministries of investment proposals by
public enterprises; similar steps are under consideration in most proposed
public enterprise reform operations in Africa. Strengthening the scrutiny
of public enterprise expenditure and borrowing, increasing interest rates
on loans to public enterprises to market levels (or at least to the rates
paid by the domestic private sector), and reducing automatic access to
credit are reform measures in at least fifteen SALS and as many sector
operations in all four regions of the developing world.



112 ISSUES OF POLICY REFORM
Another major financial problem is that of cross-debts and arrears.
Settlement of arrears, and the installation of mechanisms to prevent their
recurrence, is the key issue in the Moroccan PERL. An emphasis on clar-
ification, settlement, and prevention of future cross-debts between public
enterprises and government and among public enterprises themselves is
a cornerstone of most adjustment loans and credits in Sub-Saharan Af-
rica; examples can be seen in Congo, Ghana, and Niger. A related issue
is the strengthening of the budgeting forecast and allocation system to
allow governments to rationally assess and meet their financial com-
mitments to public enterprises; this is a continuing theme of adjustment
reform in Senegal and elsewhere-again particularly in Africa.
Reform of the Institutional Framework and Rehabilitation of
Enterprises
Relations between the public enterprises and the state have received a
great deal of attention in every PERL and in most sALs as well. A large
number of SECALs have also addressed this issue. This general theme has
been disaggregated into many parts.
More than twenty-five public enterprise operations have dealt with the
creation or improvement of control and guidance agencies in the central
managing organization (the president's or prime minister's office, the
ministry of finance or of planning, or the central bank); in the technical
ministries, particularly industry and agriculture; or in holding companies.
Although several African countries have created ministries of state en-
terprises during the period of World Bank-assisted reform, this has not
been at the request of the Bank or a requirement of the adjustment loan
except in the case of Togo. Indeed, the World Bank has generally opposed
the creation of specific ministries of this nature on the grounds that, in
the ministerial competition for resources, these entities become uncritical
defenders of public enterprises rather than stimulators and evaluators of
their performance. Training programs for government supervisors and
evaluators of public enterprise operations, in both the technical and cen-
tral ministries, are often included in public enterprise operations.
Closely related is the creation or strengthening of information and
monitoring systems to provide the data on public enterprises needed to
guide and evaluate their performance. This is done both at the level of
the firm, to improve internal management, and at the subsectoral or
national level, to provide essential information for holding companies or
ministerial reviewers. Many of the more than twenty-five operations that
attempt to deal with this issue are also working to improve state guidance
and control agencies.
A common reform is to enhance the power and autonomy of the boards
of directors. Efforts focus partly on enlarging a board's legal powers and



Reform of Public Enterprises 113
responsibilities and partly on changing its composition from a majority
of civil servants to a majority of independent individuals, preferably rep-
resentatives of the private sector, creditor agencies, user groups, chief
suppliers, technical experts, and other interested groups.
One very important issue has been addressed more through studies
than through direct action. Performance evaluation that is based on a
comparison of results with the initial objectives is often mentioned and
is consistently regarded as desirable. But performance evaluation of this
type requires that information systems and competent monitors be largely
in place. Thus, activities in this field have been mainly of a preparatory
nature, except in Gambia, Ghana, Korea (where an innovative perfor-
mance evaluation system was supported in the single adjustment oper-
ation), and in seven francophone countries where performance contracts
have been or are being installed.
Rehabilitation of individual public enterprises has been equally mul-
tifaceted. A more or less standard package of diagnosis and prescription
has emerged that can be applied in all developing regions. Diagnostic
studies of the problems of a key enterprise, group of enterprises, or the
sector as a whole have been attempted in more than fifteen operations.
Management and financial audits frequently follow the diagnostic studies
(in a slightly smaller number of operations), followed in turn by recovery
plans or rehabilitation programs. (These actions have been called for
directly in the adjustment operations, or they have been included in ac-
companying technical assistance loans.) A common theme among coun-
tries and regions is the weakness of accounting practices within firms;
adjustment operations or the projects that parallel them propose training
or technical assistance in many instances. In francophone countries, and
increasingly in anglophone Africa and Latin America, contract plans, or
performance agreements, are being initiated and installed in public en-
terprises working in largely noncompetitive markets. Training for the
personnel of firms has been featured in sixteen adjustment operations
and is a prominent part of most of the public enterprise restructuring
loans in preparation. Finally, capital support and restructuring has been
advanced in only five operations, four in Sub-Saharan Africa and one in
Turkey.
Divestiture
Divestiture refers to the full range of mechanisms by which the state
reduces its direct involvement in the economy: the full or partial sale or
transfer of ownership, the sale of assets, leasing arrangements, contract-
ing out, and liquidation (in its various forms). Close to forty adjustment
operations have contained divestiture elements, which in most cases con-
centrate on the full or partial sale of ownership. The sale of a specifically



114 ISSUES OF POLICY REFORM
named enterprise or set of enterprises has been clearly set out in loan or
credit conditions in only five adjustment cases: Ghana, Guinea, Jamaica,
Panama, and Senegal. The closure and liquidation of specific public en-
terprises has been a subject for conditionality in seven countries, six of
them in Sub-Saharan Africa: Benin, Burundi, Central African Republic,
Guinea, Niger, Panama, and Senegal.
A more common conditionality for divestiture is that governments
prepare, and discuss with the World Bank, studies or action plans on the
classification of enterprises and the future of problem enterprises. The
best course of action is frequently a matter of dispute, with the World
Bank generally favoring divestiture and the government favoring restruc-
turing. Adjustment operation conditions often require investigation of or
progress on privatization, but not the sale of a particular enterprise.
Recurring phrases state that the government "will initiate a divestiture
program"; "will agree with the International Development Association
on a strategy for rationalization and disengagement of the state enterprise
sector"; "will implement a divestiture program"; "will accelerate its pro-
gram on divestiture"; and that the government's progress on divestiture
"will be satisfactory to the Bank." Such phrases-almost always in the
future or conditional tense-are common, occurring in the documents
for at least twenty-five adjustment operations, roughly half of them in
Sub-Saharan Africa. The intention, of course, is that the future tense will
change to the present as the scheduled divestitures occur. (There are
several sound financial and economic reasons for this vagueness, which
are discussed below.) Hard conditionality on divestiture is concentrated
in the smaller and poorer countries of Africa.
Other divestiture actions have been scattered; they include leasing ar-
rangements (prominent in Togo) and management contracts (common
in hotel sectors). There has been relatively little experimentation in ad-
justment operations with such changes as the contracting out to private
suppliers of services performed by public enterprises, although such steps
are often suggested in consultants' reports prepared in the early stages.
Assessment of Results
Public enterprise aspects of adjustment lending pose at least two problems
for the evaluator. First, many of the objectives of these reforms, partic-
ularly those in the institutional realm, are qualitative and not easily sub-
ject to precise measurement. Second, even in cases in which the desired
outcome can be precisely stated, several years of effort and implemen-
tation will be required before any assessment of achievement can be
meaningful. There are also elements that are both qualitative in nature
and long term. With regard to the qualitative objectives, one is thrown
back on impressionistic assessments of whether progress has been sat-



Reform of Public Enterprises 115
isfactory. Although these judgments can be of some benefit, they are
obviously less than ideal. The problem of assessing the long-term objec-
tives is equally intractable; not many of the public enterprise-related
adjustment operations have been in place long enough to permit a judg-
ment concerning which targets have been hit and which missed.
Therefore, in this chapter, an attempt is made to assess the experience
of nine countries that meet the following criteria: (1) public enterprise
reforms have figured prominently in adjustment operations, (2) the re-
forms have been in place long enough to produce at least some discussable
(if not always measurable) results, and (3) the results have been reviewed
by World Bank staff. These countries are C6te d'lvoire, Ghana, Jamaica,
Morocco, Niger, Panama, Senegal, Togo, and Turkey. This list also is
representative of public enterprise-related adjustment operations: just
over half the countries are in Sub-Saharan Africa, and most are small
countries.
Types of Reform Attempted
The type and frequency of public enterprise reforms attempted in the
thirty-one adjustment operations in the nine countries are presented in
table 6-1. The figures require some interpretation. With regard to di-
vestiture, the twenty-nine operations do not equate to twenty-nine dif-
ferent programs. Privatization actions started by the first adjustment loan
in Togo, for example, were pursued and deepened by later loans, and
the same is the case for Jamaica, Niger, and Senegal, all of which have
had at least two adjustment operations. These countries needed subse-
quent adjustment operations to reinforce initial efforts, to actually launch
a divestiture after initial preparation, or-commonly because of an over-
ambitious original timetable-to provide additional time to prepare and
complete the arrangements. This said, it is still true that divestiture re-
forms were far and away the most common efforts in these thirty-one
operations.6 They were followed, in roughly equal measure, by pricing
reforms, institutional strengthening, and rehabilitation of specific enter-
prises.
Table 6-1 lists the same wide range of reform activities reviewed earlier
in this chapter. But whereas the overall review indicated that the emphasis
was more often on preparation for divestiture than on actual sales or
closures, in the smaller sample the opposite is clearly the case: imple-
mentation actions are almost three times more numerous than prepa-
ratory activities. Moreover, the conditionality called for in the divesti-
tures in the sample tends to be precise and action oriented, in particular
in Jamaica, Niger, Panama, and Togo. In addition, many of the public
enterprise reforms initiated in this particular group of countries tend to
be concrete, measurable, and monitorable. Enterprises are either sold or



116 ISSUES OF POLICY REFORM
Table 6-1. Type and Frequency of Public Enterprise Reforms in
Adjustment Operations in Nine Sample Countries
Number of operations
in which
Reform                                                   reform is stressed
Divestiture
Implement full or partial sale, lease, or liquidate          21
Examine cases, classify, set policy, prepare                   8
Pricing                                                         15
Institutional framework                                         15
Rehabilitation and restructuring in specific firms              15
Reduction in work force                                        12
Financial audits of troubled firms                              11
Preparation of policy statements                                10
Settlement of cross-debts and arrears                            9
Contract plans                                                   7
Reduction in transfers and subventions                           6
Management audits of troubled firms                             4
Financial restructuring                                          *
Training                                                         *
Export promotion                                                 t
Elimination of monopolies                                        a
Revision of legal framework                                      *
Specification of rate of return on assets                        *
Preparation of corporate plans                                   *
Reform of management pay scales                                  a
Increase in competition                                          a
Subsidies and credit
* Mentioned in three or fewer adjustment operations.
Source: World Bank data.
closed or they are not; prices are either increased by the assigned date
and by- the stipulated percentage or they are not; workers are either
dismissed from an enterprise or they are not (although whether they stay
dismissed is another matter); and the same is true for the completion of
financial audits, contract plans, rehabilitation plans, and the like. Thus,
the operations in this sample provide a relatively tangible record for
review.
Results of Reform Efforts
DIVESTITURE. Reforms related to the privatization and liquidation of
public enterprises or to the preparation for divestiture were featured in
adjustment operations in eight of the nine countries. Adjustment loans
signed in Turkey did not deal directly with divestiture (the energy SECAL
contained a minor reference to the process), but large-scale privatization



Reform of Public Enterprises 117
was the leading element in a PERL, the three-year preparations for which
were halted at the request of the Turkish government-without any en-
terprise having been sold. In contrast, in Jamaica, Niger, Panama, and
Togo all of the divestiture programs either stipulated or suggested have
been carried out or are in an active stage of preparation.
In Panama, for example, even though the disbursement of the second
tranche of the second structural adjustment loan was canceled at the
request of the government, five of the scheduled seven divestitures were
completed. In Jamaica the government was committed under the public
enterprise restructuring loans to divest its public enterprise portfolio by
a value of 150 million Jamaican dollars by December 1987; it far ex-
ceeded the goal by divesting to the value of 500 million Jamaican dollars.
Thirty-two of the forty-five firms on the list for divestiture were fully or
partially privatized, leased, or dissolved by November 1987. In Togo
four public enterprises were leased to private operators. The best known
of these four was the STS steel mill, a financial and political success but
a dubious economic proposition. The management of a clothing firm was
handed over to a Togolese in the private sector-one of the first examples
of a management contract being given to an African entrepreneur. Two
public textile firms were merged and a sale was negotiated with a Korean-
American group, but the sale was seriously delayed at the last moment.
Five other public enterprises were liquidated and their assets sold. Almost
twenty other commercial and industrial public enterprises are on the sales
block; arrangements for several are quite far advanced. In Niger, as of
December 1987, three public enterprises had been fully privatized, eight
partially privatized, and four liquidated, while divestiture work was
under way for eight others. In many of the cases cited there were financial
and economic drawbacks, some of which are reviewed below. With re-
gard to the divestiture conditions of adjustment agreements with these
four countries, however, the conclusion is that the conditions have sub-
stantially been met.
In Ghana a loan condition was that divestiture be initiated in at least
ten public enterprises, one of which was specified by name. The objective
was to put five companies on the market and to prepare five inactive
public enterprises for liquidation. A March 1988 World Bank report
notes that some preparatory work has been completed but that "no de-
cisions have yet been taken." In a statement that could easily be applied
to a number of countries and divestiture cases, the report argues: "This
slow progress need not be construed as lack of commitment or effort on
the part of the Government, but rather as reflecting the inherent diffi-
culties associated with the implementation of such a program."
In Morocco the PERL and various SECALs have emphasized studies and
preparation rather than overt action (although the agricultural SECAL
required some action on a liquidation, the status of which is unknown).



118 ISSUES OF POLICY REFORM
In the case of the PERL, the World Bank reasoned that the policy frame-
work and the openness of public enterprise operations needed to be im-
proved first. Only after demonstrating that its concern was with efficiency
and that ownership issues were secondary (as one of many mechanisms
to improve efficiency) would the Bank broach the issue of divestiture.
The Moroccan PERL thus called for consideration, studies and classifi-
cations, and a draft law on privatization, all of which were achieved on
schedule. (The king of Morocco, in his speech opening Parliament on
April 8, 1988, announced a major privatization initiative, and a law
proposing the sale of most Moroccan public enterprises is now being
debated.) The divestiture question will receive more concrete attention
in the proposed second PERL.
Senegal has moved very slowly on divestiture, exhibiting the full range
of difficulties in the privatization process: reluctance to sell to foreigners,
fear of a concentration of ownership of divested firms, difficulty recon-
ciling market value with book value, and problems finding an appropriate
and open sales mechanism in the absence of an adequate equities market.
Thus, although studies and preparation for divestiture have figured in
adjustment lending to Senegal since 1980, not until 1987-88 was more
demanding conditionality introduced. Results have been modest. In Oc-
tober 1987 the Special Commission on the Disengagement of the State
published a list of ten public enterprises in which the share of state own-
ership was to be reduced through the sale of stock issues. In only two
of the ten would ownership pass from a public majority to a private
majority; in the rest, the government already had only a minority share,
which was to be reduced or eliminated. The accompanying explanations
left many technical and mechanical questions unanswered. In early 1988
the government published a list of ten additional firms that were to be
put up for sale. But the authorities removed from the first list two of the
larger firms, stating that they needed to be restructured prior to sale. No
buyers have yet been found for any of the enterprises on either list. The
government has managed to liquidate seven small enterprises (that in
total accounted for 1 percent of government equity in nonfinancial public
enterprises) but has not fulfilled its commitment to identify a second list
of public enterprises to be closed. The conclusion is that Senegal is creep-
ing forward on this issue without enthusiasm, at a pace far slower than
anticipated.
In Cote d'Ivoire, conditionality was met by the liquidation of two of
the six sugar firms.
The first lesson learned from these experiences is that the divestiture
process takes longer than anticipated. Divestiture cannot be easily syn-
chronized with rapid loan disbursement. Trying to do so usually only
makes matters worse. For example, specifying a date by which a partic-
ular enterprise must be sold can be counterproductive. The announce-



Reform of Public Enterprises  119
ment provides an incentive for purchasers to wait until the deadline nears
and then offer a low price, while governments will feel obliged to accept
either a low price or an offer contingent on the granting of protection
or competition-reducing distortions. (This was the case in the privati-
zation of a group of firms in Guinea, a country not included in the sample
reviewed here. In that divestiture program, which was a matter of strong
conditionality in the loan, five of the fourteen companies sold were trans-
ferred as monopolies, and another company was given the exclusive right
to import fuel oil, duty free, for fifteen years.)
The second lesson is that although most governments are complying,
albeit with varying degrees of speed and enthusiasm, with the conditions
and suggestions for divestiture, there is as yet little information on the
postprivatization performance of the entities sold. A small feeder airline
privatized in Panama has expanded its fleet and opened new routes. Pri-
vate traders allowed to compete with a food-distributing public enterprise
in Niger are providing more goods at lower prices to a greater number
of outlets, while the public enterprise itself, faced with competition, has
cut costs and is now making a profit. In countries where divestiture and
monopoly-breaking has taken place, there are a fair number of encour-
aging anecdotes of this nature. But anecdotes, no matter how consistent
with the tenets of neoclassical economics, do not amount to a conclusive
demonstration of a generalized superior capability of the private sector
to put resources to efficient use. Future review work must concentrate
on this essential issue.7
A third lesson is that the World Bank has not proven to be particularly
adept at advising governments on the timing or detailed mechanics of
privatization ventures. For example, when Togo received two offers for
its textile enterprises, the Ministry of State Enterprises asked for advice
in selecting the better offer. The World Bank provided a detailed report
on the pros and cons of the two offers. The report was circumspect but
clearly favored the offer that proposed to bring in more restructuring
capital, create more jobs in Togo, and produce for export rather than
for the protected domestic market. From an economic point of view, that
offer was clearly superior to the other one, and it was duly accepted by
the government. But the foreign investors proved unable to raise the
stipulated amount of restructuring capital, and the sale was delayed. This
led some Togolese to think that the other, smaller, more traditional
offer-which would undoubtedly have gone through more rapidly-
should have been taken. The point is not that the World Bank's economic
advice was bad; on the contrary, it was good. The issue, rather, is that
economic analysis is only part of the deal-making process.
The fourth and final lesson is an obvious one: liberalization should
precede, or at a minimum parallel, privatization. Above all, privatization
must not become an obstacle to liberalization (this could happen when



120 ISSUES OF POLICY REFORM
new, protected, and powerful private owners take steps to prevent further
liberalization). The point is simple: divesting into uncompetitive markets
may do more harm than good. In practical terms, this means taking more
time at the beginning of or during adjustment operations to reform the
policy and regulatory environment and encourage competitive forces so
that privatization can produce the needed benefits of efficiency and pro-
ductivity. Recent actions of the Togolese government, which is in the
forefront of privatizers, are illustrative. The government has become con-
cerned about the nature and pace of the divestiture arrangements so far
concluded and has announced a pause in privatization actions. The World
Bank has supported this pause for evaluation and consolidation and is
now concentrating more on eliminating the remaining policy distortions
and reforming the industrial incentive system than on setting numbers
and dates for the sale of public enterprises.
PRICING, POLICY REFORM, AND REHABILITATION. Pricing issues have fea-
tured prominently in about half the operations under review. In the vast
majority of cases-C6te d'Ivoire, Jamaica, Morocco, Niger, Togo, and
Turkey-the called-for price increases have been achieved. Compliance
has not generally been a problem. Contrary to usual practice, action has
proven easier than contemplation; pricing studies in several countries
have taken much longer to mount and review than anticipated. (It may
well be that calling for a study is the way the World Bank and the bor-
rower government agree to resolve disputes over particularly contentious
pricing issues, which would account for the slow pace of these actions.)
The lengthy and well-studied Turkish experience is particularly inter-
esting in regard to price issues. The Turkish government's primary goal
in public enterprise reform was to reduce the burden on the budget posed
by resource-consuming enterprises. A variety of price adjustments and
increases was enacted, and the financial position of the enterprises im-
proved while government transfers remained even and then declined. But
it has been argued in a World Bank evaluation report that financial im-
provements came about because monopoly public enterprises were al-
lowed to raise prices and that no gains in efficiency can be observed in
the utility enterprises in question, or indeed in the sector as a whole.
In Jamaica the loss-making railway raised its tariffs, laid off a sub-
stantial number of workers, and closed several uneconomic lines, all in
the framework of various SALS. The cases in Jamaica and Turkey are but
two instances in which the financial positions of firms were improved
through pricing revisions but the gains in efficiency are unknown or
unobserved. In the Jamaican railway cost-cutting measures such as staff
reductions and line closures presumably also contributed in the long run
to the return to profitability. But it is not possible on the basis of available



Reform of Public Enterprises 121
data to sort out which element accounted for the greater percentage of
the variance.
The degree of compliance with requirements and requests for staff
reduction has been high, indeed surprisingly so given the politically sen-
sitive nature of the issue. Ghana is the outstanding case. In a two-year
period, the infamously overstaffed cocoa board lost 54,013 names from
its staff list: 25,000 "ghost," or nonexistent, workers were dropped from
the payroll, and a further 29,013 staff were laid off. This action was
taken in response to a loan conditionality that "excess labor" be identified
and eliminated. Early SALS in Turkey had requested or required the same
sort of program; 30,000 workers were laid off from public enterprises
during 1979-82. Staff reductions of a significant if less dramatic nature
have also taken place in C6te d'lvoire, Jamaica, Niger, Panama, and
Togo. (The labor issue was not addressed in the Morocco PERL.) The
layoff program in Ghana is at present running into difficulties, however,
not because political will is exhausted but because of a lack of funds for
the generous severance payments previously agreed to by the public en-
terprises.
The pace and impact of the rehabilitation and restructuring of indi-
vidual enterprises vary greatly from country to country. Improvements
in management and financial performance are reported in five key public
enterprises in C6te d'Ivoire, in several formerly troubled firms in Jamaica,
and across the board in Turkey. In the Jamaican case the assertions of
progress are backed by data on the physical and financial improvements;
in the Turkish case the improved macrofinancial situation is taken as a
proxy for improvements within firms. In Sub-Saharan Africa much of
the rehabilitation work has been diagnostic and preparatory: financial
and management audits, consultants' reports and proposals, improve-
ments in the capacity to review investment proposals, training programs,
and the like. As with many other reform measures, particularly in Africa,
the proposed schedules for rehabilitation and restructuring activities have
consistently underestimated the time needed-further confirmation of
the poor fit between the time frame required to make improvements in
efficiency and the disbursement schedules for adjustment loans.
Inasmuch as World Bank-supported public enterprise reform has been
under way for ten years in Senegal, there is a clear record to evaluate,
and indeed one can point to improvements associated with adjustment
operations: in financial and management information systems, in setting
objectives (through contract plans, discussed below), in pricing policies,
and especially in accounting and auditing. But the economic impact of
these reforms is less easy to discern. The sector continues to accrue fi-
nancial losses, and the trend is downward. In 1982-84 the sector lost
an average of 4 billion CFA francs a year; in 1986 public enterprises lost
7.8 billion CFA francs. (A significant part of the loss is due to price sup-



122  ISSUES OF POLICY REFORM
ports to farmers paid through public enterprises for rural development
and so cannot necessarily be attributed to operating inefficiencies.) Pre-
sumably, the counterfactual argument applies: losses would have been
greater without the adjustment-associated reforms. There are more en-
couraging signs as well: debts of the public enterprises have declined over
the past few years, and operating subsidies to public enterprises have
been reduced-by 5 percent in 1986 and by a further 15 percent in the
1987-88 budget.
Senegal has had extensive experience with the performance agreement,
or contract plan. The results have been less than encouraging. Goals have
been set, the mutual responsibilities and commitments of the two con-
tracting parties clearly set out, and performance measures clearly estab-
lished. In the financial field, pricing regimes have been specified, invest-
ment programs stated, noncommercial operations costed out and
compensation arranged for, and operating subsidies set. But in case after
case the government has proven unable or unwilling to honor the finan-
cial commitments agreed upon, and the agreements have had to be re-
peatedly and extensively revised or have fallen into abeyance. Experience
with the device in other countries (C6te d'lvoire, Niger, and Togo) has
been too short to permit an assessment. The first months of experience
with the contract plans negotiated in Morocco (their signing was a con-
dition of the PERL) have been positive, however, and for the first time in
years the government has honored financial commitments to two key
firms.
Some progress has been made in reducing transfers and subventions.
A direct approach was pioneered in Turkey, where a structural adjust-
ment loan specified that in 1983 investments in public enterprises were
to be held to 6.5 percent of GNP and that overall the financing of public
enterprises was not to exceed 5.9 percent of GNP. The actual figures were
7.2 and 6.4 percent, respectively. Nonetheless, overall transfers were sub-
stantially reduced from previous levels. To meet its loan conditions, Sen-
egal managed to reduce operating subsidies to public enterprises by sim-
ply decreeing that total subsidies in year X would be Y percent less than
the previous year.
INSTITUTIONAL REFORM. Reform of the policymaking, supervisory, and
evaluating institutions has absorbed considerable attention and re-
sources in all the countries in the sample. In Cote d'lvoire, Ghana, Niger,
Togo, and particularly in Morocco and Senegal, World Bank-supported
efforts are under way to strengthen the data base and the monitoring and
managerial capacity of the central institutions guiding public enterprises.
A central goal of reform is to shift governmental methods of supervision
away from extensive and rigid controls and approvals and toward a
smaller number of more flexible mechanisms for evaluating performance.



Reform of Public Enterprises 123
Introducing contract plans, reforming boards of directors, lightening fi-
nancial review procedures, increasing both management's latitude and
the competence and capacity of government supervisors, introducing
legal changes that place more enterprises in the more lightly supervised
categories-these are but some of the institutional changes being intro-
duced. For countries other than Senegal, about all that one can reasonably
say at this stage is that the process has been launched and that the first
signs are encouraging.
For example, in Morocco the PERL-supported Directorate of Public
Enterprises (within the Ministry of Finance) is supervising the preparation
of the contract plan and ensuring that the Treasury honors its financial
commitments, drafting legislation on divestiture procedures, hiring con-
sultants to build a sectoral data base and management information sys-
tem, advising a central policymaking committee on public enterprise
strategy in general, coordinating a training program for government and
enterprise staff, supervising the complicated process of clearing arrears,
and conducting studies on relevant questions such as the optimal powers
and composition of boards of directors. Serious and numerous obstacles
remain, but in general the directorate is tackling the problems with dy-
namism and commitment and is disseminating a changed and positive
attitude regarding the need for efficiency within public enterprises. Prepa-
rations for the PERL and the conditionality associated with it have helped
improve the directorate's position, and resources furnished under the PERL
provided consultants and training.
Roughly similar strengthening measures are being supported in Ghana,
Niger, and Togo; early SALs did the same in Turkey, with modestly pos-
itive results. In Senegal SALs and related technical assistance loans to the
Court of Public Enterprise Accounts have turned that body into an ef-
fective, respected organization. The main agency supervising public en-
terprises in Senegal, however, has not so far played its anticipated role-
despite considerable input from the World Bank. Problems have included
rapid turnover in its leadership, the junior status and lack of experience
of its cadre, and its inability to change the budgeting system to better
reconcile government commitments with revenues (the government's fail-
ure to honor contract plans reveals the need for such change). Of course,
one small agency cannot reasonably be expected to overcome single-
handedly the poverty and deficiencies of an entire system; moreover,
recent changes in the organization's management and personnel are ex-
pected to lead to improvements in performance.
Conclusions
Adjustment operations affecting public enterprises have required or re-
quested reforms in many areas: divestiture (in all its many forms); policy



124 ISSUES OF POLICY REFORM
changes, particularly with regard to pricing; measures to improve man-
agement of the resources used by the enterprises, including labor; better
guidance and evaluation by the state; and restructuring and rehabilitation
of firms. So far there has been a variable but generally acceptable degree
of compliance with conditionality. Adjustment operations have under-
lined both the importance and the intractability of the institutional issues
in public enterprise reform: management, guidance, and evaluation of
performance. Progress in these fields is slow and subject to wide variation
from country to country. Across the board, but especially in the insti-
tutional realm, much remains to be done. There is little information on
the performance of firms after they have been divested and disturbingly
little evidence that gains in efficiency have been responsible for the per-
ceived improvement in the financial situation of many enterprises (im-
provement seems attributable largely to across-the-board cuts in invest-
ment and price increases in monopolies). Nonetheless, in the context of
adjustment operations many governments have succeeded in reducing the
flow of resources into poorly performing public enterprises.
More specifically, in the nine countries reviewed in detail (and there
is no reason to believe that this is an unrepresentative sample), the fol-
lowing reforms are discernible over the adjustment period:
* A freeze on new hiring in public enterprises and, in several instances,
a reduction in the number of employees, a sometimes substantial one
* A freeze on the creation of public enterprises and, in many instances,
a reduction in their number through divestiture and amalgamation
into the regular line administrations
* A general but not universal reduction in the budgetary burden of public
enterprises (mainly through price increases, investment cuts, and staff
layoffs)
* Considerable increases in the quantity, quality, and availability of in-
formation on the physical, financial, and economic performance of
public enterprises
* Widespread experimentation with institutional mechanisms to better
monitor and manage state enterprises
* Equally widespread experimentation with private sectors as alternate
users of the resources previously invested in public enterprises.
These are not dramatic advances, but they are steps forward.
Much of the public enterprise reform now being instituted through
adjustment operations can be regarded as preliminary or "brush-clear-
ing" steps; that is, difficult, tedious, and time-consuming measures that
do not individually or immediately turn the firms around but that set
the scene for more dramatic improvements. The complexity and slow
pace of the reform packages being implemented are regrettable but prob-
ably unavoidable. Getting prices right and selling off the industrial and



Reform of Public Enterprises 125
commercial public enterprises that are not natural monopolies were once
viewed as "simple" solutions; they have proven to be either not so simple
or only partial solutions, although both are, and will continue to be, of
critical importance.
The public enterprise reform packages proposed and implemented
under adjustment operations have been based on "best practice," that
is, on the prevailing economic and managerial wisdom and on interna-
tional experience with improving performance.8 As expected, these re-
forms have brought macroeconomic benefits, as measured by a reduction
in the flow of funds from governments to public enterprises, in many
countries (Jamaica, Turkey, and more recently Ghana and Niger) but
not yet in others (Senegal and, surprisingly, Togo, despite its lead role
in privatization). But a reduction of the burden on the budget, no matter
how important in the short run, is not the ultimate justification for re-
form. In the long run, the value of the reforms will be measured by the
extent to which they promote efficiency. On this crucial issue, the data
are not yet sufficient to allow a conclusion.
The World Bank, in its conditions for adjustment lending, should con-
tinue to move away from specifying the number of public enterprises to
be sold by a particular date; this approach is counterproductive. The loan
to Jamaica that specifies divestiture of a stated dollar amount of the state
portfolio is worth further examination by those who favor concrete goal
setting. But in general the divestiture process is too uncertain to allow
the setting of a deadline, especially in the short term. Furthermore, it
appears unwise to make purely institutional reforms a matter of hard
conditionality because the time frame for their design and implementa-
tion exceeds that of even several sequential SALs.
Institutional development projects, running parallel to adjustment op-
erations, may be one solution to this perplexing problem. Some of the
factors involved in meeting such conditionality lie outside the control of
governments; moreover, it is unlikely that institutional reform will make
or break an adjustment operation. Indeed, the World Bank might con-
sider establishing two levels of conditions. Primary conditions would
include a few key, clear, easily monitored objectives, for which lack of
complete compliance would be automatic grounds for bringing an op-
eration to a halt. Secondary conditions would include most institutional
reforms and most public enterprise reforms in general, but nonfulfillment
would provoke milder sanctions that would not halt the program.
Notes
1. The most succinct review of the economic and organizational obstacles to
good performance of public enterprises is found in Shirley (1983). For an analysis
of problems specific to public enterprises in Sub-Saharan Africa, see Nellis (1986).



126 ISSUES OF POLICY REFORM
2. For discussions of this point, see Jones (1982).
3. Since many countries have had several adjustment loans or credits affecting
the public enterprise sector, the total number of borrowers is less than 100.
Moreover, counts vary according to how reform activities are classified; in April
1989 the World Bank's Industry Department concluded that eighty-three op-
erations contained public enterprise actions.
4. This chapter deals with public enterprise reform under adjustment lending;
for a thorough study of the World Bank's experience with public enterprise re-
form as a whole, see Shirley (1989b).
5. For a review of contract plans, see Nellis (1989).
6. This confirms the conclusions of Shirley and Kikeri (1987), which listed
divestiture as the issue most often tackled in World Bank operations to develop
the private sector.
7. A World Bank research project titled "Ex-Post Performance of Divested
State-Owned Enterprises" was launched in August 1989. The study will measure
changes in economic efficiency and fiscal incidence in a sample of eight public
enterprises, six of them in developing countries. It will analyze the factors that
led to perceived changes in performance (if any) and draw lessons for future
divestiture programs. For a preliminary statement on why the World Bank has
sponsored privatization efforts, and with what results to date, see Nellis and
Kikeri (1989).
8. The best survey of issues and reform measures is found in Shirley (1989b).
For case studies of performance evaluation and improvement mechanisms, see
Park (1986) on Korea, and Shirley (1989a) on Pakistan.
References
Jones, Leroy P. 1982. Public Enterprise in Less Developed Countries. Cambridge,
U.K.: Cambridge University Press.
Nellis, John R. 1986. Public Enterprises in Sub-Saharan Africa. World Bank
Discussion Paper 1. Washington, D.C.
. 1989. Contract Plans and Public Enterprise Performance. World Bank
Discussion Paper 48. Washington, D.C.
Nellis, John R., and Sunita Kikeri. 1989. "Public Enterprise Reform: Privatization
and the World Bank." World Development 17 (5):659-72.
Park, Y. C. 1986. A System for Evaluating the Performance of Government-
Invested Enterprises in the Republic of Korea. World Bank Discussion Paper
3. Washington, D.C.
Shirley, Mary M. 1983. Managing State-Owned Enterprises. World Bank Staff
Working Paper 577. Washington, D.C.
. 1989a. "Evaluating the Performance of Public Enterprises in Pakistan."
PPR Working Paper 160. Policy, Planning, and Research, World Bank, Wash-
ington, D.C.
. 1989b. The Reform of State-Owned Enterprise: Lessons from World
Bank Lending. Policy and Research Report 4. Washington, D.C.: World Bank.
Shirley, Mary M., and Sunita Kikeri. 1987. "Survey of the Work of the World
Bank Group on Private Sector Development." Country Economics Depart-
ment, World Bank, Washington, D.C.



Reform of Public Enterprises 127
Comments
Leroy P. Jones
As AN ECONOMIST from academia commenting on a paper by a political
scientist from the World Bank, I take it my task is to oversimplify. Ac-
cordingly, I first collapse John Nellis's fully nuanced and documented
statements into a few bold propositions. Then I attempt to identify un-
derlying causal patterns.
At one level, the chapter gives an exceptionally useful description of
the World Bank's involvement in more than 100 public enterprise reform
efforts. Given the nature and magnitude of the problems in the World
Bank's client countries, this is not a remarkable level of activity. It is,
however, a dramatic change from just ten years ago, when there was
debate over whether public enterprise was even a useful unit of obser-
vation for World Bank work. The Marris (1979) report's conclusion that
it was a potentially fruitful area correlates with the beginning of serious
World Bank work in the area, although no outsider should attempt to
attribute causation in such a complex organism as the World Bank. In
any event, a great deal has been accomplished since then, and Nellis's
chapter superbly documents public enterprise reform embodied in ad-
justment loans.
At another level, the chapter frankly admits the difficulties of trying
to act as a change agent on a sensitive and complex reform in a foreign
country. If I read it correctly, three messages are among the more im-
portant ones.
First, a decade ago we learned that problems observed in individual
public enterprises often originated at the higher level of controlling bod-
ies. Accordingly, effective reforms had to be pursued not only through
project lending but also through policy reform at the sector level via
adjustment lending. This policy work was essential for creating an en-
vironment in which enterprise work could be effective. For example, the
reform of a steel mill is difficult if the manager lacks the autonomy to
hire the right people and reward them for a job well done.
Second, after a decade of working at the sectoral level, we have learned
that many public enterprise problems do not have their genesis within
the sector alone but are rooted in broader imperfections in markets and
political systems in which the sector functions. To return to the example
of the steel mill, it is not necessarily an improvement to give the mill



128 ISSUES OF POLICY REFORM
manager hiring autonomy in a political system that encourages him to
hire politically well-connected individuals (or those who can otherwise
make the manager better off without making the company better off)
and in an economic system with protected markets that allows the man-
ager leeway to indulge such costly behavior.
Third, the most important-or at least the most common-area of
reform has been divestiture, but results to date have not been dramatic,
to say the least. Among the nine countries Nellis selected for intensive
study, Jamaica sold enterprises worth 500 million Jamaican dollars; Togo
leased four public enterprises, shut down five, and has twenty on the
block; Niger privatized three companies fully and eight partially; and
Panama sold off five public firms. These are the successes. Considerably
less happened in the other five countries. Even the successes represented
only a small fraction of the total number of public enterprises in the
country and an even smaller fraction of total sales or value added. The
success of reforms other than divestiture is harder to quantify, but the
impression is one of less than outstanding results.
I will concentrate my comments on the divestiture issue. Given the
limited successes, we need first to determine whether the World Bank
has been doing the wrong things, doing the right things poorly, or doing
the right things well but in a difficult arena in which success necessarily
has a long gestation period.
A clue to what is going on is provided by the observation that success
in divestiture appears to be a positive function of per capita income. That
is, divestiture has gone furthest and fastest in developed countries (West-
ern Europe, New Zealand, and Japan) and slowest in developing coun-
tries, with some newly industrialized developing countries or middle-
income countries in between (Chile, Korea, and Mexico). The correlation
is very far from perfect but is perhaps strong enough to attribute some
explanatory power to the following hypotheses:
• Divesting into competitive markets is unambiguously good. Divesting
into monopoly or otherwise imperfectly competitive markets entails
a tradeoff: one may get the desired improvements in efficiency but at
some loss of market power or bureaucratic and regulatory power.
Letting the private sector exploit a monopoly position is an economic
evil that is not eliminated (although its manifestations are certainly
changed) by either regulation or public ownership. The issue is that
of choosing the lesser evil here-and a legitimate case can be made
for any of the options.
* Market imperfections are, tautologically, far more widespread in de-
veloping countries than in developed countries. This is true in the sense
that enterprises not only sell in imperfect product markets, but also
acquire fundamental inputs in imperfect factor markets: stock markets



Reform of Public Enterprises 129
are thin or nonexistent, and credit and foreign exchange are allocated
by other than competitive forces, as are a variety of other government
permits and protections.
* Therefore, divestiture is liable to be a much slower process in devel-
oping countries. This is so because there is less scope for unambiguous
gains and because it takes time to construct necessary safeguards in
the intermediate cases. Furthermore, legitimate concerns about di-
vestiture have an exponential impact on the gestation period because
they provide a comfortable cloak for illegitimate opposition.
In sum, to paraphrase Adam Smith, the extent of divestiture is limited
by the scope of the market.
Although divestiture in developing countries could and should be used
as an opportunity to reduce some of these market imperfection, it can
often work in the opposite direction. An all too typical story of divestiture
runs as follows: visit country X and be told that enterprise Y is a di-
vestiture success story with vastly enhanced profitability; visit company
Y and be shown how costs have been reduced and demand increased by
a variety of impressive management reforms; visit the political opposition
and be told that profits were turned around primarily because of a side-
condition of divestiture that competing imports be banned for five years.
Was this divestiture a success from the public point of view? Clearly
tradeoffs were involved, and we need a detailed quantitative study (a
commodity, incidentally, which is also in lamentably short supply).
The alternative to divestiture is reform of enterprises within the public
sector. Here again obstacles are substantial, and once again they are, at
least to some extent, larger in developing countries. A common feature
of public enterprise reform is to assign greater autonomy to managers.
If the enterprise faces competitive markets, however, what is really hap-
pening is that one form of control is traded for another: the manager is
more subject to the discipline of market forces and less subject to bu-
reaucratic forces. If, however, the enterprise faces imperfect markets, then
lessened bureaucratic control is not countered by increased market dis-
cipline and the manager or enterprise may be left free to pursue individual
interests rather than the interests of the nation. Granting borrowing au-
tonomy to public enterprises whose lenders know (or think they know)
that the resources of the state ultimately guarantee the loan is at best a
mixed blessing, as the Latin American debt crisis has shown.
Once again, the point is not that increased autonomy is not a desirable
goal. Rather, it is that in developing countries progress toward that goal
may be slow and may vary from decision area to decision area. And
sometimes increased autonomy in one area is accompanied by increased
control (for example, over results) in others. Or, to paraphrase Smith
again, the extent of autonomy is limited by the scope of the market.



130 ISSUES OF POLICY REFORM
The conclusion, then, is that the slow progress reported by Nellis is
largely explicable as a concomitant of attempting reforms that necessarily
have a long gestation period in developing countries. Furthermore, the
obstacles to reform are in part a function of the scope of markets, and
these obstacles will not disappear as long as developing countries are
developing countries. It may be salutary to recall the words of the late
Senator Estes Kefauver, describing the U.S. aid program: "We are fighting
twenty-year problems with five-year plans, two-year people, and one-
year money."
Reference
Marris, Richard. 1979. Survey and Critique of World Bank-Supported Research
on International Comparisons of Real Product. World Bank Staff Working
Paper 365. Washington, D.C.



7 Agricultural Policy
Odin Knudsen
John Nash
IN THE EARLY to mid-twentieth century, agriculture led many countries
into relative prosperity, based on natural comparative advantage and
export-led growth. Beginning in the 1950s, however, industrial devel-
opment was financed through taxation of agriculture, pricing policies
that depressed food and agricultural commodity prices so that wages
could be kept low, overvalued exchange rates and commercial policy that
discriminated against exports, establishment of parastatal agencies to
carry out government policies in all sectors, and heavy borrowing from
abroad. The inflow of foreign capital, invested to a large extent in the
industrial sector, led to further overvaluation of the real exchange rate
and shifted the internal terms of trade against agriculture.
As a result, by the late 1970s many developing countries had large
foreign debts to service, little potential for growth because industrial
sectors had failed to develop as planned, and agricultural sectors that
were functioning poorly after years of hostile policies. In low-income
Sub-Saharan Africa, for example, the agricultural growth rate fell from
2.6 percent a year in 1965-73 to 1.4 percent in 1973-84 (World Bank
1986).
It soon became clear that World Bank project lending in this distorted
policy environment further discouraged agricultural growth by perpet-
uating overvaluation through the provision of foreign exchange and by
indirectly sanctioning the continuation of antiagricultural policies. To
foster macro-level and sectoral reforms while providing countries with
the foreign exchange they urgently needed to carry out these reforms,
the World Bank developed new loan instruments: the structural adjust-
ment loan and structural adjustment credit (hereafter collectively called
SALS) and, after 1983, the sectoral adjustment loans (SECALS). Although
not aimed only at agricultural policies, the early SALS contained conditions
that were intended to correct the policy bias against agriculture. Agri-
cultural SECALS, however, generally included conditions related solely to
the agricultural sector and generally followed measures for macroeco-
nomic stabilization.
This chapter reviews the experience with and lessons of these policy-
based loans. The first section summarizes the types of conditions related
131



132 ISSUES OF POLICY REFORM
to agriculture in SALS and agricultural SECALS. The following sections draw
out policy lessons for the design of agricultural reform programs and
examine issues relevant to the future of lending directed toward agri-
cultural policy.
Policy Reforms in Agricultural Adjustment Loans
Agricultural conditions accounted for about 10 percent of all conditions
in the sample of adjustment loans and for 9 percent of all legal require-
ments.1 This distribution varied little by region except in the case of Latin
America and the Caribbean, where the figures were 5 percent and 6
percent. The importance of these conditions was about proportionate to
their distribution, with 11 percent judged critical to the success of the
loan. The following subsections examine the main policy areas covered
by agricultural conditions (see also tables 7-1 and 7-2).
Pricing
Since prices are the main determinant of incentives for agriculture, pricing
has received special attention in adjustment loans. Almost 60 percent of
SALS and all but one agricultural SECAL have included agricultural pricing
conditions. In Africa 80 percent of SALS and SECALs have contained such
conditions. Input and output pricing conditions have been about equally
common. Almost one-third of the pricing conditions have been judged
to be critical to the loan's success. Implementation of agricultural pricing
conditions has been successful in about 68 percent of the cases, which
is somewhat below the average of 73 percent for all conditions. The
success rate is only 44 percent for Latin America and the Caribbean,
however, which may be related to the unstable macroeconomic environ-
ment in which the price reforms were carried out.
PRODUCER PRICES. Conditions related to producer prices have generally
concerned the need to increase them.2 Some adjustment loans have re-
quired that producer prices be set at specific levels agreed to with the
World Bank. The agricultural SECAL to Mexico linked official domestic
prices for some crops to international prices via a price band. This mech-
anism, if used for all official prices and if complemented by measures to
reduce industrial protection, would ensure both that the prices for ag-
riculture as a whole are approximately correct and that relative prices
within the sector are appropriate. Other adjustment loans, for example,
to Morocco and Jamaica, have required changes in the mechanisms or
formulas for establishing prices. By improving the pricing strategy rather
than merely setting new prices, this kind of condition offers greater hope
for sustained reform.



Agricultural Policy 133
Because of the reluctance of governments to relinquish control of ag-
ricultural markets, only four of the twenty-one SECALs have required the
abdication of the official price-setting authority and withdrawal of the
government from the market. Madagascar was required to refrain from
setting an official price for rice and from procuring rice domestically,
although it was allowed to operate a buffer stock using imported rice to
prevent prices from rising above a trigger level. Ecuador was required
to cease procurement activities directed at establishing minimum prices.
CONSUMER PRICES. Only the agricultural SECAL to Ecuador and some
SALS in Africa have required decontrol of consumer prices. Some others
(for example, in Mexico) have required that global subsidies administered
via low consumer prices be reduced or not be expanded.
INPUT PRICES. Fifteen of the twenty-one agricultural SECALS have in-
cluded conditions related to input pricing. These loans have required
increased prices (or reduced subsidies) for the three most important in-
puts-fertilizer, water, and credit. A few (for example, in Kenya) have
had conditions related to certified seeds, although most of these have
been of a nonprice, institutional nature. None of the loans has required
decontrol of input prices or disengagement of the state from these input
markets, even though sectoral analysis has often identified state involve-
ment as a problem.
TRADE LIBERALIZATION. Since policies that cause domestic prices to
diverge from their international levels must be reinforced by international
trade controls, many loans (eight of the twenty-one agricultural SECALS
and forty-nine of the seventy-nine SALS and SECALS) have required some
liberalization of trade. Of the eight SECALs requiring trade liberalization,
all but two were linked to pricing reforms. Loan conditions have usually
involved removing restrictions (or reducing taxes) on exports of agri-
cultural products or imports of agricultural inputs. Some loans, in the
interest of reducing all barriers to trade, have required reductions in
barriers to imports of agricultural products (for example, in Mexico and
Ecuador). Taken alone, these steps would exacerbate the antiagricultural
bias by removing trade protection for agriculture while leaving industrial
protection in place. But as part of a comprehensive effort that includes
a reduction of trade barriers in the industrial sector and devaluation of
overvalued exchange rates, these measures can open the economy to trade
opportunities. Conditions intended to liberalize trade in inputs have been
rare (only three of twenty-one SECALS).
Institutional Reform
Conditions related to institutional reform appear in 71 percent of the
seventy-nine SALS and SECALS in table 7-1 and constitute the largest single



Table 7-1. Conditions for Agricultural Policy in Seventy-Nine Structural Adjustment and Sectoral Adjustment Loans,
by Region and Policy Area, Fiscal 1980-87
Europe, Middle        Latin America and
Africa                  Asia             East, North Africa          Caribbean                 Total
Agricultural policy         Number    Percent    Number    Percent    Number    Percent    Number    Percent    Number    Percent
Agricultural pricing (input
and output)'                 30         80           3          27          3           33           9          39          45          57
Trade                          16         44           6          55          5           56         23          100          49          62
w    Institutional reform           29          81           5         45           7          78          15           65         56          71
Credit and bankingb            11         31           3          27          1           11           9          39         24           30
Macroeconomic                  20         56           2          18          6           67          10          43          38          48
Public investment budget      21          58           1           9          5           56          10          43          37          47
Environment                     3           8          1           9          1           11           2           9           7           9
Total loans evaluated          36                     11                      9                      23                       79
Number of countries            23                      5                      3                       12                      43
a. Includes interest rate subsidies.
b. Excludes interest rate subsidies.
Source: An internal Word Bank study by Nelson Ridley and Colleen Roberts of seventy-nine SALs and SECALS in fiscal 1980-87 that contained agricultural conditions.



Agricultural Policy 135
category of agricultural sector conditionalities. The two main subcate-
gories of institutional reform are privatization and deregulation of mar-
kets.
PRIVATIZATION. Most analyses of agricultural policies have noted two
problems with the operation of government-owned companies or public
enterprises. First, the legal monopoly power enjoyed by these companies
usually leads to inefficient operation and large fiscal losses. Second, they
are the instrument through which the state carries out the pricing and
regulatory policies that have done so much harm.
Only the first problem has been addressed through adjustment lending.
Seven SECALs have required either that the private sector be allowed to
compete with the public enterprises in a particular market or that some
public enterprises be divested or closed. Four of these conditions covered
product markets and three covered input markets. Some loans, especially
SALS in Africa, have required reorganizing the public enterprises so that
they operate more efficiently. Twelve (all but two of them in Africa) of
a sample of twenty-nine adjustment loans required some divestiture. For
the seven loans for which implementation data were available, imple-
mentation was generally found to be successful.
Virtually no adjustment loans, however, have required that public en-
terprises be abolished or completely privatized or that they withdraw
from direct participation in the market. Because private institutions are
weak in some countries, especially in Africa (partly because their devel-
opment has been suppressed by the public enterprises), public enterprises
cannot be withdrawn precipitately from the market. With few exceptions,
however, withdrawal of public enterprises does not appear even to have
been considered as an explicit long-term objective. Two exceptions were
Ecuador, where the crop-marketing public enterprise was required to
withdraw completely from procurement activities, and Panama, where
the agricultural marketing public enterprise was limited to providing mar-
ket information and managing warehouses.
DEREGULATION. Three of twenty-one SECALS have contained conditions
for the removal of regulatory constraints on markets. Tanzania had to
abolish permits for internal movements of grains. Morocco was required
to dismantle an elaborate system for regulating upstream flour marketing,
including licenses that limited market access and fixed allowances for
distribution margins.
MACRoEcoNoMIc. Macroeconomic policies, particularly exchange
rate policy, have been as important as direct agricultural policies in de-
pressing the sector. Most agricultural adjustment programs, therefore,
have required exchange rate reform of some kind. Forty-eight percent of



136 ISSUES OF POLICY REFORM
the seventy-nine SALS and SECALS had conditionalities for macroeconomic
policy related to agriculture. World Bank President's Reports on twelve
SECALS cite progress on the macroeconomic front preceding the loan. The
most common requirement has been for more rapid devaluations. Some
adjustment loans have also been preceded by other measures to correct
antiagricultural biases in the trade regime. This has usually been done
through SALS or trade policy loans rather than directly through agricul-
tural SECALS (Colombia is the main exception).
Other Policy Areas
Two other policy areas occasionally mentioned in adjustment loan con-
ditionalities are the environment and the alleviation of poverty. A few
loans have required regulation of pesticide use; in Jamaica one required
measures to control erosion, and another required the sale or lease of
government-owned land to improve land use. In general, however, despite
the impact of agricultural policies on the environment, environmental
issues have been ignored in adjustment lending.
Poverty alleviation, too, has been relatively neglected. From fiscal 1980
to fiscal 1987 only one loan (to Ghana in 1984) included a poverty
alleviation program. In fiscal 1988, however, two adjustment loans (to
Mexico and Morocco) included such programs. A loan to Colombia
addressed this issue by targeting counterpart funds for investment in
regions where incomes would fall during adjustment.
Issues for the Design of Agricultural Reform Programs
The World Bank's experience in promoting agricultural reform has been
too short to permit firm conclusions about what does and does not work.
Only one agricultural sector adjustment operation has had a completion
report. Nonetheless, this experience and the general literature on agri-
cultural development suggest some broad guidelines for future loans.
One caveat is particularly important: adjustment is a long-term pro-
cess. Experience with agricultural adjustment has shown mixed results
in the short run. The evidence on the relation between adjustment op-
erations and agricultural growth is ambiguous. Countries that received
adjustment loans increased the average annual growth rate of their ag-
ricultural production from 1.6 percent to 2.8 percent in the three years
following the loan, with a further increase of 3.2 percent in the years
until 1987. Agricultural growth also increased more in these recipient
countries than in comparable groups of nonrecipient countries. But these
averages are deceptive. Of twenty-nine recipient countries, thirteen
showed lower agricultural growth rates between the loan year and 1987
than for the three years before the loan (Balassa 1988b).



Agricultural Policy 137
Chile's experience with agricultural reform suggests that the sequenc-
ing and pace of reforms are important to their effectiveness in the long
term. In late 1973 Chile began, with some World Bank involvement in
the early stages, a long-term program of general economic and agricul-
tural reforms that now appears to have been successful (Knudsen and
Yotopoulos 1985). Before the military takeover in 1973, the government
had administered the agricultural sector through direct ownership of im-
portant production, distribution, and financial enterprises. It also had
employed a multitude of controls and subsidies, including price controls,
quantitative restrictions, interest rate controls, and tariff preferences.
Starting in late 1973 the government introduced changes based on free-
market principles. It deregulated prices, divested itself of state land and
of marketing, distribution, and agroindustrial processing enterprises; re-
duced technical assistance and public expenditures; and attempted to
rationalize the budget, exchange rates, and banking. These sudden and
extensive changes led to pronounced instability in the agricultural sector
in the short term. Agriculture became decapitalized, agricultural debt
accumulated and many loans became nonperforming, and the land
planted to the fourteen main traditional crops declined from 1.2 million
to 0.87 million hectares by 1983. Many farmers who had acquired their
land through agrarian reform lost it.
Beginning in 1983, the government attempted to stabilize agricultural
prices through variable tariffs and government procurement of some
commodities. Retrenchment was not complete, and the government's role
remained limited. The measures taken were successful, and in a period
of continuing macroeconomic stability, the agricultural sector rebounded
and stabilized. Fifteen years after the commencement of the agricultural
reform, the sector is booming, with rapid growth (7.1 percent a year
from 1984 to 1986) of both traditional crops for domestic consumption
and fruits for export.
Chile's experience shows that agricultural reform takes years, that too
much change too quickly can lead to instability, and that temporary
retrenchments may be necessary. Despite a profound commitment to
laissez-faire, the government considered it necessary to give agriculture
special consideration through price stabilization measures and through
protection for wheat, oilseeds, sugar, and milk-markets it viewed as
artificially depressed and unstable because of the policies of developed
countries.
Price and Nonprice Factors in Agricultural Growth
A consistent goal of conditions in World Bank loans for agriculture has
been to move crop and input prices toward their international levels.
Except in cases where initial distortions are severe, however, price reform



138 ISSUES OF POLICY REFORM
alone cannot produce major gains in sectoral or national growth rates
in the short run.3 They must be complemented by nonprice policy mea-
sures, and even then the growth rate will respond only in the medium
to long term.
PRICE REFORMS. In the short run price reforms will improve efficiency
but not growth rates. Because the supply response is high for individual
crops, "getting the prices right" causes a rapid adjustment in relative
production between crops in favor of those that can be produced most
efficiently. Long-run supply responses of individual crops in developing
countries are greater still and have been estimated at ten times or more
the short-run response (World Bank 1986). These production shifts be-
tween crops provide gains in allocative efficiency and increase the eco-
nomic value of production in a step-like fashion.4
But such correction of relative prices alone will not lead to sustained
increases in agricultural growth. First, the aggregate supply response in
agriculture is much lower than the response of individual crops, even in
the longer run, since the supply of land, a major factor of production,
is relatively inelastic in many countries (Binswanger and others 1987).
Even Argentina, which until recently had a fairly elastic supply of land
and well-developed infrastructure, has an estimated aggregate supply
elasticity of between 0.21 and 0.35 for the short run, although it may
be close to 0.8 for the long run. Countries with less developed infra-
structure have a much lower aggregate supply elasticity, for example,
between 0.2 and 0.3 in the short and long run for India, about 0.1 in
the short run and 0.16 in the long run in Kenya (Chhibber 1988).
Second, the price adjustment is often comparatively small. A study of
nineteen developing countries (Krueger, Schiff, and Valdes 1988) found
that domestic prices for export crops averaged only 11 percent below
their international equivalents in 1975-84. This study also found that
domestic prices were actually higher than their international counterparts
in the case of crops whose domestic supplies were in part imported.5
Third, aggregate response to price changes is limited to crops grown by
producers who market their crops or use purchased inputs; the changes
have virtually no impact on the production of large numbers of subsist-
ence farmers.
Of course, if the reduction of price distortions causes production to
shift to outputs with high income elasticities, such as livestock or fresh
fruits and vegetables, it may improve growth rates. The switch to fruits
and vegetables contributed to the growth of Chilean agriculture, for ex-
ample. But with the exception of livestock in some countries, such as
Argentina and Uruguay, these products are not heavily taxed by direct
pricing policies.



Agricultural Policy 139
NONPRICE FAcrORS. Public goods such as roads, irrigation, and re-
search and extension have been shown to be extremely important in
sustaining agricultural growth (for a survey of related studies, see Lele
and Mellor 1988). Chhibber (1988) found that, on average, a 1 percent
increase in irrigation infrastructure increased output by 1 percent in the
countries studied. Other studies show a significant response of output to
increased density of rural roads as well (Binswanger 1989). Although
such findings might lead one to ask whether price or nonprice factors
are more important, a more useful approach is to view them as com-
plementary. Investment in irrigation or drainage expands the fixed factor
of production, land, and gives farmers a greater choice of crops. Improved
transportation makes it profitable for farmers to produce some outputs,
such as perishables, that could not be marketed before. Nonprice factors
lower the cost of adjusting to changes in relative prices and increase
responsiveness to these changes.
Price and nonprice factors are complementary in another, longer-run
sense as well. Recent research has emphasized the importance of relative
prices in guiding the direction of technological change in agriculture
(Hayami and Ruttan 1985; Mundlak 1988). Where one input is relatively
expensive, technological change will proceed so as to economize on its
use. Beyond this, decisions affecting both public and private investment
in physical infrastructure and human capital, research and extension ac-
tivities, and the adoption of new varieties or techniques by farmers are
all guided by relative prices. Where prices are close to their opportunity
costs, investment and technological change are more efficient. This is
particularly true of intersectoral pricing. A study (Cavallo and Mundlak
1982) that included technological change induced by changes in inter-
sectoral prices estimated a higher aggregate long-run supply elasticity
than have other studies.
Since public goods such as roads, irrigation systems, and research and
extension services take many years to develop and yield returns, their
beneficial effects on agricultural growth become apparent only in the
long term. Meanwhile, the austerity in public expenditures required by
structural and macroeconomic adjustment may initially slow growth,
despite the increase in agricultural prices that results from adjustment
measures. There is thus a danger that the sustainability of the reforms
will be threatened as the government and farmers become discouraged
with the slow, painful process.
The Chilean experience with agricultural reform over many years, with
some periods of retrenchment before the effects on growth were felt,
together with what is known generally about the determinants of agri-
cultural growth, suggest certain lessons for agricultural reform and ad-
justment lending. In planning the first agricultural SECAL to any country,
the World Bank and the government should agree on a series of adjust-



140 ISSUES OF POLICY REFORM
ment operations to take place over at least five years. Some of these
operations could include hybrid loans that combine fast-disbursing ad-
justment loan funds with slower-disbursing project or program funds.
The first operation should not be too ambitious or politically painful.
What is important is to secure a clear commitment to the longer-term
objective and see that precedents are set. Ideally, there should be a public
commitment by the government to the basic elements of a long-term
program. Loan effectiveness should be judged according to the imple-
mentation of policy reforms rather than an immediate improvement in
agricultural growth and trade balances. Some retrenchment should be
anticipated so that temporary adjustments can be made when unforeseen
events occur or reforms do not work as expected. Only after the program
is well advanced-perhaps after five to ten years-can its effectiveness
in increasing output be assessed.
The experience with price and nonprice factors and long-term agri-
cultural growth also demonstrates the importance of maintaining public
and private investment. Because of fiscal austerity, public investment
programs have suffered, particularly in agriculture. In Mexico, where
fiscal constraints are particularly severe, public investment in agriculture
dropped from US$1,824 million in 1982 to less than US$500 million in
1987. In many countries a tremendous potential exists for switching some
current expenditures to investment while still reducing overall expen-
ditures. For example, although credit programs have been found to be
ineffective in encouraging agricultural growth (Binswanger 1989), credit
subsidies accounted for 4 to 99 percent of total govemment spending for
agriculture in a sample of six Latin American countries in 1970 and 1980
(Elias 1985). In most countries under adjustment, however, funds re-
leased as a result of lower expenditures for salaries and input subsidies
have not been used to augment investment budgets. In Morocco, agri-
culture's share of public investment fell from 15 percent in 1984 to 7
percent in 1986, despite substantial fiscal savings from reductions in input
subsidies. Net investment in agriculture has probably been negative, with
depreciation in agricultural and rural infrastructure outweighing the
small amounts of investment that have continued.
Sector loans have rarely addressed both the quality and the specific
level of funding of the public sector investment portfolio. About 50 per-
cent of SECALs contain conditions related to a satisfactory investment
budget (see table 7-2), but frequently the emphasis is on stopping "white
elephant" projects rather than on a thorough review of the investment
portfolio or of government procedures for project selection. The Mexico
agricultural SECAL is the only loan to specify a minimum level of funding
for agricultural investments. Future adjustment loans should require that
at least part of the savings produced by a reduction in current expenditure
be converted to investments. Also, to protect investment budgets and



Agricultural Policy 141
Table 7-2. Conditions in Twenty-One Agricultural Sectoral
Adjustment Loans (SECALs), Fiscal 1980-88
Number of agricultural SECALS
Reforms preceding         Conditions in
Condition                                    loan                     loan
Product price                                 17                       15
Decontrol consumer price                     2                        4
Decontrol producer price                     3                        1
Change official consumer price               5                        1
Change official producer price               7                        5
Reduce subsidies                             2                        2
Liberalize trade                             9                        8
Conduct studies                              3                        7
Input price                                    14                      15
Decontrol or set at world level               1                       1
Raise official price                        10                       12
Lower official price                         0                        0
Liberalize trade                             4                        3
Conduct studies                               1                       3
Privatization                                  10                       9
Product distribution                         7                        5
Input distribution                           3                        3
Conduct studies                              0                        1
Institutional reform                          10                       17
Deregulation                                 0                        3
Other                                        9                       14
Conduct studies                               1                       6
Macroeconomic                                 12                        3
Exchange rate                                12                       1
Deficit                                       4                       1
Other                                        2                        3
Public investment                              4                       10
Environment                                    0                        2
Poverty alleviation                            2                        2
Source: Authors' examination of World Bank President's Reports on twenty-one agricultural
SECALS (virtually all those through fiscal 1987 and one for Mexico in 1988).
encourage a more efficient allocation of government-provided goods and
services, both current and capital government expenditures for agricul-
ture should be increasingly financed by farmers through user fees and
other cost-recovery measures. The design of administrative structures to
permit this cost recovery should be part of World Bank financing. By
exerting stronger pressure to reduce current spending and recover costs,
the World Bank could ensure the survival of investment budgets without
compromising the macroeconomic objective of deficit reduction.



142 ISSUES OF POLICY REFORM
Price Stabilization and Food Policy
Almost all developing countries have programs for price stabilization and
food subsidies. The international prices of food staples are volatile, and
this volatility is magnified in domestic markets by the erratic devaluations
typical of countries with macroeconomic imbalances. Almost all gov-
ernments, whatever their views on the role of the public sector, attempt
to stabilize the prices of important agricultural crops. Since food con-
stitutes a large part of the total expenditure of the populace, especially
the poor, programs to correct price volatility and alleviate its effects
receive high priority. Given these aspects of food policy in developing
countries, reform options in this area are relatively constrained. To be
sustainable, reforms should be based not on the abandonment of these
objectives but rather on the development of more efficient and less in-
trusive mechanisms for stabilizing food prices and protecting food con-
sumption.
The mechanisms used to stabilize prices vary widely, ranging from
domestic procurement or importation schemes directed toward a specific
commodity to financial schemes such as buffer funds, which are financed
through variable taxes. Whatever the mechanism, the result has usually
been to tax producers. Although the effect of price instability on pro-
duction plays a part in government thinking, the welfare of urban con-
sumers is a prime concern. Indeed, governments often go beyond price
stabilization and subsidize consumption through price controls or general
subsidies. (General subsidies are those that are available to all consumers
of a product.) Fear of political unrest produced by sudden increases in
basic food prices reinforces this tendency. Exchange rate instability also
accentuates the desire to stabilize commodity prices. When governments
devalue as part of an adjustment program, they usually want to tax the
windfall gains to producers and holders of commodity stocks in order
to keep consumer prices down. The use of export taxes to partly offset
movements in the exchange rate is well documented for Uruguay (Jarvis
and Medero 1988) and Argentina (Sturzenegger forthcoming).
Stabilization of commodity prices may also help attain macroeconomic
objectives. Large movements in commodity export earnings that consti-
tute a significant part of aggregate foreign exchange earnings can de-
stabilize the economy through effects on the money supply or "Dutch
disease" effects (see, for example, Pinto 1987). When governments are
unable to institute countervailing fiscal and monetary programs because
of general macroeconomic instability, export taxes can exert a stabilizing
effect on the economy. For this to happen, however, the increased revenue
from the taxes that are imposed when export earnings are high must be
used to reduce the creation of domestic credit, not to increase spending.



Agricultural Policy 143
Used in this way, export taxes sterilize part of the change in the flow of
foreign exchange.
The extensive theoretical debate on price stabilization has not produced
any clear conclusions about whether a theoretically well-designed sta-
bilization mechanism increases or decreases the welfare of producers and
consumers. Price stabilization schemes in practice, however, clearly im-
pose enormous costs. Price supports reinforced by trade controls not only
insulate producers from short-term fluctuations in world prices but also
divorce domestic prices from long-term trends in border prices, thereby
perpetuating inefficient patterns of production. Public enterprises
preempt private sector participation in the procurement, storage, trans-
port, processing, and sale of basic foods, while their purchases are sub-
sidized by fiscal transfers or low producer prices. Huge financial losses
and economic inefficiency are the results (Knudsen and Nash forthcoming
a and b).
Experience with sector operations shows that most governments will
not completely abandon the objectives of stabilizing prices and subsi-
dizing consumption. Thus, for example, if a public enterprise is the mech-
anism used to implement this goal, the government will be unwilling to
privatize or abolish it before having a viable alternative for stabilizing
prices. The attempt to remove the government from procurement ac-
tivities in Ecuador failed because no viable alternative was specified.
What is the lesson for agricultural adjustment lending? The desire of
governments to stabilize prices or other macroeconomic variables should
be recognized, but more efficient means should be sought to achieve these
objectives. Government intervention should be limited to a small number
of commodities and should not involve direct handling of the commodity.
Ideally, the intervention should be based on publicly announced pricing
rules and prices linked to international prices. If the exchange rate is
unstable, then an ad hoc price stabilization scheme may be needed until
the macroeconomic situation is under control. When the concern is to
stabilize prices to avoid macroeconomic fluctuations, an alternative to
ad hoc export taxes deserve consideration, such as the revenue stabili-
zation fund for copper that is used in Chile or the buffer funds that are
used for agricultural exports in Papua New Guinea.6
Consumer price controls and subsidies on important foods are also
unlikely to be entirely removed. General subsidies, however, should be
replaced by a targeted food system for the poor, as recent World Bank
operations have attempted to do. The Mexico agricultural SECAL limited
the maximum level of global subsidies, while also requiring a minimum
level of funding for direct, targeted subsidies to cushion the impact on
the poor. The team preparing the Venezuela SAL is aware of this tradeoff
and is focusing on the development of indirectly and directly targeted
food subsidies.



144 ISSUES OF POLICY REFORM
Food stamps are a cost-effective method of targeting subsidies, but the
system cannot be developed overnight. Much administrative work is re-
quired to minimize abuses and keep down costs. The food stamp system
in Mexico took more than a year to design and begin implementation-
and even then serious errors were made.
In designing alternative subsidization schemes, it should be recognized
that leakages are inevitable and may even be used to broaden support
for the alternative system and thus contribute to its sustainability. In
Mexico, for example, food stamps were provided to unions as part of a
political bargain when removing global food subsidies. To limit the pro-
gram's costs and to avoid creating strong incentives for rural-urban mi-
gration, the subsidy should not be open-ended. Furthermore, the price
of the stamps should be linked to the price of the food product to prevent
program costs from ballooning quickly in a period of rapid inflation, as
happened in Mexico. (Food stamps redeemable for a given physical quan-
tity of food had a fixed price unrelated to the food's actual price.)
The impact on the income of the target population of price rises re-
sulting from the adjustment program should be estimated and measures
taken to ensure the transfer of approximately equivalent resources to this
population through the targeted subsidy program. The Mexico agricul-
tural SECAL required targeted food programs to redress the estimated
US$225 million reduction in the purchasing power of the poorest 20
percent of the population that would occur as a result of food price
adjustments once general subsidies were eliminated. Obviously, decisions
on levels of funding for direct subsidies must be balanced by macro-
economic and fiscal considerations.
Trade Liberalization
One of the most common objectives of agricultural adjustment loans has
been trade liberalization (see table 7-1). The success of such trade reforms
is determined by a number of factors, however, including macroeconomic
considerations, pricing objectives (including stabilization), and trade in-
centives in other sectors. This section discusses some of these connections
and their implications for policy reform.
The most significant discrimination against agriculture has occurred
through protection of industry and overvaluation of the exchange rate.
A study of agricultural incentives in sixteen developing countries found
that the "disprotection" of agriculture resulting from overvaluation and
industrial policy was 25 to 29 percent; that is, farmers received real prices
that were 25 to 29 percent lower than they would have been in the absence
of such policies. This discrimination generally overwhelmed the advan-
tage received from various input subsidies to producers or direct pro-
tection from restrictions or taxes on imports (Krueger, Schiff, and Valdes



Agricultural Policy 145
1988). Liberalization of agricultural trade should therefore be preceded
or accompanied by a general trade liberalization program for other com-
modities, including adjustment of the exchange rate to realistic levels.
This has in general been done in most programs.
The agricultural trade regime cannot be treated independently of the
agricultural pricing policy. If, for example, quantitative restrictions are
to be removed from a commodity for which minimum price or other
price stabilization measures are in effect, the replacement regime must
be able to fulfill the same objective. A variable tariff might appear to be
an adequate substitute that will provide the same degree of protection,
but in practice it may not be able to do so. Nominal exchange rate
movements may overwhelm even a variable tariff, and adjusting the tariff
to exchange rate fluctuations may prove administratively difficult, es-
pecially in the presence of parallel exchange markets and capital controls.
In addition, the stated or invoice price on which the tariff is calculated
may be false or may not take into account other favorable terms, such
as credit concessions. Also, when capital controls are in effect, it may
pay to import agricultural goods at a loss, thereby permitting false in-
voicing and providing a mechanism for moving capital abroad.
If, under a minimum price scheme, the government is committed to
purchase all available supplies at a fixed price, the budgetary repercus-
sions of undercutting trade protection may be serious; the government
may find itself paying the support price for domestic production that is
displaced by a flood of imports. Under these circumstances, it will be
difficult to obtain government cooperation in converting quantitative
restrictions to tariffs. If variable tariffs are put in place when the exchange
rate is unstable, the government should not try to enforce a minimum
price rigidly. Price movements below this price are inevitable as exchange
rate movements undercut prices.
Because of these connections between direct procurement schemes and
trade controls, other means must be found to stabilize prices. On the
premise that a well-functioning storage market will best stabilize prices,
the Madagascar agricultural SECAL introduced a staged process for build-
ing up private capabilities for rice procurement and storage. As a first
step, the government withdrew from domestic procurement but operated
a buffer stock of imported rice, and it applied a pricing mechanism that
sought to avoid large price increases from speculation while allowing a
profit to those engaged in private trading and storage. This scheme
worked well in establishing the private sector in these activities. The next
step is to be government withdrawal from running the buffer stock as
all storage functions are taken over by the private sector.
Because of import barriers that indirectly create disincentives to export,
some governments have attempted to promote exports. Measures have
included reduced export taxes, tax or duty drawback schemes, duty-free



146 ISSUES OF POLICY REFORM
zones, special credit lines, insurance or other special subsidies for ex-
porters, and the creation of government agencies to promote exports in
other ways. Often poorly designed for correcting antiexport distortions,
these measures may turn out to be as costly and inefficient as overzealous
import protection. A study of export promotion through subsidized credit
and insurance found the schemes to be ineffective and socially costly
(Fitzgerald and Monson 1989). These programs seldom include primary
agriculture, although they often include agroindustry, and they adversely
affect agriculture through their effect on the exchange rate and through
higher taxation of the sector to support the export promotion activities
(Nogues 1989). An alternative approach that relies on technical assist-
ance to governmental trade promotion organizations has almost never
been successful because of problems inherent in such a role for the public
sector. The only exception found (by Keesing and Singer 1989) was the
official trade promotion organization in Chile, which was marginally
helpful in promoting exports of grapes and salmon.
Evidence from Africa suggests that the best way to promote exports
(including agricultural exports) may simply be to follow a realistic ex-
change rate policy and to reduce government interference in markets.
Balassa (1988a) compared the export performance of nineteen Sub-
Saharan African countries in 1974-78. He found that, on average, those
following market-oriented policies gained 15 percent in export market
share, while those following interventionist policies lost 24 percent, with
intermediate countries losing 10 percent.
The Deepening of Reforms
The strategy behind policy-based lending is to make basic economic re-
forms under the auspices of a SAL or a broad-based SECAL and then, once
the economy has been sufficiently stabilized, to introduce deeper and
broader reforms. Reforms undertaken in this way are considered more
likely to be institutionalized and so to become permanent. Thus, for
example, Mexico's commitments upon joining the General Agreement
on Tariffs and Trade (GArr) institutionalized, or "deepened," its trade
reform process. This deepening of reforms is essential to a program's
success because the private sector will not make the desired investment
decisions unless the reforms are viewed as permanent (Rodrik 1988).
Most adjustments undertaken previous to or as conditions of SECALs
have not helped to deepen reform. Experience with product pricing pro-
vides an example of why this is so. Although measures were introduced
to raise the levels of controlled consumer prices or official producer
prices, the institutional apparatus for determining prices was usually left
intact and thus subject to the same political forces that had caused prices
to be set low to begin with. This failure to dismantle the old apparatus



Agricultural Policy 147
is especially troublesome because as inflation erodes the price increases
the political battles need to be fought over and over again to ensure that
price increases are maintained in real terms.
In a few cases the mechanisms used to set official prices have been
altered rather than the specific price levels, for example, by creating a
link with world prices. This approach alleviates some political pressure.
But in only four of the twenty-one SECALs examined has the real structural
problem been addressed by eliminating controls on both trade and con-
sumer prices (see table 7-2). Given the ease with which superficial price
changes can be reversed, the private sector will be reluctant to base long-
term investment decisions on these reforms.
The situation is similar with respect to inputs, an area in which reforms
are generally superficial and do not affect the basic structure of the de-
fective system. In the case of fertilizer, for example, instead of removing
the government from the business of manufacturing and distributing fer-
tilizer, reforms have raised the official price in order to reduce the heavy
subsidies required by the fertilizer sector, which is frequently a govern-
ment monopoly. It is doubtful, however, whether the pressure to allow
prices to lag behind inflation can be resisted once the government's press-
ing budgetary problems are ameliorated.
Most reforms directed at the privatization of state enterprises have also
been quite superficial. Many loans have required that the legal monopoly
of certain parastatals be eliminated, but only one loan has required that
the government actually withdraw from a market. Simply eliminating
the legal monopoly of the government has not generally been sufficient
to overcome the private sector's reluctance to enter a market, given the
threat of unfair competition from a public enterprise with unlimited re-
sources and the power of government at its disposal. (In Mexico, for
example, the private sector did not respond quickly when a public en-
terprise's monopoly in grain imports was abolished.) In many cases
World Bank lending has supported the same publicly owned marketing
organizations that were specifically designed to suppress the small trad-
ers, who, according to a recent review of agricultural marketing projects,
"tare often best placed to meet the needs of small farmers" (World Bank
1988). Criticizing this pattern, the review concluded it resulted from
unwarranted assumptions that private traders are inefficient or exploi-
tative and that strategic crops should be marketed by state monopolies.
The legal regulatory framework is another problem that has not really
been touched. In most developing countries the regulatory environments
impede change and thwart the objectives of trade liberalization. Although
these obstacles are usually not a direct burden on the agricultural sector,
they are of special concern to agroindustries, many of which expand in
response to changes in the trade regime or other adjustments. But only



148 ISSUES OF POLICY REFORM
three of the twenty-one SECALS acted to alleviate regulatory burdens, and
none required action on this issue before the loan (see table 7-2).
It is not difficult to understand why this pattern of superficial reform
has emerged. A deepening of reforms requires very detailed and complete
information on the legal framework in each country. In addition, gov-
ernments are reluctant to make changes that are not easily reversed.
Nonetheless, now that the adjustment process is supposed to be well
advanced in many countries, the World Bank should put more emphasis
on the legal framework, and governments, if they are truly committed
to adjustment, should be more willing to relinquish controls in ways that
cannot readily be reversed. Privatization of state enterprises, while not
a guarantee of continuity, certainly raises the cost of reversal. The GATr,
an international forum for making trade reform binding, can exert direct
pressure that raises the costs of reversal and makes reform more durable:
under GATT accords, reversal of trade reductions in tariffs or other re-
strictions can invite retaliation from other member countries.
Because of the political difficulty of some institutional reforms, the
World Bank should be prepared to accept that reforms may be gradual.
Nonetheless, a basic condition of initial lending should be the govern-
ment's willingness to commit itself to significant long-term institutional
reforms by policy declarations and precedent-setting dismantling or sales
of at least some public enterprises.
The World Bank review of structural adjustment lending raised the
question of whether sectoral adjustment lending can proceed before ma-
croeconomic stabilization has been achieved (World Bank 1989). Al-
though progress on agricultural reform may be possible in an unstable
macroeconomic situation, it is rare, especially in the case of pricing and
trade policy. With inflation or unstable exchange rates, not only are
relative prices and trade unlikely to respond quickly to uncertain signals
but also the political will for change deteriorates. Agricultural SECALS in
Argentina and Ecuador failed in large part not because the designs were
faulty or too ambitious, but because the macroeconomic situation was
so unstable that the political will to act was eroded.
The Future
What does the future hold? Have the World Bank and its borrowers
learned the lessons of the early 1980s on adjustment lending?
The answer is mixed. A certain fatigue is evident. New agricultural
SECALs are not planned for 1988-92 for twelve of the twenty-one coun-
tries that previously received such loans (Burkina Faso, Central African
Republic, Colombia, Ecuador, Madagascar, Morocco, Nigeria, Sierra
Leone, Sudan, Uruguay, Yugoslavia, and Zambia). Only four countries
that received such loans in 1986 or 1987 are currently expected to receive



Agricultural Policy 149
follow-up loans during the next five-year period, and even in these cases
the lags are often as long as three to four years. Although this does not
mean that reform will halt in many of these countries, it does imply that
the World Bank's role will be limited. Furthermore, it raises doubts about
whether reforms will stick or be deepened.7
On the positive side is evidence that the World Bank has begun in
recent loans to address some problems in innovative ways that may give
momentum to the reform process. One area in which this has been oc-
curring is price stabilization in the context of agricultural trade liberal-
ization. An agricultural SECAL for Morocco has assisted in building a
variable tariff system to replace trade controls, and one for Mexico in-
itiated a study on the feasibility of implementing a similar system. The
Madagascar adjustment loan is exploring the use of buffer stocks to
stabilize prices. These schemes are still in the experimental stage, and in-
depth evaluations will not be possible for some time.
The situation is similar with respect to food subsidy programs. Al-
though recent adjustment loans (to Mexico and Morocco, for example)
have recognized the need for cushioning price adjustments with targeted
subsidies for the poor, the World Bank operational staff is just beginning
to build up the specialized knowledge needed to deal with food programs.
The World Bank and developing countries are learning from the dif-
ficult process of agricultural reform. It is a new process that has yet to
become self-sustaining, and the World Bank's participation will continue
to be necessary. With the Uruguay Round of Multilateral Trade Nego-
tiations in its last phases, it may be that agricultural reform efforts in
developing countries will be complemented by global reform so that mar-
kets rather than governments will guide agricultural development in the
1990s.
Notes
We wish to thank Bela Balassa, Ajay Chhibber, and Jaime de Melo for very
thoughtful comments on early drafts of this chapter. Remaining errors are our
responsibility.
1. The discussion in this section is based on an analysis of a sample of ad-
justment loans by the World Bank's Industry Development Division and on tables
7-1 and 7-2. The true contribution of agricultural conditions to adjustment loans
is probably understated here, since agricultural SECALs are underrepresented in
the Industry Development Division's sample.
2. A few loans, such as those to Colombia and Morocco, have put a ceiling
on producer prices or required that they be reduced. For some other countries,
such as Burkina Faso, C6te d'lvoire, Ecuador, Ghana, the Philippines, and
Uganda, the conditions have concerned improving relative prices so as to en-
courage diversification in the sector.
3. That is, increasing allocative efficiency (eliminating the classical "triangle"
losses in traditional welfare analysis) shifts GDP to a higher plane but does not
by itself raise the underlying growth rate of the economy.



150 ISSUES OF POLICY REFORM
4. The agricultural reform in China is an interesting example. Although output
of several crops increased at unprecedented rates (grains at 5 percent, cotton at
17 percent) from 1979 to 1983, output slowed considerably after 1984 (to 1.1
percent and 1.2 percent, respectively) as the effect of the initial price and insti-
tutional reforms dissipated. Even the growth between 1980 and 1983 was at-
tributable largely to nonprice factors such as institutional reform. An econometric
study found that 60 percent of the expansion was due to institutional reform,
25 percent to input use and technology changes, and only 15 percent to price
changes (Lin 1987).
S. Indirect taxation through overvalued exchange rates and other measures,
however, was found to be quite substantial, overwhelming even positive price
differences between domestic and international prices. These distortions are not
ameliorated by reducing the differences between domestic and border prices.
6. These funds not only partially neutralize the macroeconomic consequences
of export earnings instability, but they do so in a way that does not tax the
agricultural products in the long run (see Knudsen and Nash forthcoming a).
7. We are indebted to Anandarup Ray for making this point in an internal
World Bank memorandum to Randolph Harris, March 8, 1989.
References
Balassa, Bela. 1988a. "Incentive Policies and Agricultural Performance in Sub-
Saharan Africa." PPR Working Paper 77. Policy, Planning, and Research, World
Bank, Washington, D.C.
c   . 1988b. "Quantitative Appraisal of Adjustment Lending." PPR Working
Paper 79. Policy, Planning, and Research, World Bank, Washington, D.C.
Binswanger, Hans P. 1989. "How Agricultural Producers Respond to Prices and
Government Investments." Paper presented at First Annual World Bank Con-
ference on Development Economics, April 27-28, Washington, D.C.
Binswanger, Hans P., Maw-Cheng Yang, Alan Bowers, and Yair Mundlak. 1987.
"On the Determinants of Cross-Country Aggregate Agricultural Supply." Jour-
nal of Econometrics 36:111-31.
Cavallo, Domingo, and Yair Mundlak. 1982. Agriculture and Economic Growth
in an Open Economy: the Case of Argentina. Research Report 36. Washington,
D.C.: International Food Policy Research Institute.
Chhibber, Ajay. 1988. "Raising Agricultural Output: Price and Nonprice Fac-
tors." Finance & Development June:44-47.
Elias, V. V. 1985. Government Expenditures on Agriculture and Agricultural
Growth in Latin America. Washington, D.C.: International Food Policy Re-
search Institute.
Fitzgerald, Bruce, and Terry Monson. 1989. "Preferential Credit and Insurance
as Means to Promote Exports." World Bank Research Observer 4 (Janu-
ary):89-114.
Hayami, Yuiiro, and Vernon Ruttan. 1985. Agricultural Development: An In-
ternational Perspective. Baltimore, Md.: Johns Hopkins University Press.
Jarvis, Lowell, and Maria del Rosario Medero. 1988. "Domestic Beef Price Sta-
bilization in a Beef Exporting Country: Uruguay, 1961-86." Paper presented



Agricultural Policy 151
at the Twentieth Conference of the International Agricultural Economics As-
sociation, Buenos Aires, Argentina, August 24-31.
Keesing, Donald B., and Andrew Singer. 1989. "Experience of the World Bank
in Support and Assistance for Export Marketing." Trade Policy Division,
Country Economics Department, World Bank, Washington, D.C.
Knudsen, Odin, and John Nash. Forthcoming a. "Agricultural Price Stabilization
and Risk Reduction in Developing Countries." In Alberto Valdes and Romeo
Bautista, eds., Trade and Macroeconomic Policies' Impact on Agriculture.
Washington, D.C.: International Food Policy Research Institute.
. Forthcoming b. "Domestic Price Stabilization Schemes in Developing
Countries." Economic Development and Cultural Change.
Knudsen, Odin, and P. A. Yotopoulos. 1985. "Agriculture in the Process of Trade
Liberalization and Retrenchment: The Case of Chile." Agriculture Depart-
ment, World Bank, Washington, D.C.
Krueger, A. O., Maurice Schiff, and Alberto Valdes. 1988. "Agricultural Incen-
tives in Developing Countries: Measuring the Effect of Sectoral and Econo-
mywide Policies." World Bank Economic Review 2 (3) September:255-71.
Lele, Uma, and John W. Mellor. 1988. "Agricultural Growth, Its Determinants,
and Their Relationship to World Development: An Overview." Paper prepared
for the Twentieth Conference of the International Agricultural Economics As-
sociation, Buenos Aires, Argentina, August 24-31. (Revised September 23,
1988.)
Lin, Justin Yifu. 1987. "Household Farms, Cooperative Farms and Efficiency:
Evidence from Rural De-Collectivization in China." Discussion Paper 91. De-
partment of Economics, University of Hong Kong.
Mundlak, Yair. 1988. "Capital Accumulation, the Choice of Techniques, and
Agricultural Output." In John W. Mellor and Raisuddin Ahmed, eds., Price
Policy. Baltimore, Md.: Johns Hopkins University Press.
Nogues, Julio. 1989. "Latin America's Experience with Export Subsidies." PPR
Working Paper 182. Policy, Planning, and Research, World Bank, Washington,
D.C.
Pinto, Brian. 1987. "Nigeria During and After the Oil Boom: A Policy Com-
parison with Indonesia." World Bank Economic Review 1 (May):419-46.
Rodtik, Dani. 1988. "Liberalization, Sustainability, and the Design of Structural
Adjustment Programs." Trade Policy Division, Country Economics Depart-
ment, World Bank, Washington, D.C.
Sturzenegger, Adolpho. Forthcoming. "The Political Economy of Price Discrim-
ination in the Argentine Pampas." In A. 0. Krueger, Maurice Schiff, and Al-
berto Valdes, eds., The Political Economy of Agricultural Pricing Policy: Latin
America. Baltimore, Md.: Johns Hopkins University Press.
World Bank. 1986. World Development Report 1986. New York: Oxford Uni-
versity Press.
. 1988. "Agricultural Marketing: The World Bank's Experience." Op-
erations Evaluation Department, World Bank, Washington, D.C.



152 ISSUES OF POLICY REFORM
. 1989. Adjustment Lending: An Evaluation of Ten Years of Experience.
Policy and Research Series 1. Washington, D.C.
Comments
Nurul Islam
MY COMMENTS touch on several of the issues discussed by Odin Knudsen
and John Nash: the influence of price and nonprice factors; the rela-
tionship among prices, aggregate supply response, and growth; input
subsidies and government expenditures on agriculture; price stabiliza-
tion; protecting the poor during adjustment; agricultural protectionism;
and the sustainability of reform.
Reforms of agricultural pricing policies, including those affecting the
level and stability of prices, liberalization of agricultural trade, and re-
duction in government involvement in the marketing and distribution of
agricultural outputs and inputs, have been among the most widely dis-
cussed and controversial aspects of agricultural structural adjustment
programs. The analysis of Knudsen and Nash deals more with policy
recommendations in the World Bank's agricultural adjustment lending
than with the impact of such reforms on performance. Still urgently
needed is an empirical analysis of how the reforms were actually imple-
mented and what their impact was on the agricultural sector.
Extensive research has quantified the extent to which economywide
policies such as those dealing with trade and the exchange rate affect
relative prices not only within agriculture but in other sectors as well.
Research has also shown that the effects of these policies dominate those
of sector-specific pricing policies implemented through subsidies and
taxes or public intervention. For reforms of price policy to be effective,
they must be combined with or even preceded by macroeconomic policy
reforms, especially in the areas of trade and the exchange rate. High
inflation or volatile exchange rates can soon erode or offset reforms of
government support prices or agricultural taxes and subsidies.
Although some years ago experts considered "getting the prices right"
to be the most effective way of generating efficient growth, opinion today
seems to be shifting toward the belief that both price and nonprice factors
are important. Experience with structural adjustment programs may have
contributed to this pragmatic conclusion. Controversy persists, however,



Agricultural Policy 153
on two issues: (1) the relative importance of price and nonprice factors
and (2) the sequencing of corrective measures on these two fronts.
The relative priority that should be attached to price and nonprice
factors depends on the particular circumstances of each country. If price
distortions are severe and long-standing, pricing policy reforms may
greatly improve efficiency and growth even if nonprice factors are in-
adequate. It would be a one-shot improvement, however, unless nonprice
factors were significantly strengthened. In other cases, the strengthening
of nonprice factors may be the dominant consideration, partly because,
as Knudsen and Nash point out, the aggregate supply response to price
changes is itself a function of nonprice factors.
The interrelations between price and nonprice factors can be illustrated
in a number of cases. For example, despite favorable coffee and tea prices
in Kenya, productivity per hectare increased little among smallholders,
primarily because of inadequate access to research and extension services,
credit, and inputs; in contrast, productivity increased on large estates
because of more ample access to these nonprice factors. In Cameroon,
despite lower cotton prices than in Kenya, Malawi, and Nigeria (both
in nominal terms and in purchasing power parity) and despite a fall in
the price of cotton relative to the price of maize, yields per hectare were
larger than in other countries because Cameroonian farmers had access
to improved technology and extension services. Similarly, in Nigeria new
technology and organizational efficiency increased cocoa production de-
spite low returns from exports due to an overvalued exchange rate (Lele
and Mellor 1988).
In the view of many developing countries, the recent heavy emphasis
of international development agencies on pricing policy reforms, without
commensurate attention to nonprice factors, is an excuse for inadequate
assistance and inaction on nonprice issues. Everyone recognizes that im-
proving nonprice factors requires substantial resources in both the short
and long runs. The supply response is not only higher in the long run
than in the short run, but it is also higher in the presence of improved
nonprice factors such as infrastructure. Achieving this response requires
a long-term commitment from both lenders and borrowers.
Another issue discussed by Knudsen and Nash is the relationship
among prices, aggregate supply response, and growth. As they mention,
aggregate supply elasticities in agriculture are rather modest, frequently
0.6 or less. The higher estimates of price responses that are obtained in
a general equilibrium framework come only after a period of ten to fifteen
years and are based on models of technological change and capital ac-
cumulation that include intersectoral capital flows and labor migration.
These models seem to suggest that favorable price changes in agriculture
are the principal engine of growth. Such price changes set in motion
technological progress, which is incorporated in physical and human



154 ISSUES OF POLICY REFORM
capital on the basis of farmers' selection of the most appropriate tech-
nologies from among a wide set of those readily available. This model,
however, tends to ignore or underplay the role of public action in the
field of technology, institutions, and infrastructure. Empirical verification
of these models, including their implicit assumptions, is still limited and
requires further research (Mundlak 1989; Cavallo and Mundlak 1982;
IFPRI 1988).
In this connection, the controversy regarding the relationship between
relative price changes in agriculture and the generation of technology
needs some clarification. First, it is generally agreed that price incentives
accelerate the diffusion of technological innovations once they are avail-
able; what is being debated is the generation of technology in response
to price incentives. Second, the relevant policies and organizations differ
for basic or fundamental research and applied and adaptive research.
Third, no distinction is made between changes in relative prices within
agriculture and changes in agricultural prices relative to nonagricultural
prices, or between the effects of short- and long-term price changes on
the generation of technology. Systematic analysis and clarification of the
various aspects of the process of technological innovation and their in-
terrelations are needed. For example, although the emphasis of private
research may be affected by relative price changes within agriculture, it
is unlikely that aggregate government research expenditures on agricul-
ture or basic science are affected to any significant extent.
In the context of adjustment lending, government expenditures on ag-
riculture, particularly input subsidies, have received much attention in
recent years. Loans for agricultural adjustment programs have usually
included the condition that input subsidies be reduced or eliminated.
Among input subsidies, the most controversial is the one for fertilizer-
the most important ingredient in the Green Revolution. Several argu-
ments have been advanced in favor of a fertilizer subsidy: it compensates
for imperfections in the capital market and for the risks associated with
adopting a new technology with high cash costs; it is required to stimulate
the process of trial and error needed initially to determine the optimum
uses of fertilizer; and it is needed temporarily to offset high costs of
distribution and transportation in the early stages of fertilizer use, when
the volume of sales is low and the unit cost of distribution is high.
The importance of these considerations depends on the circumstances
of each country and the effect of fertilizer use on output relative to the
prevailing ratio of fertilizer price to output price. Although it is often
argued that, however justified, temporary subsidies are inadvisable be-
cause they tend to become permanent, several countries have shown that
subsidies can be reduced or even eliminated over time. It has also been
argued that the resources spent on subsidies would be better spent on
agricultural investments with higher returns such as research, extension,



Agricultural Policy 155
infrastructure, and so on, and that such expenditures would in turn help
increase returns from the use of fertilizer. Experience with structural
adjustment programs in many countries demonstrates, however, that the
reduction of subsidies is often accompanied by a reduction rather than
an increase in public investment in agriculture, with adverse effects on
growth and employment in agriculture.
Economists continue to debate whether the stabilization of agricultural
prices has a desirable impact on consumer and producer welfare. Some
argue that substantial benefits may accrue from price stabilization if ma-
croeconomic and dynamic considerations are taken into account. Price
instability adversely affects investment not only by farnmers but also by
those engaged in marketing, distribution, and processing. Moreover, con-
sumers, especially poor consumers, incur high transaction costs every
time prices change because they need to reallocate their expenditures in
order to fulfill their minimum nutritional requirements (Timmer 1988;
Ahmed 1989). Furthermore, the very poor-who in many countries con-
stitute a large part of the population and have almost no access to credit-
cannot hold back sufficiently on consumption when prices are low to
save enough to avoid hunger when prices are high. Even farmers who
do not produce enough food for their own needs must sell some of their
output immediately after harvest when prices are low to meet urgent cash
requirements and must buy more during other times when prices are
higher.
Most governments, in both developing and developed countries and
whatever their position on the role of markets, intervene to achieve some
price stability. Governments are unwilling to stand idle and let people
go hungry when world prices have risen, even if their action may worsen
the problem at some time in the future; nor do they allow their farmers
to go bankrupt and join the urban unemployed if prices collapse in the
short run, even if the commodities they produce have limited markets in
the long run. Policymakers therefore need a mechanism for addressing
price fluctuations. Analytical studies and practical measures are needed
so that a cost-effective method for stabilizing prices can be designed.
Prices are affected not only by fluctuations in world prices and ex-
change rates but also by variations in domestic production due to the
weather or the availability of inputs and technological innovations.
Knudsen and Nash concentrate on price instability arising from external
causes rather than from fluctuations in domestic production. This em-
phasis naturally leads to a discussion of taxes and subsidies on foreign
trade as the principal mechanism for stabilizing domestic prices. Expe-
rience, however, is very limited in this respect, and the subject requires
further study. In addition, countries differ widely in their capacity to
formulate and implement such taxes and subsidies.



156 ISSUES OF POLICY REFORM
The fluctuations in domestic output and prices can be reduced by vary-
ing exports and imports, by varying domestic stocks, or by doing both.
For countries that are not consistent exporters or importers, an exclusive
reliance on trade adjustment mechanisms for stabilizing prices is not
necessarily the most effective approach. Occasional exporters cannot sell
their surplus output in the world market except at a considerable loss,
while occasional importers cannot import a small amount at short notice
to offset shortfalls in domestic production without incurring delays and
high costs. Furthermore, a country's staple food products may not com-
monly be traded in international markets. These considerations support
the case for using domestic stocks, in appropriate combination with varia-
tions in exports and imports, wherever feasible as a method for stabilizing
domestic prices.
What is needed is to analyze for a given country the tradeoffs among
(1) the degree of price stability being sought, (2) the size and cost of
stocks, and (3) the degree of reliance on imports. The narrower the price
band (the margin between ceiling and floor prices) within which stability
is sought, the larger the size of the necessary stocks and the larger the
cost of stabilization. For example, a study in Kenya found that holding
the producer price of maize between US$129 and US$151 per ton would
cost between US$1.0 million and US$1.5 million more annually than
would holding prices between US$125 and US$155 per ton (Pinckney
1989). Even a moderate degree of flexibility in response to changes in
world prices or domestic production can considerably reduce the costs
of price stabiliiation. In the Kenyan example, a policy that translates a
US$10.00 change in the world price to a US$1.00 change in the domestic
price would save about US$350,000 annually on average, while a policy
that allows the official price to increase by 6 percent when production
falls by 10 percent would save about US$600,000 annually (Pinckney
1989).
A narrow price band that does not meet the costs of private traders
for marketing, distribution, and processing will discourage private stock
holding and so increase the amount of public stocks required. In view
of the high cost of stabilizing prices through the management of domestic
stocks, such efforts should be confined to a very few vital commodities.
It is unlikely, however, that a public stock policy for price stabilization
can be carried out without some control over imports and exports. If
exports and imports are unregulated while domestic prices are held within
a band, the cost of holding stocks will be enormous. The actual importing
and exporting can be undertaken by private traders under a system of
licensing, and domestic trading in food grains can also be left to private
traders, with occasional intervention by the public stock-holding agency.
At least in the early stages of the adjustment process, the need to protect
and compensate the poor is increasingly being recognized-for example,



Agricultural Policy 157
when food prices rise after the elimination of subsidies or employment
and income fall at least in part because of a reduction in government
expenditures. Targeting subsidies to the poor is easier said than done,
however. It confronts not only the political opposition of those excluded
from the subsidy but also the problems of logistics and inadequate in-
formation about the target group. If the very poor constitute a high
proportion of households, marketwide subsidies may be more cost-ef-
fective than narrowly targeted schemes that are subject to large leakages.
A recent study by the International Food Policy Research Institute found,
for example, that the total cost (the amount of subsidy plus administrative
costs and leakages) of transferring US$1.00 to the target groups in an
economically depressed rural area in the Philippines was US$1.19 if all
households received subsidies, US$1.60 if only households consuming
less than 80 percent of their calorie requirements were targeted, and
US$3.61 if only households with malnourished preschool children were
targeted. The major reason for the differences in costs was leakages to
nontargeted groups in the latter two cases (IFPRI 1988).
The discussion by Knudsen and Nash of agricultural protectionism in
developing countries makes it appear stronger and more pervasive than
it is. It is true that agricultural protectionism is on the rise in the middle-
income developing countries, particularly for import-competing food
products. In other developing countries, however, usually only one or
two commodities have domestic prices that are higher than world prices
at the official exchange rate. The current depression in world prices has
contributed in several instances to an improvement in the relative price
of import-competing agricultural commodities in domestic markets. In
some cases, exchange rate adjustments have led to a rise in domestic
prices. This emphasizes once again that trade reforms cannot be divorced
from exchange rate reforms.
The reform of protectionist policies in agriculture or of trade policies
generally should take place within a framework of overall trade liber-
alization so that intersectoral discrimination does not persist. One forum
for reciprocal trade liberalization is the Uruguay Round of Multilateral
Trade Negotiations. In the Uruguay Round, developing countries, es-
pecially high-income developing countries, will be requested to grant
reciprocity in trade concessions. In this context, they should receive credit
for unilateral trade liberalization undertaken under structural adjustment
programs. Furthermore, as Knudsen and Nash emphasize, the incorpo-
ration within the GAIT framework of concessions already given by de-
veloping countries is likely to improve their durability and reduce the
likelihood of policy reversal.
To contend that the only way to deepen and broaden recent policy
reforms in agriculture is to dismantle all government undertakings or
public enterprises in the agricultural sector seems an extreme position.



158 ISSUES OF POLICY REFORM
Recent experience confirms two major risks or difficulties of rapid and
wholesale privatization: (1) the public monopoly may be replaced by a
private monopoly or oligopoly, with government failures replaced by
market failures of a greater magnitude, and (2) the public marketing and
distribution system may collapse before a viable private one can be de-
veloped. Examples of such outcomes abound. International development
agencies need to maintain a nonideological approach on this matter so
that efforts can be made both to improve the efficiency of public enter-
prises and to encourage competition between public and private enter-
prises. The durability of agricultural policy reforms depends less on the
dismantling of all government institutions capable of interfering with the
market mechanism and more on each country's conviction concerning
the desirability and extent of policy changes. Such reforms should be
based on a reasoned analysis undertaken by the recipient countries in an
atmosphere of open debate in which dialogue with the lending agencies
is only one input. This process requires a national. capability for analyzing
and formulating policy. Assistance by international organizations in
building up such a capability will make an important contribution to the
sustainability of policy reforms.
References
Ahmed, Raisuddin. 1989. Rice Price Fluctuations and an Approach to Price
Stabilization in Bangladesh. IFPRI Research Report 72. Washington, D.C.: In-
ternational Food Policy Research Institute.
Cavallo, Domingo, and Yair Mundlak. 1982. Agricultural and Economic
Growth in an Open Economy: The Case of Argentina. IFPRI Research Report
36. Washington, D.C.: International Food Policy Research Institute.
IFPRI. 1988. "Commentary: Achieving Growth with Equity Through Appropriate
Agricultural Pricing Policy." IFPR1 Report 2. Washington, D.C.: International
Food Policy Research Institute.
Lele, Uma, and John W. Mellor. 1988. "Agricultural Growth, Its Determinants
and Their Relationship to World Development: An Overview." Paper prepared
for the Twentieth Conference of the International Agricultural Economics As-
sociation, Buenos Aires, Argentina, August 24-31.
Mundlak, Yair. 1989. Intersectoral Factor Mobility and Agricultural Growth.
IFPRI Research Report. Washington, D.C.: International Food Policy Research
Institute.
Pinckney, T. C. 1989. Storage, Trade, and Price Policy under Production Insta-
bility: Maize in Kenya. IFPRI Research Report 71. Washington, D.C.: Inter-
national Food Policy Research Institute.
Tinmer, Peter C. 1988. Agricultural Pricing and Stabilization Policy. Cambridge,
Mass.: Harvard Institute for International Development.



8 Industrial Performance: Some
Stylized Facts
James R. Tybout
ROUGHLY A DECADE AGO, external shocks in the form of oil price increases,
world recession, deterioration in the terms of trade, and high interest
rates began to create major adjustment problems for developing coun-
tries. In many cases these problems were exacerbated by excessive public
spending, overvalued exchange rates, and various stopgap policies de-
signed to maintain macroeconomic stability or external balance. By the
early 1980s some combination of adverse external shocks and unwise
domestic policies had placed a number of countries on the brink of eco-
nomic crisis.
Either voluntarily or as a condition for external assistance, many coun-
tries began a process of policy reform. The new policies, which amounted
largely to devaluation and trade promotion, had two basic objectives:
(1) to restore stability and external balance without distortionary inter-
ventions and (2) to lay the foundation for a new era of sustained growth.
This chapter analyzes the performance of the manufacturing sector
during the adjustment process in a sample of fifteen countries that were
heavily involved in adjustment programs and had received adjustment
loans from the World Bank.' The analysis focuses on short-run patterns
of growth and trade. The first part of the chapter briefly reviews the
consequences for the industrial sector of external and policy-induced
shocks and the conditionalities of World Bank adjustment loans and then
examines the performance of the country sample. These countries' growth
rates are juxtaposed with average growth rates for their respective regions
during three periods: precrisis, crisis, and adjustment. Next, a decom-
position of the demand-side sources of growth is used to study the changes
in domestic demand and manufacturing trade flows that have accom-
panied changes in growth. Intertemporal and regional comparisons are
made once again. In the second part of the chapter, an attempt is made
to associate loan conditions with patterns of manufacturing growth and
trade among the sample countries.
Performance of the Manufacturing Sector
Just as trade regimes typically favor industry at the expense of agriculture,
policymakers typically attempt to minimize the adjustment burden im-
159



160 ISSUES OF POLICY REFORM
posed on industrial firms during periods of adverse external shocks and
contractionary macroeconomic policies. There are several reasons for this
concern. Successful industrial development is often viewed as a critical
prerequisite for rapid growth in productivity and gross national product
(GNP). Industrial technology is subject to scale economies and rapid tech-
nological advance; similarly, the industrial sector is an incubator for
entrepreneurial development and the accumulation of human capital.2
Because capital is concentrated in the industrial sector, a large fraction
of national savings originates there; for the same reason, the sector is an
important tax base.
Failure to maintain stable market conditions for industrial producers
can also create short-run problems. One obvious consequence of indus-
trial malaise is urban unemployment; another, perhaps more insidious
consequence is general financial distress. When austerity programs, high
interest rates, or overvalued exchange rates cut industrial profits, liquidity
falls, business failures increase, and the quality of banks' loan portfolios
deteriorates. Especially where financial regulation has been wanting, the
situation can quickly deteriorate into runs, financial crises, and expensive
government bailouts (Tybout 1986).
Unfortunately, it has probably been especially difficult for policymak-
ers in countries with structural adjustment loans (SALS) to insulate in-
dustrial producers, given the shocks they have experienced. First, the
industrial sector relies heavily on imported materials and capital equip-
ment, so foreign exchange shortages have probably constrained current
industrial production, fixed investment, and technical efficiency. Second,
because a substantial fraction of industrial goods are tradable, the ov-
ervalued exchange rates in the early 1980s typically caused significant
reductions in the demand for the output of manufacturing industries.
Third, many industrial products, particularly consumer durables and cap-
ital equipment, are relatively sensitive to the effects of demand fluctua-
tions. Industrial sectors thus tend to produce goods that are sensitive to
national income contractions. Fourth, in addition to these inherent char-
acteristics of industry, there is evidence that financial structures in the
industrial sector were generally at greater risk in the early 1980s than
they had been in the 1970s.3 Hence the sector was already fragile before
the reforms began, a situation that was exacerbated by the growing re-
luctance of international creditors to provide new funds. The balance of
this chapter is concerned with whether the SAL countries did manage to
insulate their industrial sectors from radical adjustment, how they at-
tempted to do so, and the role of World Bank conditionality in shaping
these policies.
Three Phases of Performance
To examine the short-run costs of crisis and adjustment for industrial
producers, several indicators of performance will be constructed by coun-



Industrial Performance: Some Stylized Facts 161
try and by broadly defined region.4 The indicators will be calculated for
three periods, each chosen to approximate an era in world economic
conditions. The first period, 1979-81, spans a time of moderate growth
in the industrial countries and relatively easy access to private credit in
world capital markets. The second, 1982-83, is characterized by reces-
sion in the industrial countries and a scarcity of private external credit
for developing countries. It corresponds to the advent of serious crises
in a number of the countries studied in this chapter. The third period,
1984-86, is characterized by recovery among the industrial countries
and increased adjustment lending to developing countries. These periods
are of course crude approximations to the turning points in specific coun-
tries, and their interpretation is further confounded by unpredictable lags
in adjustment. Nonetheless, it will be seen that clear patterns emerge.
A brief elaboration of conditions in the SAL countries during each pe-
riod provides useful background for later sections. During the precrisis
years (1979-81) these countries were surprisingly similar. First, trade
protection was typically substantial. Except in Chile, import-competing
producers enjoyed considerable protection through both nontariff and
tariff barriers. Second, except in Kenya, Korea, and Malawi, this pro-
tection was at least partly offset by overvalued currencies. Third, the SAL
countries were generally suffering from the second oil-price shock (Mex-
ico excepted), and their terms of trade problems were typically com-
pounded by deteriorating world prices for their exports.5 Fourth, sig-
nificant price controls were in force in Morocco, Pakistan, the
Philippines, Turkey, and many of the African countries. Fifth, these same
countries tended to have a large public sector and state ownership of
industrial enterprises.
The combination of overvalued exchange rates, terms of trade shocks,
protection, and state intervention made the SAL countries relatively sus-
ceptible to balance of payments problems. These materialized when in-
ternational credit markets tightened and the world moved into recession.
The crises did not occur simultaneously in all countries, but it is generally
reasonable to describe 1982-83 as the period when the greatest impact
was felt. In roughly half the sample the crises were exacerbated by further
terms of trade shocks.6 Two initial policy reactions typified the stopgap
attempts of SAL countries to deal with foreign exchange shortages: ov-
ervalued exchange rates were abandoned and trade protection was in-
creased.'
As countries tried to regain their economic footing, a process of struc-
tural adjustment ensued. World Bank adjustment lending helped ease the
foreign exchange shortage, which by 1984 was severely exacerbated by
high debt-service obligations and the unavailability of private external
funds. World Bank loan agreements included various conditions, espe-
cially with respect to exchange rates and trade policy. In part because of



162 ISSUES OF POLICY REFORM
this conditionality, the structural adjustment period saw further deval-
uations and some dismantling of the new trade barriers that had been
instituted during the crisis years.
Conditionality
Loan conditionality has spanned the entire policy spectrum from mac-
roeconomic policies to those for trade, finance, and the productive sec-
tors. For industry, these sectoral policies pertain to pricing, entry and
exit, subsector planning and restructuring, investment incentives, tech-
nology, marketing, and various others (see table 8-1).
Several patterns of conditionality during fiscal 1982-87 merit mention.
First, "firm" policy conditions have pertained most frequently to trade
and exchange rate policies; firm policy conditions specific to the indus-
trial sector have been relatively less common.' Second, when specific to
industry, conditions have most frequently concerned subsector restruc-
turing, followed by investment incentives and pricing policies (see table
8-1). Other industrial policy areas have been mentioned relatively infre-
quently (none of them has been at issue in more than 10 percent of the
loan packages in the SAL country data base). Third, about 53 percent of
industrial policy conditions have been fully implemented, which is almost
identical to the implementation rate for all provisions for conditionality.
Although regional variations in conditionality are limited, some pat-
terns do emerge. Subsector restructuring and pricing policy reforms are
relatively common in Africa. In Latin America and the Caribbean con-
crete conditionality was relatively limited, with most of the emphasis on
quantitative restrictions, export regimes, and exchange rate policies. Pub-
lic investment and export policies appeared relatively frequently among
the less concrete measures. The relative emphasis in Europe, the Middle
East, and North Africa (EMENA) has been on export promotion; some
trade liberalization has also been required. Macroeconomic and financial
sector conditionality has been relatively light, emphasizing tax policy and
interest rate reform respectively. Industrial conditionality has also been
relatively light in this region. Conditionality in Asia has been heaviest in
the areas of import duties and tariffs, although industrial policy has also
figured rather prominently, particularly with respect to issues of entry
and exit, technology, and subsector restructuring. Macroeconomic and
financial conditions have been less common.9
Manufacturing Growth
Before discussing the manufacturing growth of SAL countries in each pe-
riod, a note of caution is in order. It is not yet possible to assess these
countries' success in realizing the long-run goals of adjustment pro-



Table 8-1. Conditionality and Implementation Affecting Industry in Fifteen SAL Countries
Percentage
of all              Percentage             Industrial conditions fully
Number               industrial               of all                     tmplemented4
of                  sector              adjustmnent
Policy area                                      conditions            conditions             conditions            Percent            Number
Pricing and subsidies                                10                   13.0                   0.7                  80.0                 5
Entry and exit of firms                               4                    5.2                   0.3                   0.0                 3
t    Other regulatory                                      7                     9.1                   0,5                 60.0                 5
Restructuring of subsector and firms                 29                   37.7                   2.1                  65.2                23
Investment promotion and incentives                  17                   22.1                   1.3                  53.9                13
Technology                                            2                    2.6                   0,2                   0.0                 2
Subsector planning (supply and demand)                1                    1.3                   0.1                   0.0                 1
Marketing                                             0                    0.0                   00                    0.0                 0
Other sector-specific policies                        7                    9.1                   0.5                   0.0                 3
Note: Figures are constructed from the ALCin data base of the World Bank's Industry Development Division. For a fuller discussion of their interpretation, see Coogan
(1989).
a. implementation data were not available for all of the conditions. The number of conditions on which the percentages are based is given in the last column.



164 ISSUES OF POLICY REFORM
grams.10 Short-run growth rates, however, as William Steel has stressed
in an internal World Bank paper, reflect largely the extent of capacity
utilization and the associated adjustment costs. They should not be
viewed as reflecting success in adjusting for two basic reasons. First, the
longer-run objectives of adjustment programs-that is, sustained growth
and efficient production-can run counter to short-run objectives related
to easing adjustment costs, especially when major stabilization attempts
are necessary. Second, despite the common elements of SAL country ex-
periences, adjustment has meant different things in different countries.
Some countries have simply been recovering from transitory shocks, such
as world recession and borrowing constraints. Others have been at-
tempting to accommodate more permanent problems, such as secular
deterioration in their terms of trade. And still others have been adopting
new development strategies altogether-typically ones more oriented to-
ward exploitation of their comparative advantage. Accordingly, different
growth paths must be expected, especially in the short run. All that is
known with certainty is the extent of output fluctuations and the short-
run adjustment burden that such fluctuations imply.
REGIONAL PATFrERNS. Table 8-2 presents annualized growth rates for
the fifteen SAL countries by region. The regional figures are averages of
real growth rates by country weighted by base-year country size (see the
appendix for details and countries used in the regional calculations). To
get a crude indication of how well insulated industrial producers were
from macroeconomic fluctuations, I begin with a general examination
of the relation between the growth of gross domestic product (GDP) and
that of industrial value added.
One way to judge the performance of the industrial sector, or its "in-
sulation" from shocks, is to compare manufacturing growth rates with
GDP growth rates. By this yardstick, the most drastic reductions in GDP
and manufacturing growth took place in Africa during 1982-86. More-
over, African manufacturing contracted more rapidly than GDP, despite
the fact that African countries are generally at an early stage of devel-
opment and should be undergoing relatively rapid industrialization.
"Deindustrialization" also took place in Latin America during 1978-81
but was reversed during the adjustment years, so in this region manu-
facturing went from being a lagging to a leading sector. In EMENA, by
contrast, industry was a leading sector during all periods under review.
Asia showed balanced growth until the 1984-86 adjustment period,
when manufacturing emerged as a leading sector. Thus regional perfor-
mance overall suggests that manufacturing producers in Asia and EMENA
were somewhat protected from macroeconomic contractions, while pro-
ducers in Latin America were less protected and those in Africa were the
least protected.



Table 8-2. Average Annual Growth Rates of Manufacturing Value Added and GDP for Fifteen SAL Countries, 1979-81,
1982-83, and 1984-86
1979-81                                1982-83                                 1984-86
Region and
country                  GDP            Manufacturing            GDP            Manufacturing           GDP             Manufacturing
Africa                  -0.01                 0.02              -0.04               -0.06               -0.02              -0.05
C6te d'lvoire         -0.04                 0.00                0.00               0.15                0.03                0.06
Ghana                  -0.12              - 0.22              - 0.03              -0.23                0.13                 n.a.
Kenya                   0.02                0.03              -0.03               - 0.04               0.04                0.03
Malawi                 -0.02                 n.a.               0.02                n.a.              -0.02                 n.a.
Zambia                  0.03                0.04              -0.06               -0.05                0.01                0.01
Latin America             0.05                0.00              -0.02                0.02                0.00                0.03
Chile                   0.07                0.06              -0.06               -0.10                0.04                0.06'
Colombia                0.03                0.06                0.02               0.01                0.07                0.06
,,    Jamaica                -0.09               -0.10                0.05                0.16                0.01               0.05
Mexico                  0.10                0.08              -0.05               -0.04               -0.02                0.02
Asia                      0.07                0.07               0.04                0.03                0.04                0.07
Korea                   0.02                0.03                0.08               0.07                0.10                0.12
Philippines             0.03                0.03                0.01               0.01               - 0.05             -0.05
Thailand                0.04                0.06                0.04               0.01                0.04                0.07
EMENA                     0.03                0.06               0.02                0.03                0.04                0.05
Morocco                 0.01                0.02                0.02               0.01                0.03                0.04
Pakistan                0.06                0.09                0.06               0.08                0.07                0.07
Turkey                  0.05                0.09                0.01               0.06                0.05                0.06
n.a. Not available.
Note: Regional growth rates are weighted averages of real growth rates (denominated in domestic currency) for all countries (see appendix for list of countries) in the
region for which data were available for all three periods. The weights are shares in regional GDP or regional manufacturing value added, depending on which aggregate
is being studied.
a. This growth rate is calculated by using a production (rather than value added) growth rate to infer the data for 1986.
Source: See appendix.



166 ISSUES OF POLICY REFORM
SAMPLE PERFORMANCE. How did the fifteen SAL countries fare? There
are several ways to evaluate their performance on the basis of the data
presented in table 8-2. One possibility is to compare each country's man-
ufacturing growth with the regional average. Such comparisons show
that only half the SAL countries had growth rates above regional norms
during both the 1978-8 1 reference period and the 1982-83 crisis period.
However, ten of the thirteen SAL countries with observable data registered
growth rates above the regional norms during the 1984-86 recovery
period. This pattern is weakly supportive of the view that the manufac-
turing sector was better able to rebound from recession and crisis in the
SAL countries than elsewhere.
A more subtle observation can be made if individual SAL countries are
followed through time. Four countries were above their regional norms
in all periods (Kenya and Zambia, Pakistan and Turkey); three were
below their region in the first period and above it thereafter (C6te
d'Ivoire, Jamaica, and Korea); and three were below their region in the
second period but above it thereafter (Chile, Colombia, and Thailand).
So six of the ten sAL countries showed positive turnarounds relative to
their regions. If adjustment lending were unrelated to country turn-
around, there should also have been about six countries that started above
regional norms and fell below them. Only Mexico did this, however, and
even that was a marginal case.i1
An alternative way to judge the performance of SAL countries is to
compare manufacturing growth rates with GDP growth rates. Since the
industrial sector typically grows faster than GDP in developing countries
(see, for example, Chenery, Robinson, and Syrquin 1986), the difference
between manufacturing and GDP growth in each region can be taken to
indicate the "normal" amount of industrialization taking place. A com-
parison of industrialization in the SAL countries with their respective re-
gional norms shows that of the thirteen countries with available data,
only five (Chile, Colombia, Korea, Pakistan, and Philippines) failed at
least to match the regional rate of industrialization during 1984-86.
During the precrisis period, however, this was the case for a different,
larger set of SAL countries (only Ghana, Kenya, Morocco, and Zambia
failed to match their regional average rates of industrialization). Put dif-
ferently, the SAL countries did manage to industrialize relatively rapidly,
but this was true both before and after adjustment programs were in-
stituted.
Sources of Manufacturing Growth
Adjustment programs are likely to affect the growth of industrial output
through several channels. First, to the extent that real expenditure is
constrained by an adjustment program, the contraction of aggregate de-



Industrial Performance: Some Stylized Facts 167
mand should reduce industrial capacity utilization across the board. Sec-
ond, and counter to this force, to the extent that an adjustment program
eases a foreign exchange constraint, the resultant increased availability
of imported inputs may allow industrial expansion. The foreign exchange
effect can dominate the aggregate demand effect if prices are rigid and
there is substantial excess demand for industrial products.'2 Third, if an
adjustment program involves real devaluation, the effects of expenditure-
switching in product markets should help both exporters and import-
competing producers. Devaluation also increases the price of imported
inputs, however, and so can hurt manufacturing producers.
To gain a better understanding of the roles played by these basic forces,
it is helpful to employ algebraic decompositions of gross manufacturing
output by demand source: domestic market expansion, import substi-
tution (given domestic market size), and export expansion. The expres-
sion used is
(8-1)        AQ, = AD,(1 - mt) - Am,D1  + AX,
where Q, is gross manufacturing output in period t, DT is domestic de-
mand for manufacturing output (for both intermediate and final use), mt
is the share of domestic demand supplied by imports, and X, is the amount
of manufacturing output exported. Bars denote averages of period t and
period t - 1 values.'3
The first term is the portion of output expansion attributable to do-
mestic market expansion, assuming that the import penetration rate, m,
remains constant. This term should pick up general aggregate demand
effects. The second term reflects the effect of changing import penetration
rates on output expansion-note that reductions in import penetration
rates amount to positive import substitution and contribute positively to
output growth. If manufactured intermediate and capital goods are im-
portant components of manufactured imports, however, import substi-
tution may reflect foreign exchange rationing and associated constraints
on output. (When this happens, domestic demand or exports, or both,
should be falling as import substitution rises.) The last term simply re-
flects export expansion. In the presentation of results (see tables 8-3 and
8-4), these three components of output expansion are expressed as ratios
to beginning-of-period output, so they will always sum to the cumulative
rate of output growth for the period:
(8-2)    AQIQ = (AD/Q)(1 - m) - (AI-/Q)D + AX/Q.
Contrasts across periods for a given country will reveal how domestic
and world market conditions contributed to expansion (or contraction)
of the industrial sectors (see table 8-4). In particular, it should be apparent
whether the industrial sector's net contribution to the current account
improved or worsened, whether this change was due to export expansion



Table 8-3. Sources-of-Growth Decomposition for the Manufacturing Sector, by Region, 1979-81, 1982-83, and 1984-86
Summary statistics                                Contribution to total output growth
Output            Import            Export          Domestic market            Import              Export
°°                             growth             share             share              expansion             substitution        expansion
Region and period        AQIQ                -m               X/D            (ADIQ) (1 - -i)          - (Am/Q)D             AXIQ
Africa
1979-81                  0.21             0.40              0.03                  0.11                  0.10              -0.01
1982-83                - 0.12             0.33              0.02                - 0.23                  0.11              -0.01
1984-86                  0.06             0.27              0.02                -0.03                   0.08                0.00
Latin America
1979-81                  0.11             0.16              0.03                  0.15                 -0.04              -0.00
1982-83                -0.04              0.14              0.03                -0.13                   0.09              -0.01
1984-86                  0.06             0.10              0.03                  0.06                 - 0.01               0.01
Asia
1979-81                  0.22             0.21              0.14                  0.17                  0.00                0.04
1982-83                  0.11             0.20              0.14                  0.08                  0.02                0.01
1984-86                  0.19             0.18              0.14                  0.14                  0.02                0.03



EMENA
1979-81                   0.03              0.18               0.06                 -0.02                    0.03                 0.03
1982-83                   0.06              0.16               0.07                   0.02                   0.03                 0.01
1984-86                   0.13              0.16               0.08                   0.15                  -0.04                 0.02
All regions
1979-81                   0.15              0.20               0.08                   0.14                  -0.01                 0.02
1982-83                   0.03              0.18               0.08                 -0.03                    0.06                 0.00
1984-86                   0.12              0.15               0.08                   0.09                   0.01                 0.02
Sample countries
1979-81                   0.18              0.17               0.09                   0.17                  -0.01                 0.02
1982-83                   0.04              0.15               0.09                 -0.03                    0.06                 0.01
1984-86                   0.13              0.13               0.09                   0.12                  -0.01                 0.02
Note: Growth rates (gross output) are cumulative over the periods indicated. (For example, figures for 1979-81 reflect a comparison of 1978 and 1981 values.) Reported
figures are weighted averages of country-by-country figures; the weights are the respective country's share in total manufacturing value added for the group. Further details
ONfi   are provided in the appendix.
'°      Source: See appendix.



Table 8-4. Sources-of-Growth Decomposition for the Manufacturing Sector for Fifteen SAL Countries, 1979-81, 1982-83,
and 1984-86
Summary statistics                              Contribution to total output growth
Output         Import          Export         Domestic market             Import                Export
growth          share          share             expansion               substitution         expansion
Country and period        AQJQ            -m             X/D            (ADIQ) (1 - Wi)            -(Am/Q)D               AXIQ
Chile
1979-81                  0.14           0.24           0.02                 0.22                   -0.08                -0.01
1982-83                 -0.11           0.21           0.02                -0.24                     0.13               -0.00
1984-86                  0.19           0.15           0.02                 0.18                     0.01                0.00
Colombia
1979-81                  0.12           0.17           0.04                 0.13                   -0.02                 0.01
1982-83                 -0.02           0.17           0.03                -0.02                     0.02               -0.02
1984-86                  0.15           0.14           0.03                 0.09                     0.05                0.02
Cote d'lvoire
1979-81                  0.24           0.31           0.04                -0.04                     0.27                0.01
1982-83                  0.08           0.20           0.04                 0.01                     0.08               -0.01
1984-86                  0.21           0.18           0.03                 0.23                   -0.02                 0.00
Ghana
1979-81                 -0.13           0.30           0.01                -0.18                     0.06               -0.01
1982-83                 -0.18           0.23           0.01                -0.28                     0.11               -0.00
1984-86                  0.35           0.19           0.01                 0.36                   -0.01                 0.00



Jamaica
1979-81                -0.15          0.45          0.45               -0.07                  -0.11                0.02
1982-83                 0.05          0.46          0.38                0.12                    0.11              -0.18
1984-86                -0.03          0.37          0.24               -0.04                    0.08              -0.07
Kenya
1979-81                 0.19          0.42           0.05              -0.05                    0.27              -0.03
1982-83                 0.06          0.28          0.04               - 0.09                   0.16              -0.02
1984-86                 0.15          0.25          0.04                0.21                  - 0.08               0.02
Korea
1979-81                 0.25          0.18          0.28                0.14                    0.05               0.06
1982-83                 0.21          0.16          0.26                0.18                    0.02               0.01
1984-86                 0.31          0.15          0.24                0.25                  -0.00                0.07
Malawi
1979-81                 0.12          0.44          0.04               -0.15                    0.24               0.03
s      1982-83                 0.09          0.34          0.04                0.01                    0.10              -0.02
1984-86                -0.01          0.30          0.04               -0.06                    0.02               0.02
Mexico
1979-81                 0.27          0.13          0.02                0.35                  - 0.08               0.00
1982-83                -0.10          0.11          0.03               -0.21                    0.11               0.00
1984-86                 0.04          0.07          0.03                0.05                  -0.02                0.01
Morocco
1979-81                 0.05          0.25          0.07               -0.05                    0.08               0.03
1982-83                 0.10          0.20          0.09                0.03                    0.05               0.01
1984-86                 0.05          0.19          0.10                0.05                  - 0.03               0.03
(Table continues on the following page.)



Table 8-4 (continued)
Summary statistics                            Contribution to total output growth
Output        Import         Export          Domestic market            Import              Export
growth         share          share            expansion              substitution         expansion
Country and period       A QIQ           m             X/D           (AD/Q) (1 - mi)           -(Am/Q)D              AXIQ
Pakistan
1979-81                 0.29          0.29           0.15                0.21                    0.04                0.04
1982-83                 0.23           0.24          0.15                0.13                    0.07                0.04
1984-86                 0.25           0.20          0.14                0.21                    0.03                0.01
Philippines
1979-81                 0.12           0.24          0.10                0.06                    0.00                0.06
1982-83                 0.05          0.22           0.12                0.02                    0.04              -0.01
1984-86                -0.08           0.19          0.12               -0.11                    0.03                0.00
Thailand
1979-81                 0.23          0.16           0.05                0.20                    0.00                0.02
1982-83                 0.12          0.15           0.05                0.11                    0.01              -0.00
1984-86                 0.15          0.14           0.06                0.09                    0.02                0.03
Turkey
1979-81                -0.03          0.11           0.03               -0.04                  -0.01                 0.03
1982-83                 0.11          0.10           0.06                0.06                    0.02                0.03
1984-86                 0.21          0.11           0.07                0.23                  -0.04                 0.02
Zambia
1979-81                - 0.02         0.28           0.00                0.04                  - 0.06              - 0.00
1982-83                 0.08          0.25           0.00               -0.07                    0.15              -0.00
1984-86                 0.05          0.19           0.01                0.04                    0.00                0.01
Source: See appendix.



Industrial Performance: Some Stylized Facts 173
or import substitution, and whether industrial output was closely tied to
domestic demand conditions. Contrasts with average regional patterns
(table 8-3) should also indicate which countries were doing well relative
to other countries in similar circumstances. Finally, once each country's
industrial performance has been assessed, a crude attempt can be made
to associate policy reforms with outcomes.
Since the real exchange rate has fluctuated considerably in a number
of the countries in the sample, it makes a difference whether exports,
imports, and production are expressed in terms of real value (as is typi-
cally done) or as indices of physical volume. For example, real deval-
uation increases the domestic currency value of imports and exports rela-
tive to goods produced and consumed at home. Hence, if the
decomposition is done in value terms, export expansion and increased
import penetration will appear to accompany real devaluation even if
physical trade flows (and the dollar-denominated manufacturing trade
balance) remain unaffected. Because the primary interest here is in
changes in physical resource allocation, not in the valuation effects of
exchange rate changes, I depart from tradition and use indices of physical
volume to implement equation 8-1 (see appendix for details).14
Annualizing the manufacturing growth rates in tables 8-3 and 8-4
would not yield exactly the same growth rates as in table 8-2. The reason
is that in tables 8-3 and 8-4 the figures are growth rates of gross output,
whereas in table 8-2 the figures refer to value added. Qualitatively, how-
ever, the patterns are similar for most countries, but for Cote d'lvoire,
Kenya, Turkey, and Zambia there are substantial contrasts. These con-
trasts imply either that physical input-output relationships changed or
that the relative prices of inputs and outputs changed. Given Africa's
heavy reliance on imported inputs, it is not surprising that most of the
growth discrepancies are found in that region.
GENERAL PAIrTERNS. Several patterns emerge clearly from the results
for all fifty countries included in the regional totals in table 8-3 (see table
8-6 for the list of countries). First, as is typically the case, the expansion
of domestic markets has generally been the most important source of
expanded demand for manufactured products. Second, and more inter-
esting, rapid import substitution took place during the crisis years 1982-
83 when domestic demand contracted. So developing countries generally
responded to the debt crisis and world recession with large cutbacks in
manufactured imports, thereby conserving foreign exchange. This re-
sponse softened the strong contractionary effects of falling domestic de-
mand in countries where recessionary forces were strong, but it probably
meant production bottlenecks in countries where imported inputs were
key. (This issue is picked up again later when regional and SAL country
patterns are examined.)



174 ISSUES OF POLICY REFORM
Third, during 1984-86 import penetration rates remained at the re-
duced levels of the 1982-83 period. (Recall that reductions in protection
during 1984-86 were often accompanied by offsetting devaluations.)
This is probably the most striking structural adjustment to emerge from
the aggregate figures. If it reflected either ongoing foreign exchange short-
ages or attempts to generate product market demand, the consequences
have most likely been damaging: imported inputs were in short supply,
and protection of product markets tended to generate the familiar col-
lection of inefficiencies."5 Fourth, the expansion of manufactured exports
has not been a major source of growth in demand among the fifty coun-
tries. After dropping to near zero during the 1982-83 period, export
expansion rebounded to precrisis levels of about 2 percent.
Surprisingly, the sample of fifteen SAL countries heavily involved in
World Bank adjustment programs performed almost identically to the
full comparator group of fifty developing countries in all periods. Al-
though the SAL countries consistently grew slightly more rapidly than the
comparator group because of more expansion of domestic markets and
experienced slight increases in import penetration during 1984-86, these
contrasts are very minor. Part of the explanation for the close similarity
in performance lies in the fact that the SAL countries are part of the
comparator group, but the similarity and its invariance with respect to
period is nonetheless surprising. One interpretation is that World Bank
lending and conditionality have served to maintain the economic position
of recipient countries in the developing world.
REGIONAL PA=IERNS. The regional weighted averages show more di-
versity (table 8-3). First, the import substitution that was apparent in the
overall figures took place mainly in Africa (during all periods) and in
Latin America (during 1982-83). In Asia and the EMENA region import
substitution during 1982-83 took place on a much more limited scale.
There are several plausible explanations for this pattern. First, the effects
of domestic market contraction were most severe in Africa and Latin
America, so the pressure was greater to generate demand for industrial
output through protectionism.16 Second, the debt crisis was most severe
in these two regions, so relatively severe foreign exchange shortages prob-
ably forced cutbacks in imported inputs. During 1984-86 the import
substitution of 1982-83 was actually undone in the EMENA region, per-
haps because of reductions in barriers to trade in Morocco and Turkey.
But Africa continued to scale down the domestic market share of its
imports at a rapid clip.
Expansion of manufactured exports has been less important than im-
port substitution in generating foreign exchange and demand. Nonethe-
less, there are clear cross-regional contrasts in export performance. In
Latin America and Africa the export expansion of 1984-86 essentially



Industrial Performance: Some Stylized Facts 175
made up the ground lost to export contraction during 1982-83. In con-
trast, the effect of growing export markets was positive and somewhat
more significant during both periods in Asia and the EMENA region. None-
theless, no region gives evidence of a structural shift toward a heavier
reliance on manufactured exports when its experience is compared with
that of the precrisis period.
SAMPLE COUNTRIES VERSUS REGIONAL NORMS. In Africa all sample coun-
tries showed the same strong pattern of import substitution during 1982-
83 that appears in the regional average. They also showed a positive but
small role for exports in the expansion of output during the adjustment
period. However, the African SAL countries did deviate in several signif-
icant ways from the regional norm.17 First, except in Malawi, output
expanded in all African sample countries during 1984-86 by servicing
a growing domestic market for manufactured products. (The regional
average shows a shrinking domestic market.) Second, except in Malawi,
import substitution was not significant during 1984-86. Indeed, sub-
stantial import penetration took place in Kenya. One hypothesis is that
World Bank adjustment lending allowed these countries to resume im-
porting intermediate goods, thereby releasing a constraint on capacity
utilization. (In C6te d'Ivoire, Ghana, and Kenya import penetration was
accompanied by relatively rapid expansion of output.) Another possi-
bility is that adjustment lending allowed (or forced) these countries to
forgo additional increases in protection. Still another explanation is that
these countries had already done as much import substitution as was
feasible by 1984. The relative merits of these perspectives cannot be as-
sessed without additional empirical work; each, however, seems to have
an element of truth.
In Latin America both Chile and Colombia appear to have followed
the regional pattern rather closely. A contraction associated with falling
domestic demand during 1982-83 was offset by import substitution.
Recovery took place during 1984-86, with import penetration rates re-
maining stable. Mexico also followed the regional pattern in 1982-83,
but during 1984-86 it showed increased import penetration and signif-
icant export expansion. So Mexico, more than Colombia and Chile, ap-
pears to have opened up during 1984-86. Jamaica differed from the
regional pattern during both periods. In 1982-83 Jamaica underwent a
large drop in manufactured exports, yet output expanded because of
strong domestic markets and import substitution. Import substitution
continued during 1984-86, but its effects were more than offset by re-
ductions in exports and domestic demand, and output fell.
In Asia both Korea and Thailand followed the regional pattern: rela-
tively rapid output growth, mild import substitution during 1982-86,
and significant export expansion (with a dip in 1982-83). The Philip-



176 ISSUES OF POLICY REFORM
pines, by contrast, deviated considerably from the regional norm with
substantial import substitution during the 1982-86 period and a large
contraction in output in the adjustment years, 1984-86. Hence this coun-
try's reduced reliance on imports reflected either the effects of foreign
exchange scarcity on intermediate imports or an expenditure-switching
policy that was insufficient to prevent domestic recession.
Among the sAL countries in the EMENA region, Morocco and Turkey
followed the regional pattern fairly closely. They both exhibited rapid
export growth coupled with domestic market contraction during the pre-
crisis period of 1978-81, shifted to import substitution during 1982-
83, and experienced a domestic market recovery and export expansion
in 1984-86 that more than offset import penetration. The only notable
distinction between these two countries' performances and the regional
norms is that Turkey showed considerable import penetration during the
precrisis years. Pakistan, however, deviated somewhat from regional
norms by expanding domestic demand during the precrisis years and
continuing gradual import substitution through all three periods.
In summary, the SAL countries whose manufacturing sectors followed
regional patterns were Chile and Colombia in Latin America, Korea and
Thailand in Asia, and Morocco and Turkey in the EMENA region. African
SAL countries deviated from the regional norm by showing much less
import substitution and more demand expansion during 1984-86. The
other countries that deviated from regional norms did so by showing
more import substitution (Jamaica and Pakistan) and/or substantially
lower growth in domestic markets (Jamaica and Philippines), except for
Mexico, which showed increased import penetration and export expan-
sion.
A Link between Conditionality and Performance?
The remaining question is whether any obvious patterns emerge that link
loan conditionality to the performance measures reviewed in the previous
section. Identifying such linkages is a daunting task and one that at best
yields merely suggestive conclusions. In the turbulent environments that
have characterized countries undergoing adjustment, diverse forces have
acted on industrial producers simultaneously. Fiscal and monetary pol-
icies, the real exchange rate, commercial policy, credit market conditions,
price controls, and world market conditions have all influenced the vol-
ume of domestic production and the fraction of this production that was
exported. These same forces have also helped to determine the volume
of imported industrial products and so the contribution of the industrial
sector to the current account deficit or surplus. Isolating the effects of
specific loan conditions, particularly ones that are designed to effect grad-
ual, long-run changes, is not a simple task.



Industrial Performance: Some Stylized Facts 177
To further complicate matters, the industrial sector has in turn affected
the rest of the economy. For example, in instances where inordinate fi-
nancial stress has been placed on firms, widespread problems of illiquidity
or insolvency have rendered much of the financial sector's portfolio non-
performing. Financial crises, capital flight, and abandonment of adjust-
ment efforts have sometimes resulted. Less dramatically, when the in-
dustrial sector has been insufficiently responsive to export incentives,
current account problems have helped to undermine efforts to restore
external balance without protection.
If it is difficult to link policies and performances, it is even more chal-
lenging to link certain policy mixes with World Bank loan conditionality.
First, it is impossible to gauge the extent of many types of reform. (For
example, how binding were quotas or price controls before they were
relaxed?) Second, it is difficult to say what impact conditionality has had
on the policymaking process. In cases where the political will to reform
has been weak but loan conditionality has been explicit and enforceable,
the impact may have been considerable. In other cases, recommended
changes may have gone largely ignored. In still other instances, policy-
makers may have undertaken major reforms regardless of whether they
were conditions of World Bank or International Monetary Fund (wMF)
lending. For all these reasons, this section evaluates the effects of actual
policies, rather than conditionality, on industrial performance.
A crude categorization of reforms in the sample countries is attempted
in table 8-5. The policy reforms that are most important for the man-
ufacturing sector's performance are assumed to pertain to protection,
export promotion, the exchange rate, and industrial policies (as described
earlier). Countries are classified according to whether these policy areas
have been significant dimensions of their reform programs. Performance
is measured according to import penetration (whether negative import
substitution took place) and export expansion (whether exports ac-
counted for more growth during the adjustment period than during the
precrisis reference period); performance on both indicators is taken from
the analysis of sources of growth in the previous section.
As already discussed, import penetration can reflect many things and
may be good or bad. When accompanied by output expansion, increased
import penetration probably reflects relaxation of a foreign exchange
constraint and improved efficiency. By contrast, increases in import pen-
etration that are accompanied by little or no output growth may reflect
a worsening of conditions, either because of an overvalued currency or
(less probably) because of trade liberalization without export promotion
or a compensating devaluation."8 The analysis of import penetration that
follows is therefore limited to an appraisal of this variable's responsive-
ness to policy reforms; judgments regarding the desirability of responses
are not attempted.



178 ISSUES OF POLICY REFORM
Table 8-5. Policy Reforms and Manufacturing Trade Flows for Fifteen
SAL Countries
Emphasis of reforms             Performance
Export             Indus-   Import    Export
Trade    promo-    Deval-    trial     pene-    expan-
Country        reform     tion      uation    policy    tration   sion
Chile             X         X         X                            X
Colombia          X                   XI                           X
C6te d'lvoire     X                                      X
Ghana             X                   X                  X
Jamaica           X         X         X
Kenya             X                                      X         X
Korea             X         X         X         X                  X
Malawi                                X
Mexico            X         X         X                  X         X
Morocco           X                                      X
Pakistan                    X         X
Philippines       X                            X
Thailand          X         X         X         X                  X
Turkey            X         X         X                  X
Zambia                                X                            X
Total             12        7         11        3        6          7
Source: Reform cassifications are based on World Bank country briefs and country studies
prepared for a report on adjustment lending (World Bank 1988).
Several patterns emerge. First, in all six countries in which import
penetration increased, trade liberalization was an important dimension
of the reform package. Moreover, three of these countries did not exhibit
significant real devaluation. If patterns were random, one would expect
to find that only 4.8 countries had liberalized trade and only 1.6 had
not devalued. Although the sample size is too small for these deviations
to be statistically significant, they suggest that trade liberalization was
real and that it tended to increase import penetration. Similarly, one is
tempted to infer that countries avoiding devaluation tended to experience
more import penetration.
Although very weak, there is also some evidence that export promotion
policies and devaluations have encouraged export expansion. Of the
seven countries that expanded exports, four had promotion programs;
the expected number under the assumption that encouragement had no
effect is 3.3. Similarly, six of the seven countries with significant export
expansion devalued, while one would have expected 5.1 countries to have
devalued in the absence of any association between exports and the ex-
change rate. One reason these associations are statistically weak is ob-
viously the small size of the sample. However, several economic forces



Industrial Performance: Some Stylized Facts 179
may also have been at work. First, since expansion of manufactured
exports requires reallocation of productive capacity, lack of confidence
in the long-term policy may have inhibited entrepreneurs from reacting.
Second, the policy reforms themselves may not have been very dramatic.
Third, export performance may have largely reflected world market con-
ditions in the short run.
Summary
Policymakers in developing countries typically view expansion of the
industrial sector as the key to development. The sector also receives care-
ful attention because industrial contraction can lead to urban unem-
ployment and financial disruptions in the short run. So, when external
shocks combined with domestic problems to create economic crises for
many developing countries, policymakers had strong incentives to pursue
adjustment programs that insulated industrial producers as much as pos-
sible from adverse conditions. This chapter has been concerned with how
successful they have been, what policies they have employed, and what
role World Bank adjustment lending has played in this process.
In a sample of fifty developing countries, growth in manufacturing
value added dropped during 1982-83 with the advent of the debt crisis
and world recession. Nonetheless, the manufacturing sector did not
shrink relative to the rest of the economy, so in this sense the adjustment
costs for manufacturers were contained. The exception was in Africa,
where rapid deindustrialization took place. Overall, manufacturing value
added rebounded in 1984-86, but in Africa recovery was very slight and
growth rates remained negative.
A subsample of countries that received substantial adjustment lending
from the World Bank had somewhat better than average growth in man-
ufacturing value added during the period when most adjustment pro-
grams were under way (1984-86). Four of the thirteen SAL countries with
complete data outperformed their respective regions in all periods ana-
lyzed, and another six went from being below average before the ad-
justment period to being above average during the adjustment years.
(Only one sAL country went from above regional norms to below norms.)
With respect to rates of industrialization, the sample tended to outpace
regional averages both before and after 1982-83.
A demand-side sources-of-growth decomposition showed that, overall,
developing countries reacted to world recession and the debt crisis with
substantial reductions in the domestic market's share of manufactured
imports. This probably reflected some combination of product market
demand switching (induced by devaluation and increases in protection)
and reduced use of imported inputs. Regional analysis revealed that this
import-substitution process was most important in Latin America and



180 ISSUES OF POLICY REFORM
Africa, where domestic markets for manufactured goods contracted most
dramatically and shortages of imported input may have limited capacity
utilization. Especially rapid import substitution began in Africa during
the 1982-83 crisis period. Although import substitution continued into
1984-86 in Africa as domestic markets for manufactured products re-
mained stagnant, it was reversed in Latin America as growth resumed
in domestic markets.
Many SAL countries followed the regional patterns. The five African
SAL countries, however, did not exhibit continued import substitution
during 1984-86. Moreover, domestic markets in three of the five coun-
tries grew considerably faster than the regional norm (C6te d'lvoire,
Ghana, and Kenya). One possible explanation is that World Bank lending
released a constraint on imported inputs.
A final issue examined was the role of conditionality in shaping policies
and performance in SAL countries. Among conditions directly affecting
the industrial sector, those relating to trade liberalization and exchange
rate devaluation were stressed relatively frequently; subsector restruc-
turing conditions were also fairly common, especially in Africa. Although
it is unclear what the SAL countries would have done in the absence of
this conditionality, most did devalue and liberalize trade somewhat. In-
dicators of industrial performance were examined for possible correlation
with these changes in the policy regime. Weak evidence was found that
countries in which trade liberalization or devaluations have been im-
portant also tended to exhibit increased import penetration during the
adjustment period.
Appendix
Calculating Growth Rates
All data on GDP and manufacturing value added were taken from the
World Bank Andrex data base and valued in current domestic prices.
Data were converted to constant prices with the use of the domestic
consumer price index (cpi), the only deflator consistently available for
all countries and periods. Once converted to real terms, the annualized
growth rates covering the period t to t + k were constructed for country
i as follows:
(8-3)              gitt+k = (ZirtkIZit)11k - 1
where Z is the variable of interest (GDP or value added).
Growth rates are constructed country by country by using the meth-
odology described above. To get regional growth rates, GR, a weighted



Industrial Performance: Some Stylized Facts 181
Table 8-6. Countries Included in Average Growth Calculations,
by Region
Latin
Africa                    America                Asia               EMENA
Benin                   Argentina             Bangladesh           Cyprus
Central African         Barbados              Fiji                 Greece
Republic              Bolivia               India                Jordan
C6te d'lvoire           Chile                 Indonesia            Morocco
Gambia                  Colombia               Korea               Pakistan
Ghana                   Dominican             Malaysia             Tunisia
Kenya                     Republic            Myanmar              Turkey
Liberia                 Ecuador               Philippines
Malawi                  El Salvador           Singapore
Niger                   Honduras              Sri Lanka
Nigeria                 Jamaica               Thailand
Senegal                 Mexico
Sierra Leone            Panama
Togo                    Paraguay
Zaire                   Trinidad and
Zambia                    Tobago
Zimbabwe                Uruguay
Venezuela
average of country-specific growth rates, is constructed using base-year
shares in total GDP or manufacturing value added as weights:
(8-4)                   GRt ,+k =  2  a'igit.t+k
iER
where R is the set of countries in the region, and cZo = Zio/(E  Z,o). The
jER
countries included in regional growth rate calculations are listed in table
8-6.
Sources-of-Growth Decompositions
Several issues arise when implementing the sources-of-growth decom-
position presented in the text. The first has to do with the treatment of
real exchange rate fluctuations. Specifically, output series are recorded
in domestic currency, while trade flows are in dollars-one must some-
how put these in comparable units and adjust for inflation. The usual
procedure is to convert trade flows to domestic currency and to infer a
domestic currency value for domestic demand. Or, defining q, d, x, and
m to be physical quantities of gross manufacturing production, domestic



182  ISSUES OF POLICY REFORM
demand, exports, and imports, respectively, one infers the nominal value
of domestic demand as:
(8-5)             Pdd= -P$m- TrP$x + Pq
where
pd=(        q     )      (  m -       x   )P$
= (q + m - x)        (q + m - x
Here pd is the average price paid by domestic users of manufactured
goods, -r is the exchange rate (units of local currency per dollar), P is
the average price received by producers of domestic manufactured goods,
and P$ is the world price of manufactured goods expressed in dollars.
All terms in equation 8-5 are then deflated with a domestic price de-
flator to obtain the variables necessary to implement equation 8-2 of the
text: M  = (TrP$IPd)m; X = (zrp$/pd)X; Q = (P-Pd)q; and D = d =
(1rP$1Pd)m - (.TrP$IPd)x + (PIpd)q. This procedure is acceptable if rela-
tive prices remain fairly stable. But major devaluations can cause large
changes in the real exchange rate (proxied by Nps/pd), so that the mag-
nitude of measured exports and imports (X and M) expands relative to
domestic demand and production. In other words, export expansion and
import penetration are automatically generated by a devaluation in this
accounting framework, even if the physical flow of goods is unaffected.
To look only at changes in physical flows, the present study employs
an alternative approach. Specifically, measured values for imports, ex-
ports, and production are deflated by their own price indices, and then
used to calculate the variables in equation 8-2: M = (P$m)/PW; X =
(P$x)IPW; Q = (Pq)/P'; and D = (P$m)/Pw - (Psx)/Pw + (pq)/PC.
Here Pw is an index of the average world price (expressed in dollars)
for traded manufactured goods, and pc is a domestic manufacturing price
index. All price indices except the exchange rate are initialized to a value
of 1.0 in the base year (1978).
Regional and Aggregate Sources of Growth
To construct regional and aggregate sources of growth, the decompo-
sition described above was done for each country and period. Then, for
a given period (say, period t - k to t), each term in equation 8-2 was
weighted by the country's base-year (1978) share in regional manufac-
turing value added (a,), and summed over all countries in the relevant
group. The weighted average sources-of-growth decomposition was thus:
(8-6)  s c,(AQitlQiQ-k) = X cxi(ADi,t/Qit_k)(1 - m,t)
- Y cxi(AmitDit/Qit-.k) + X 0ti(AXit1Qit-k)-



Industrial Performance: Some Stylized Facts 183
A sources-of-growth decomposition could not be constructed for several
of the countries listed in table 8-6, although it was possible to construct
one for several countries not listed there. So the set of countries used in
calculating equation 8-6 differs slightly from the set of countries listed
in table 8-6 by the exclusion of Benin, Liberia, and Panama and the
inclusion of Egypt, Guatemala, and Portugal.
Data Sources
The first data problem encountered is that Andrex does not include gross
manufacturing production data. The World Bank's economic and social
data base (BESD) does provide gross manufacturing production data for
most countries, but the series does not extend to 1986 and is often based
on establishments above a certain size (for example, with more than fifty
workers) and does not correspond closely to production indices directly
available from the countries. The BESD series was thus used only to get
an approximation to the base-year (1978) domestic currency value of
production."9 The base-year production series was then used to generate
a series on real manufacturing production for 1978-86, with the use of
the manufacturing production index available from Andrex. (In several
spot checks, this series looked very close to the series available directly
from the countries.) The only problem with this approach was that the
production index stopped in 1984 or 1985 in some countries. When this
occurred, it was brought forward to 1986 with the use of manufacturing
value added growth rates for the missing years. When value added was
itself not available through 1986, the country was left out of the analysis.
The second problem was that manufacturing price deflators were gen-
erally not available for individual countries, nor was there a world (dol-
lar-denominated) price index for manufactured goods. The former was
approximated with domestic cPIs, the latter with the U.S. cPi.
The final data problem was one of consistency between trade and
production data. Trade data are collected according to the Standard In-
ternational Trade Classification (sITc) system, while production and
value added data are available only by the International Standard In-
dustrial Classification (isic) system. Differences in the systems make it
impossible to get a strictly conformable set of production and trade flow
statistics.20
Several solutions to this problem were tried. First, detailed (five-digit)
SITC data were taken directly from the Geneva System and aggregated to
get an approximation to the isic definition of manufactured products.
This approach had the most appeal conceptually; unfortunately, trade
data on certain products were missing for some countries for recent years.
So the approach worked well in certain countries but was unacceptable
as a general procedure. Next, data on manufacturing trade were taken



184 ISSUES OF POLICY REFORM
from the Andrex data bank. (These data originated in the Geneva System,
but their "holes" have been filled with extrapolations based on regional
trends.) The main problem with this approach is that Andrex defines
manufacturing more narrowly than does the isic system for purposes of
trade statistics.21 Hence the Andrex trade data do not suffer from sudden
drops because of missing data, but neither do they conform to the defi-
nition of manufacturing used for the available production series. In an
attempt to get the best of both systems, Andrex trade data were aug-
mented by data taken directly from the Geneva System on the manufac-
turing categories not included in the Andrex system. This hybrid system
yielded figures of the correct order of magnitude and was less sensitive
to problems of missing data. Like the pure Geneva data, however, the
resultant figures were subject to holes in certain cases and proved un-
acceptable for comparative exercises.
Since it was necessary to work with Andrex trade data, the reason-
ableness of these figures was examined for a sample of countries for which
all three options could be tried (that is, where missing data did not appear
to be a problem). It was found that although exports are understated
considerably by the omission of food and mineral products, trends are
generally correct. Imports were much less problematic because devel-
oping countries import few primary products (except oil), processed or
otherwise, and other manufactured products are captured by the narrow
Andrex definition.
Notes
I wish to thank Bela Balassa, Jaime de Melo, Mieko Nishimizu, and especially
William Steel for many helpful comments on early drafts. I am also indebted to
Julie Stanton for excellent research assistance and to Diane Coogan for generous
assistance in preparing the conditionality data reported herein. Any remaining
errors are naturally my own responsibility.
1. These fifteen countries are Chile, Colombia, C6te d'lvoire, Ghana, Jamaica,
Kenya, Korea, Malawi, Mexico, Morocco, Pakistan, the Philippines, Thailand,
Turkey, and Zambia.
2. In developing countries, growth in industrial productivity tends to outpace
that of the rest of the economy (see, for example, Chenery, Robinson, and Syrquin
1986). This assertion does not appear to hold for the developed countries.
3. In almost all developing countries for which data are readily available,
corporate leverage grew and liquidity fell between 1975 and 1981.
4. These indicators are specific to manufacturing. Ideally, the analysis would
include the nonmanufacturing industries, mining and construction, but these sec-
tors are quite distinct from manufacturing and would require separate treatment.
The scope is therefore limited to manufacturing, which is typically the target of
attempts to promote changes in product mix, volume of output, and trade flows.
5. There are a few exceptions. For example, in Morocco phosphate prices
were beginning to improve in 1980.
6. Chile, Ghana, Jamaica, Kenya, Pakistan, the Philippines, and Zambia en-
dured further deterioration in their terms of trade.



Industrial Performance: Some Stylized Facts 185
7. Chile, Ghana, Kenya, Mexico, Morocco, Pakistan, the Philippines, and
Zambia had significant real devaluations; only Colombia and Thailand registered
significant revaluations. Exceptions to increased protectionism were Kenya,
Korea, and Turkey, which underwent significant trade liberalization.
8. A "firm" condition is one that is quantifiable or can be evaluated objec-
tively. The characterization of exchange rate policy as one of the policies most
frequently covered by conditionality requires some clarification. In looking at
World Bank loan data alone, exchange rate conditionality does not appear to
be very common. But, as the 1988 World Bank report on adjustment lending
explains, as a "result of the IMF's responsibility for exchange rates, [condition-
ality] figures understate the importance given by the World Bank to movements
in exchange rates. In the context of adjustment programs, governments are usu-
ally expected to establish and maintain exchange rates that are competitive in-
ternationally, but in only a few cases (such as Chile, Colombia, and Ghana) has
there been explicit conditionality" (World Bank 1988).
9. The infrequency of macroeconomic and financial conditions in Asia has
been interpreted to reflect the absence of problems rather than shortcomings of
the lending process (World Bank 1988). In Africa, however, it has been argued
that more might have been done to deal with high wages and low productivity.
In general, it has been suggested that there was insufficient attention to industrial
policy (pricing, regulations, incentives, and technology) and to the financial sector
(World Bank 1988).
10. Such an exercise would involve detailed microeconomic analysis of in-
vestment patterns, industrial productivity, and competitiveness. Several research
projects are now under way at the World Bank that focus on these issues.
11. Mexico came very close to the "normal" growth rate for Latin America
during the recovery years. It might be objected that this type of comparative
exercise is invalid because the countries that received adjustment assistance were
not randomly selected from their respective regions. If the inferences discussed
above are to be invalidated, however, it must be argued that the World Bank,
wittingly or not, systematically chose to lend to countries with the best prospects
for accelerating their growth. I am unaware of convincing evidence that this
occurred.
12. This type of effect is formalized in the fix-price temporary disequilibrium
literature (for example, Gunning 1983 and Standaert 1985).
13. This expression, known as a direct decomposition, is based on the identity
Q = D(1 - m) + X.
14. For purposes of comparison, demand-side decompositions based on real
domestic currency values were also done. Generally, the effects of devaluation
on the relative magnitudes of domestic production and trade flows tended to
influence the findings heavily.
15. There is, of course, the possibility that sustained import substitution rep-
resented displacement of foreign producers by efficient domestic producers-
some have argued that this happened in the Chilean textile industry, for example.
On average, however, the evidence suggests that import substitution leads to
lower growth in factor productivity than other forms of demand expansion (Nish-
imizu and Robinson 1984) and disperses productivity growth much more across
sectors (Nishimizu and Page 1987).
16. There was a notable twist in the composition of imports away from con-
sumer goods during 1980-86 in a sample of thirty-seven countries receiving
World Bank lending for trade reforms (see chapter 4, table 4-4). Although com-
parable data have not been compiled for the SAL countries in this study, it is
reasonable to suppose they followed a similar trend.



186 ISSUES OF POLICY REFORM
17. One reason might be that Nigeria, which accounts for about 50 percent
of manufacturing value added in the region, heavily influences this average pat-
tern.
18. 1 am grateful to Bela Balassa for stressing the ambiguities of import pen-
etration as a performance measure.
19. When even this was not possible, base-year manufacturing value added
from Andrex was multiplied by 2.5 to get an approximate base-year production
figure, since the ratio of value added to production was very close to 0.4 in
countries where both were observable.
20. For example, foods are considered manufacturing in the isic system if they
have been processed. But some detailed SITC categories include both processed
and unprocessed varieties of certain foods, which makes it impossible to isolate
manufactured food exports. Similar problems arise with other primary products,
such as copper and wood.
21. For example, Andrex excludes processed food products.
References
Chenery, Hollis, Sherman Robinson, and Moshe Syrquin. 1986. Industrialization
and Growth. New York: Oxford University Press.
Coogan, Diane. 1989. "Conditionality and Implementation Data from Adjust-
ment Loans: An Analytical Guide to the ALCID." Industry Development Di-
vision, Industry and Energy Department, World Bank, Washington, D.C.
Gunning, J. W. 1983. "Rationing in an Open Economy: Fix-Price Equilibrium
and Two-Gap Models." European Economic Review 23 (1):71-98.
Nishimizu, Mieko, and John Page. 1987. "Economic Policies and Productivity
Change in Industry: An International Comparison." Industry Department,
World Bank, Washington, D.C.
Nishimizu, Mieko, and Sherman Robinson. 1984. "Trade Policies and Produc-
tivity Change in Semi-Industrialized Countries." Journal of Development Eco-
nomics 16 (1 and 2):177-206.
Standaert, Stan. 1985. "The Foreign Exchange Constraint, Suppression of the
Trade Deficit, and the Shadow Price of Foreign Exchange in a Fix-Price Econ-
omy." Journal of Development Economics 18 (1):37-50.
Tybout, James R. 1986. "A Firm-Level Chronicle of Financial Crises in the South-
ern Cone." Journal of Development Economics 24 (2):371-400.
World Bank. 1988. Adjustment Lending: An Evaluation of Ten Years of Ex-
perience. Policy and Research Series 1. Washington, D.C.



Industrial Performance: Some Stylized Facts 187
Comments
Mieko Nishimizu
THOSE WHO HAVE WORKED on policy conditionalities have probably
learned that one must write words down very precisely and, if possible,
quantify them. James Tybout's chapter focuses on two terms-"indus-
trial performance" and "structural adjustment"-and I am not certain
whether the author and I understand these terms in exactly the same way.
First, the term "industrial performance." In answer to the question,
"What is industrial performance?" any manager of an industrial com-
pany will respond, "To make the best and the cheapest." A little more
honest answer would be, "Well, sometimes to make the best, sometimes
to cut unit cost, sometimes both." In economic terms this means pro-
ductivity change, or in economic jargon it means total factor productivity
change, if quantity and quality changes can be measured accurately. So
industrial performance is a dynamic concept. Tybout's study, however,
uses demand-side sources-of-growth decomposition to represent perfor-
mance. This methodology is a static one and hence limits the analysis
substantially.
Second, the term "structural adjustment." I am not sure about Tybout's
time horizon with respect to structural adjustment, so I will exaggerate
to make the point. In the short run there are quasi-fixed inputs, so that
short-run industrial performance depends on capacity utilization, what-
ever the cause. If demand is the cause, business managers would have to
find new markets, including exports. With the right financing or financial
leveraging to tide a business over, it would survive. Otherwise, the com-
pany would go under. The supply-side phenomenon of more firms per-
forming and surviving instead of going under is an important part of the
stabilization-with-growth or short-run adjustment-with-growth scenario
at the microeconomic level. Of course, one must keep in mind that there
are causes other than domestic absorption for varying capacity utilization
in industry, particularly in developing countries.
In the long run all inputs are variable, and technology changes as well.
Industrial performance, reflected (one hopes) in profits, is then part of
the long-term transformation of the structure of production. It is part of
the natural process of shifting resources from less productive to more



188 ISSUES OF POLICY REFORM
productive activities, along with all the alterations in taste and compo-
sition of demand that accompany rising income. The good performance
of fast-growing firms continuously redefines the comparative advantages
of the economy over time-the process by which the industrial sector of
developing countries catches up with the rest of the world. But there are
obstacles preventing this natural process from occurring-inadequate
economic policies being among the most important ones.
I mention these points because Tybout's chapter asks a very important
question: Do policies matter for industrial performance and structural
adjustment? The chapter focuses primarily on trade policy, which has
both short- and long-term implications, and the policy discussion suffers
from the ambiguous time frame concerning structural adjustment. Al-
though the chapter is quite indecisive about whether polices do make a
difference, from the perspective of the battlefield of industry there is no
question that policies matter. The important question is how.
In the development literature a number of arguments have been ad-
vanced concerning the relationships among development policies,
growth, and productivity change. A substantial body of literature ad-
dresses the appropriate use of trade policy to increase growth and pro-
ductivity. Perhaps equal attention has been given to the role of market
mechanisms in guiding the allocation of resources and providing incen-
tives for disciplining costs and improving productivity.
One obvious link between trade policy and productivity is the positive
association between output growth and productivity change. In devel-
oping countries the size of the market has long been viewed as an im-
portant determinant of growth and structural transformation. The ex-
istence of economies of scale implies that a widening of the market
through trade should lead to reductions in real production costs. In the
context of an outward-oriented (or export-led) development strategy, this
argument is usually cast in terms of the benefits of increased demand
through expansion of exports, but it applies equally to import substi-
tution in large domestic markets.
The literature on foreign exchange constraints suggests a second link
between trade policy and productivity. This link exists because in many
developing countries imported intermediate and capital goods are not
easily substituted by domestically produced ones. Policies that limit the
availability of these goods reduce the potential for domestic producers
to utilize capacity fully and therefore reduce productivity in the short to
medium term. The increased availability of foreign exchange, however,
can raise capacity utilization and lead to short-term increases in total
factor productivity.
A different type of trade policy argument is related to the effects of
international competition on the efficiency of domestic producers. It is
possible that the "challenge-response" mechanism implied by interna-



Industrial Performance: Some Stylized Facts 189
tional trade forces domestic industries to adopt new technologies, to
improve efficiency, and generally to reduce costs and improve quality
wherever possible. According to this argument, both export rivalry and
import competition are beneficial to productivity. Although an import
liberalization policy may restrict the demand for domestic goods, such
a policy also raises competition and therefore improves production ef-
ficiency. The counterargument is also widely asserted: protectionist pol-
icies designed to promote import substitution can have dynamic negative
effects on efficiency (beyond the allocative losses) arising from reductions
in competitiveness and increased inefficiency in production.
Care must be taken, however, not to overstate these arguments re-
garding the benefits of international competition. Excessive response to
export promotion policies, for example, may distort incentives and lead
to rising inefficiency in export production. There is also the infant in-
dustry argument for according protection to high-cost industries that
cannot compete immediately but are expected to become internationally
competitive.
With regard to the loss in efficiency caused by protectionist policies,
however, there is logical clarity on one issue: the instruments by which
protection is afforded to domestic industries can affect productivity quite
apart from the levels of protection granted. Quantitative restrictions on
imports break the link between domestic and international prices and
thus insulate domestic producers from movements in international prices.
Such an import regime will permit rising costs to be passed forward in
a fashion that is not possible under a system based on equivalent tariff
protection. Changes in international prices convey substantial informa-
tion on productivity trends in the rest of the world, but instruments that
interfere with the transmission of these prices can make the productivity
performance of competitors irrelevant and can lead over time to a de-
terioration in comparative advantage.
Moreover, the relationship between resource allocation and produc-
tivity performance depends on the nature of the planning and regulatory
practices that are applied in such economies. Theoretically, it is possible
for the planners or regulators to "simulate the market" so that the dis-
tribution of productivity growth rates will be very like that in a market
economy. In practice, however, it is humanly impossible for planners or
regulators to have ideal information or models that continuously replicate
market solutions, particularly in large and complex economies. As such,
planned or highly regulated economies tend to generate incentives related
to averages that are easily observed as opposed to marginals that are
more difficult to see. Thus nonmarket economies may differ significantly
from market economies in both the level and distribution of their pro-
ductivity growth rates.



190 ISSUES OF POLICY REFORM
The policy arguments discussed above are not mutually exclusive, how-
ever. Nor are the postulated effects of policy choices on productivity
performance independent of each other. In the real world one observes
the net effect of a combination of these phenomena simultaneously, so
that discriminating among policies and their consequences is difficult. In
addition, one must always be cautious in implying the direction of cau-
sality from a statistical association.
Perhaps with a more careful distinction between short and long term
and more attention to the nature of the reforms in each country, this
chapter would have presented a more convincing argument about as-
sociations-at best-between policy changes and industrial perfor-
mance. This brings me to my penultimate point.
Before asking whether policies mattered, the chapter does not examine
exhaustively whether the reforms in the countries with adjustment loans
were good ones. Such an examination of the design of policies that formed
the various reform packages would have benefited the study, as would
a more systematic investigation of the appropriateness of the macro-
economic framework in each country and the sequencing of reforms.
A last point: it is a fact of life in industry at the microeconomic level
that improving performance is essentially a highly complex and multi-
faceted management problem. How companies transfer and master tech-
nology, how managers train workers and engineers, how maintenance
functions are performed, how the company's financial programming and
planning are conducted, and similar practices are all important parts of
productivity performance, and they ultimately boil down to the quality
of human capital called "good managers." Good managers, in turn, are
scarce human resources indeed in many developing countries. Therefore
good policies may be necessary to challenge industry to perform, but they
need not be sufficient to induce supply response.



PART II
Evaluation of Adjustment
Programs






9 Introduction to Part II
THE EVALUATION of adjustment lending operations would be relatively
straightforward if the objective were simply to determine the extent to
which agreed-on policy actions had been implemented. The task would
involve consideration and assessment of a few key loan conditions-say,
four or five-within the monitorable component of the action program.
Since these conditions refer to policy intentions or actions rather than to
performance, the task could, at least in principle, be more easily under-
taken for World Bank adjustment loans than for International Monetary
Fund (IMF) operations. In evaluating IMF operations one cannot avoid the
vexatious problem of how to disentangle the influence of program and
nonprogram factors.
Such an evaluation, however, would fail to address the broader and
fundamental question of how effective adjustment lending operations
have been in enabling the World Bank to assist developing countries in
attaining long-term growth and development. Support of these objectives
has, of course, always been a major part of the World Bank's mandate.
To execute this mandate throughout the 1950s, 1960s, and 1970s, the
World Bank financed development projects. In the 1980s it gradually
added fast-disbursing financing through its adjustment lending facilities,
both in support of balance of payments needs and as a catalyst for policy
and institutional reforms. The acute deterioration of the external pay-
ments position of many developing countries in the early 1980s was the
main rationale and impetus for the introduction of adjustment lending.
Since then, however, there has been a growing awareness of the com-
pelling need for policy and institutional reforms at both the macroeco-
nomic and sectoral levels as a means of adjusting to the changed global
environment of the 1980s and of resuming long-term sustained growth.
Thus adjustment lending operations have become the main vehicle of
policy dialogue between the World Bank and its developing country mem-
bers.
The chapters in Part II evaluate three distinct but related aspects of
adjustment programs: (1) the degree of policy change that has been as-
sociated with conditionality in the loan agreements, (2) the link between
policy reforms and macroeconomic performance, and (3) the social con-
193



194 EVALUATION OF ADJUSTMENT PROGRAMS
sequences of adjustment programs. These are not the only relevant as-
pects of adjustment lending operations, nor does the discussion touch on
the broader issue of the evolution of the role of the World Bank in the
process of global adjustment and intermediation. The chapters cover the
major areas of current concern and debate, however, and provide a useful
starting point.
The tradeoffs inherent in the adjustment process pose some important
questions for the World Bank. How should loan conditionality be for-
mulated so as to cover both macroeconomic and structural aspects in
view of the fact that the underlying structure of the economy is an object
of reform?1 Should loan conditions be tied to policy intentions or to
performance? How can conditions related to short-term macroeconomic
requirements be integrated with those related to institutional and policy
reforms, which are by nature long-term processes?
In chapter 10, William McCleary assesses the nature of the conditions
attached to adjustment loans and their implementation record. Although
his data for the sample of fifteen countries with adjustment loans do not
permit him to determine whether conditionality was well designed, they
do permit him to describe the kind of policy measures that were agreed
on and to assess the extent to which they were implemented. This in-
formation allows several conclusions about the conditions that have been
the most successfully implemented and those that have been the most
difficult to carry out. An important conclusion is that future adjustment
operations would probably benefit from reducing the number of con-
ditions in loan agreements.
In chapter 11 Riccardo Faini and associates undertake a cross-country
analysis to evaluate the macroeconomic impact of adjustment programs.
They employ the familiar "before-after" and "control-group" method-
ology that has also been intensively applied in the evaluation of IMF-
supported programs.2 In addition, they provide quantitative information
regarding the macroeconomic impacts of World Bank-supported ad-
justment lending programs. Together with the information provided by
the previous studies of IMF programs, these results should provide a useful
basis for evaluating the joint impact of the Bretton Woods institutions
on the economic performance of loan recipients.
The authors draw on a large set of data gathered from ninety-three
developing countries, especially the thirty countries that received their
first adjustment loan before 1985. This large sample gives them several
years of observations and a conveniently large group of comparator coun-
tries against which to assess the performance of loan recipients. Their
findings, however, are affected by several methodological and statistical
pitfalls inherent in this type of exercise.
First, because the set of adjustment loan recipients is not a random
sample, the estimated coefficients of the impact of adjustment lending in



Introduction to Part II 195
recipient countries relative to changes in nonrecipient countries are sta-
tistically biased.3 Second, macroeconomic performance may reflect the
influence not only of reforms stipulated in adjustment programs but of
other factors as well. For example, the positive effect of reform may be
muted by a negative external shock, or a country's performance may
improve without any policy change in the presence of a positive shock.
Third, countries are likely to exhibit different degrees of responsiveness
to policy measures, with responsiveness itself often being a function of
policy reform. The supply response to policy change has been found to
be less strong in low-income countries than in middle-income countries,
for example. Fourth, the reliability of macroeconomic indicators, that is,
of their informational content, is in question when the underlying struc-
ture, policy, rules, and regimes are themselves the focus of and subject
to change. Most traditional analyses of macroeconomic policy are based
on the implicit assumption that the underlying structure is given. The
complexity of evaluating adjustment programs lies in the fact that the
underlying macroeconomic structure can also be the focus of change
under the program and so may be continually changing.
At a more fundamental level, adjustment processes involving policy
reform and institutional change inevitably have implications for both
intertemporal resource allocation and sociopolitical factors. First is the
tradeoff between present costs and future benefits, which needs to be
understood and taken into account. This distinction, however, is seldom
made explicitly. Second is the potential for social conflict and a shift in
income distribution, which needs to be resolved both politically and eco-
nomically.
The social cost of adjustment is receiving increasing attention in the
design and implementation of adjustment lending operations, as Elaine
Zuckerman demonstrates in chapter 12. She provides a general analysis
of the distributional implications of adjustment programs. After exam-
ining the experience of adjustment programs in which little was done
explicitly to direct social expenditures toward the poor, Zuckerman con-
cludes that more attention should be placed on alleviating the adverse
impact of adjustment programs on the poor. What actions should be
undertaken, however, will remain controversial for a long time not only
because it is difficult to identify the poor accurately by occupation and
socioeconomic characteristics, but also because little is known about the
effectiveness of special targeting of services and programs.
Nevertheless, the distributional implications of adjustment programs
will remain a key issue in the design of adjustment programs, if only
because sustainability is likely to be jeopardized if the distributional im-
plications are too severe. Since the benefits of adjustment programs are
often slow in appearing, other ways must be found to ensure that the
program is sustained. Closely considering the distributional implications



196 EVALUATION OF ADJUSTMENT PROGRAMS
of the proposed measures is one way to support sustainability or at least
to avoid undermining it.
Notes
1. For an insightful and historical discussion of these issues in the context of
IMF conditionality, see Sidney Dell, On Being Grandmotherly: The Evolution of
IMF Conditionality, Princeton Essays in International Finance 144 (Princeton,
N.J.: Princeton University, 1981).
2. For a discussion of this issue, see Morris Goldstein and Peter J. Montiel,
"Evaluating Fund Stabilization Programs with Multi-Country Data: Some Meth-
odological Pitfalls," IMF Staff Papers June 1986):344-64.
3. For an excellent recent survey of literature on evaluating iMF-supported
programs, see Mohsin S. Khan, "The Macroeconomic Effects of Fund-Supported
Adjustment Programs: An Empirical Assessment," IMF Working Paper WP/88/
113 (Washington, D.C.: International Monetary Fund, 1989).



10 The Design and Implementation of
Conditionality
William A. McCleary
BETWEEN 1980 AND 1987 the World Bank made adjustment loans totaling
US$15 billion to fifty-one countries. Of this total, fifty-one structural
adjustment loans (SALS) accounted for US$5.90 billion and seventy sec-
toral adjustment loans (SECALS) for US$9.35 billion. Almost half the num-
ber of loans and credits for adjustment went to Sub-Saharan African
countries, but because of the relatively small size of individual operations
in Africa, the total amounts lent to countries in Latin America and the
Caribbean and in Europe, the Middle East, and North Africa (the EMENA
region) were larger.
Throughout the period of adjustment lending, the World Bank has
conducted internal reviews to evaluate how well adjustment programs
were working and to draw lessons from the experience. Relatively in-
formal progress reports were prepared in 1982, 1984, and 1986. Also
in 1986 the World Bank's Operations Evaluation Department conducted
a more substantial examination of fifteen SALS for which audit reports
had been completed (World Bank 1986), and in 1988 the Strategic Plan-
ning and Review Department surveyed experience to date (World Bank
1988b) and posed some questions for consideration in the report on
adjustment lending that was completed later that year (World Bank
1988a). Virtually all these reviews complained that not enough time had
passed to allow an evaluation of the full effects of adjustment programs,
that it was difficult to disentangle policy changes arising from adjustment
loans from those stemming from other sources, and that the effectiveness
of programs was hard to judge when countries were being subjected to
external exogenous shocks, such as slow world growth or drought.
Having acknowledged these difficulties, the evaluation reports were
still able to reach a number of broad conclusions:
* Implementation performance for adjustment loans was generally sat-
isfactory but ranged from successful (for example, C6te d'lvoire,
Ghana, Korea, Thailand, and Turkey) to moderately successful (for
example, Pakistan, Philippines, and Senegal) to outright failure (for
example, Bolivia, Guyana, and Zambia).
197



198 EVALUATION OF ADJUSTMENT PROGRAMS
* In early adjustment loans the World Bank was often guilty of over-
optimism with respect to government commitment and capacity to
implement; loans were too complex and took longer to implement
than expected.
* Government commitment to an adjustment program was more im-
portant to implementation than was external pressure.
* Providing greater attention to institutional reforms in order to build
up government capacity took time and required specialists with ex-
pertise that differed from that required to supervise policy reforms.
Institutional support was likely to be better handled through free-
standing projects than in the context of adjustment loans.
* Reforms would be more likely to be sustained and the commitment
of government would be more likely to be manifested if adjustment
loans were given more on the basis of action already taken and less
on promises of future actions.
This chapter attempts to push the work of the previous reviews a bit
further by providing more precise answers to three questions. How well
has conditionally been implemented? Does performance in implementing
conditions vary substantially across the various policy areas? What are
the most important determinants of this performance? Each of these ques-
tions is addressed in turn after a brief discussion of the structure of World
Bank-supported adjustment program's and of the important differences
in programs, especially those for countries in Africa and for highly in-
debted countries. The chapter draws on evidence on implementation for
fifty-one adjustment loans in fifteen countries.
The Nature of Adjustment Programs
Both SALs and SECALS are used in support of adjustment objectives. SALS
have generally supported programs to increase domestic resource mo-
bilization and improve efficiency. They do this through reforms of trade
policy, changes in prices and regulations in key productive sectors, in-
creases in government revenues, and changes in the size and composition
of government expenditures. SECALS, though more narrowly focused, nor-
mally support the same broad objectives with measures that can have
considerable macroeconomic impact. These measures include reforms of
trade policy or the financial sector; reforms of sectoral pricing, institu-
tions, and the composition of public expenditure; and still more narrowly
focused measures such as those affecting input pricing (elimination of
fertilizer subsidies, for example). Both SALS and SECALS have also sup-
ported institutional changes to strengthen a government's capacity to
plan and implement reforms, for example, through programs to improve
project evaluation, tax administration, or the monitoring of public en-



The Design and Implementation of Conditionality 199
terprises. SALS are distinguished from SECALS more by the breadth of policy
and institutional reforms in a particular operation than by the nature of
the reforms themselves.
Government programs supported by adjustment lending have tended
to have three distinct elements. First, the programs set out the broad
objectives to be achieved in the medium term. These include targets for
the key macroeconomic variables (such as savings rates, current account
deficits, or aggregate or sectoral growth rates) and for important sectoral
variables (such as crop diversification and the development of export
industries or domestic sources of energy). Second, the programs lay out
a broad set of measures to be introduced in the medium term (including
those that might be supported by future SALS or SECALS) in support of
these objectives. Examples are measures to increase the scope of taxation
or cost recovery or to improve incentives for exports or agricultural prod-
ucts. Third, the programs detail a set of specific actions that the gov-
ernment plans to take over twelve to eighteen months and that the World
Bank will monitor.
To facilitate monitoring and ensure that progress is appropriate, the
loans generally are disbursed in tranches. Disbursements depend on sat-
isfactory compliance with key conditions and with implementation of
the program in general. The World Bank is increasingly relying on more
specific tranche conditions rather than more general statements of in-
tentions. All SALS and 87 percent of SECALS were tranched (included
tranche-release conditions) in fiscal 1986-88, whereas 83 percent of SALS
and 8 percent of SECALS were tranched during fiscal 1979-83. The SECALS
that have not been tranched in recent years have tended to be small repeat
operations or those in which conditionality has been "up front," or
"front-loaded"; that is, countries are required to undertake most of the
conditionality before the first disbursement of the loan.
Differences in Adjustment Programs among Countries
It was clear from the outset of adjustment lending that programs would
vary among countries because the problems differ. The need for stabi-
lization, the relative importance of various sectors, and the nature and
degree of distortions are different. In addition, political considerations
and institutional constraints affect the types of reform that can be at-
tempted and the speed with which they can be planned and implemented.
Moreover, the World Bank's knowledge and preparation, and so the
status of its dialogue with governments, varies substantially among coun-
tries and even between sectors in the same country.
Differences in country programs are illustrated by the two sets of coun-
tries that have absorbed the bulk of World Bank adjustment lending: the
Sub-Saharan African countries and the highly indebted middle-income



200 EVALUATION OF ADJUSTMENT PROGRAMS
countries. Although these two groups share several characteristics, in-
cluding severe macroeconomic imbalances, heavy debt-servicing burdens,
and stagnating or falling per capita consumption, the differences between
them are more important than the commonalities.
In the 1980s the Sub-Saharan countries saw an intensification of long-
standing difficulties in finding a workable development strategy. Central
to their problems were a bias against agriculture, inefficient state enter-
prises, unproductive government investment, a tendency to emphasize
public service employment over expenditures for maintenance and re-
habilitation, and severe shortages in qualified manpower. Combined with
rapid population growth, these constraints reduced or even reversed im-
provements in living standards. With the onslaught of drought, sharply
declining commodity prices, and higher oil prices in the 1980s, painful
adjustments became inevitable.
In the highly indebted countries the crisis was more sudden and un-
expected. For a long period extending through the 1970s, these countries
enjoyed rising living standards, expanding modern industry and services,
and growing public services, financed in part by large overseas borrow-
ing. Then a crisis erupted in the early 1980s with the sharp fall in the
terms of trade, higher real interest rates, and greatly reduced voluntary
lending by commercial banks. Slower growth and increasing protection-
ism in industrial countries compounded their problems. Countries such
as Mexico and Nigeria, which overexpanded their public sector expen-
ditures in response to initial gains from higher petroleum prices, found
themselves in need of structural adjustment.
The structure of loan conditionality, defined by the percentage of con-
ditions falling into each major policy area, is shown in table 10-1 for
Sub-Saharan countries, highly indebted countries, other developing coun-
tries, and all three groups together. Table 10-2 shows the proportion of
total loans in the sample containing conditions related to a specific policy
area. Adjustment programs for all groups have emphasized fiscal and
budgetary policies to reduce domestic absorption, and trade and ex-
change rate policy to provide appropriate incentives for the production
of tradable goods.
In the Sub-Saharan countries the focus of adjustment has been on re-
forming agriculture and public institutions. Because of the dominant role
of agriculture and the promising potential for agricultural exports, em-
phasis has been on raising producer prices, reducing the taxation of farm-
ers through the high profits or low efficiency of state marketing boards,
and improving the quality of public services in such areas as extension
and research. Programs for public enterprises and agencies have empha-
sized improving profitability by raising prices and efficiency, reducing
overstaffing, restructuring activities and finances, and divesting some en-
terprises. The World Bank recognized that progress in Sub-Saharan coun-



The Design and Implementation of Conditionality 201
Table 10-1. The Structure of Conditionality in World Bank
Adjustment Loans, 1980-88
(percentage of all conditions by region)
Sub-Saharan        Highly          Other
African         indebted       developing         All
Policy area              countries'      countriesb       countriesc     countries
Exchange rated                4                2               0              2
Trade                        2S               32              25             28
Fiscal                        8               11              15             11
Budget and public
expenditure                12               9              10              10
Public enterprises           19               17              12             16
Finance                       4               13              13             11
Industry                      7                2               2              3
Energy                        2                3              15              6
Agriculture                  17               10               7             11
Other                         2                1               2              2
Total                       100             100             100             100
Note: Data are based on an analysis of fifty-one SALs and SECALS in fifteen developing countries.
There were a total of S04 legal conditions in those fifty-one loans. Figures may not total 100
because of rounding.
a. Includes Ghana, Kenya, Malawi, and Zambia.
b. Includes Chile, Colombia, Cote d'lvoire, Jamaica, Mexico, Morocco, and the Philippines.
c. Includes Korea, Pakistan, Thailand, and Turkey.
d. Because of the IMF'S responsibility for exchange rates, these figures understate the importance
given by the World Bank to alignment of exchange rates. In the context of adjustment programs,
governments are usually expected to establish and maintain exchange rates that are competitive
internationally, but only in a few cases (such as Chile, Colombia, and Ghana) has there been
explicit conditionality to this effect.
Source: World Bank data (supervision reports, tranche-release reports, and performance audits).
tries would likely be slower than elsewhere because of their shortage of
trained manpower and weak institutional support, and correspondingly
greater effort would be needed to obtain lasting results.
In the highly indebted countries adjustment programs have stressed
trade and financial policies. Trade policies have emphasized improving
the incentive structure for exports while progressively liberalizing imports
through reductions in quantitative restrictions and tariffs. Changes in the
financial sector have emphasized decontrolling interest rates and credit
allocation, developing new capital market instruments, and, where
needed, strengthening the portfolios of lending institutions through fi-
nancial restructuring and improved auditing and supervision. Adjustment
programs for the "other" developing countries in the sample more closely
resemble those of the highly indebted countries than those of Sub-Saharan
countries, but with somewhat less emphasis on trade policy and more
on energy policy.



202 EVALUATION OF ADJUSTMENT PROGRAMS
Table 10-2. The Content of World Bank Adjustment Loans, 1980-88
(percentage of sample loans containing conditions in each policy area)
Sub-Saharan      Highly         Other
African       indebted      developing       All
Policy area            countrieSa     countriesb     countriesc    countries
Exchange rate             31             18              0            16
Trade                     77             91             63            78
Fiscal                    62             73             56            65
Budget and public
expenditure             69             50             38            51
Public enterprises        62             55             44            53
Finance                   39             36             44            39
Industry                  54              9             25            26
Energy                     8             14             50            24
Agriculture               77             41             38            49
Other                     23              9             13            14
Number of loans           13             22             16            51
a. Includes Ghana, Kenya, Malawi, and Zambia.
b. Includes Chile, Colombia, C6te d'lvoire, Jamaica, Mexico, Morocco, and the Philippines.
c. Includes Korea, Pakistan, Thailand, and Turkey.
Source: World Bank data (supervision reports, tranche-release reports, and performance audits).
The Implementation of Conditionality
The following analysis of experience in implementing conditionality is
based on data from supervision reports, tranche-release reports, and,
where available, performance audits done by the World Bank after com-
pletion of a loan. These sources were supplemented by economic reports
and information supplied orally by persons with expert knowledge of
the countries. Data were reviewed on fifty-one SALS and SECALs in fifteen
developing countries to determine progress in implementing each item
of conditionality during the period of the loan.' Implementation success
was graded according to a five-point scale: (1) little or no progress, (2)
moderate progress (some steps taken but not much progress), (3) sig-
nificant progress (substantial steps taken-more than halfway to goal),
(4) implementation fulfilled, and (5) implementation more than fulfilled.
These assessments were summed across policy areas and across countries.
This grading process naturally has limitations. First, it obviously in-
volves judgment: the appropriate grade was not always clear. Second,
although full implementation is a desirable objective beforehand, it may
not always remain desirable or reasonable. Economic circumstances can
change unexpectedly, political resistance can be greater than anticipated,
and errors can be made in estimating how long it takes to implement
particular policy changes. Moreover, conditions differ in the difficulties



The Design and Implementation of Conditionality 203
they pose for implementation, their impact on the economy, and the pain
they may inflict on various segments of the population. Striving for per-
fect implementation of a relatively large number of conditions could lead
the World Bank to be unnecessarily inflexible. It might also have the
undesirable side effect of leading World Bank staff to frame conditionality
so that it could more easily be fulfilled. Plainly, any grading system must
be supplemented by analyses of country performance such as those that
appear in other chapters of this book, especially the case studies in Part
III.
The results of this grading of the implementation of conditionality
(table 10-3) support several generalizations:
* About 60 percent of the conditions in SALs and SECALs have been im-
plemented fully or more than fully.
* This rate increases to more than 80 percent if conditions on which
progress has been substantial are included.
* Progress varies considerably among policy areas, with implementation
of conditions of energy policy (especially energy pricing) being the
strongest and export finance and tax reforms being the weakest.
* Conditionality has been more successfully implemented in SALS than
in SECALS, but this difference disappears if the conditions on which
there has been substantial progress are included.
F Performance in implementing conditionality does not vary sharply
among groups of countries, but highly indebted countries are some-
what more successful than Sub-Saharan countries and "other" de-
veloping countries. The observed differences probably reflect differ-
ences in the composition of conditionality and random factors rather
than any systematic differences in performance among the three groups
of countries.
Variations in Implementation among Policy Areas
What stands out in this analysis of implementation progress is the varia-
tion in outcomes by policy area. Policy changes have clearly been im-
plemented successfully in some areas, such as exchange rate management,
the energy sector (especially pricing), agricultural pricing, the financial
sector, and budgetary and public expenditure policy. Policy change has
equally clearly been slower or less successful in the industrial sector,
taxation, and some aspects of public enterprises. There also are variations
within policy areas. In trade, for example, reductions in tariffs and quan-
titative restrictions have been more fully implemented than improvements
in export financing. The reason is not that the former are more politically
palatable than the latter-the reverse is almost certainly the case. Rather,
the reason is that export finance often involves institutional changes that



204 EVALUATION OF ADJUSTMENT PROGRAMS
Table 10-3. Extent of Implementation of Conditionality by Policy
Area during and after the Loan Period, 1980-88
During the loan period       Current situation'
Conditions     (1) plus    Conditions     (3) plus
fully     substantial      fully     substantial
implemented    progress   implemented    progress
Policy area                    (1)          (2)          (3)           (4)
Exchange rate                  70           90            63            88
Trade                          55           84            63            89
Quantitative restrictions    63           93            69            90
Import duties                62           77            73            82
Import and export
finance                    20           80            43            86
Export incentives            61           82            63            92
Other trade policies         33           76            41            94
Fiscal                         53           78            70            95
Tax policy                   46           54            87           100
Budget and public
expenditure                  68           78            72            85
Public enterprises
(including restructuring)    61           87            70            90
Finance                        71           86            74            90
Industry (excluding
restructuring)               53           93            43            86
Energy                         79           83            83            89
Energy pricing               85           85           100           100
Agriculture                    57           82            58            84
Agricultural pricing         64           86            62            81
All conditions                 60           83            68            89
All SAL conditions             68           84            74            92
All SECAL conditions           61           83            60            85
Sub-Saharan countries          52           85            62            87
Highly indebted countries      67           89            73            91
Other developing
countries                    53           80            56            84
Note: Based on an analysis of fifty-one SALS and SECALS in fifteen countries.
a. Based on most recent data available (generally 1987 but sometimes 1988).
Source: Authors' assessments based on World Bank data.
are more difficult and take more time to implement than do many other
types of change. The case is similar for export incentives: it has been
easier to remove disincentives to exports such as taxes or quotas than to
establish effective duty drawback or bonded manufacturing systems,



The Design and Implementation of Conditionality 205
which are technically difficult to set up and have been implemented very
slowly.
Among reforms of industrial policy there has been substantial success
in reducing price controls (as in Ghana, Malawi, Morocco, and Pakistan)
but not much progress in rationalizing industrial tax incentives (as in
Kenya, Korea, Philippines, and Thailand). For public enterprises, what
works relatively well are covenants on pricing, financial targets, im-
provements in information systems and monitoring, and reductions in
overstaffing, including even fairly large reductions, as in Ghana and Tur-
key. What works more slowly are improvements in management and
operations, the restructuring of firms or subsectors, and, slowest of all,
liquidation and divestiture. Thus Ghana, Mexico, and Pakistan have
taken preparatory steps but have made little progress in closing or selling
off public enterprises. But targets were more than met in Jamaica, and
progress was substantial in C6te d'Ivoire. Liquidation and privatization
programs often face considerable political opposition, and privatization
is often delayed by difficulties in valuing assets and finding suitable buy-
ers.
The policy areas in which implementation has been most successful
are those involving changes in prices, such as exchange rates, interest
rates, or agricultural and energy prices; those in which political consid-
erations are the least sensitive, as for most restructuring of government
expenditure; and those that do not require institutional changes, such as
changes in staffing, training, procedures, laws, and regulations.
Continued Progress with the Passage of Time
Because policy change is continuous and because full implementation of
many institutional reforms can follow adjustment loans only with a con-
siderable lag, it is important that the reforms supported by adjustment
lending continue after loan disbursement ceases. Currently the imple-
mentation of conditionality shows continuing modest progress (table 10-
3). Today 68 percent of all conditions could be considered fully imple-
mented, compared with 60 percent during the loan disbursement period.
Most striking are tax reforms implemented after an unforeseen lag (for
example, in Turkey, institution of a value added tax, indexing of rates,
and improvements in collection procedures; in the Philippines, indirect
tax reforms). Progress has also continued in areas involving the man-
agement and restructuring of public enterprises. Examples include re-
structuring of a major holding company in Malawi, subsector restruc-
turing in the Philippines, and rehabilitation of a number of public
enterprises in Cote d'Ivoire. Similarly, in the period after disbursement,
further progress was made in establishing or improving export credit and
insurance schemes in C6te d'Ivoire and Mexico.



206 EVALUATION OF ADJUSTMENT PROGRAMS
Although there has been some retrogression in implementation since
the loan disbursement period, in no policy areas has backsliding been
substantial. Instead, most backsliding has been piecemeal. Tariff reform
has stalled in Kenya because of foreign exchange shortages and in Thai-
land because of a potential loss of government revenue. Measures to free
up market entry and to alter incentives favoring capital intensity lagged
in the Philippines, as did measures to rationalize incentives for new in-
dustry in Kenya. Public sector price increases were abandoned in Thai-
land because of political opposition. The only example of large-scale
policy reversal is in Zambia, which overturned important reform efforts
in foreign exchange auctioning, import licensing, and consumer subsidies.
The Importance of Key Conditions
Among the various conditions attached to any adjustment lending op-
eration usually four or five might be called key conditions. Although
their legal status is no different from that of other conditions, they deserve
to be singled out because government officials and World Bank staff
designing the loan operation put particular emphasis on them or because
they were expected to make a significant contribution to stabilization or
adjustment in a short time.
Such conditions are expected within one or two years to have an impact
on, for example, savings, the level and composition of investment, budget
deficits, or the production of tradable goods. These key conditions span
the range of stabilization and adjustment policies but fall most heavily
in trade policy (35 percent), public expenditure and fiscal policy (19
percent), public enterprise reforms (14 percent), and pricing, with agri-
cultural and energy pricing together accounting for 14 percent. Specific
examples of key conditions include measures to achieve public sector
savings targets of 4.5 to 4.7 percent of GDP in Chile, agreement on the
1988-90 public investment program (with protection for twenty-one key
projects) and reductions in price and distribution controls in Ghana, and
phased reductions in quantitative restrictions and tariffs in Turkey.
The key conditions in the fifty-one adjustment loan operations were
isolated and analyzed for performance. On balance, performance on key
conditions was better than performance on all conditions: 68 percent of
key conditions were fully implemented within the loan period (table 10-
4) compared with 60 percent of all conditions (table 10-3). And in public
enterprise reform, tariff changes, and industrial policy, the performance
is strikingly better for key conditions. By contrast, progress in tax reform
and export finance is little affected by key-condition status.
Factors Affecting Implementation
This investigation of the implementation of conditionality in fifteen coun-
tries has revealed the importance of macroeconomic policies, tranche-



The Design and Implementation of Conditionality 207
Table 10-4. Extent of Implementation of Key Conditions in
Adjustment Loans during the Loan Period, 1980-88
(percent)
Successful            Distribution of
Policy area                           implementation           key conditions
Exchange rate                              100                        4
Trade                                       59                       35
Quantitative restrictions                 59                       16
Import duties                             77                        8
Import and export finance                  0                        2
Export incentives                         62                        8
Other trade policies                       0                        1
Fiscal                                      64                        7
Tax policy                                40                        3
Budget and public expenditure               71                       13
Public enterprises                          78                       14
Finance                                     79                        8
Industry                                    80                        3
Energy                                      90                        6
Energy pricing                            88                        5
Agriculture                                 56                       11
Agricultural pricing                      60                        9
Other policies                               0                        0
All key conditions                          68                      100
Source: Authors' assessments based on World Bank data.
release conditions, government commitment, the quality of World Bank
advice, and the economic environment. Each of these factors is discussed
below.
The Macroeconomic Policy Framework
The evidence shows that appropriate macroeconomic policies are essen-
tial to the success of sectoral reforms. Otherwise, appreciating real ex-
change rates, excessively high real interest rates, and general inflation
can threaten the viability of reforms. This is particularly the case for
reforms whose success depends on such factors as better incentives for
the production of tradable goods, the right choice of production tech-
niques, and the correct assessment of relative profitability. Therefore the
World Bank has generally agreed to support sectoral reform programs
only when a stabilization program supported by the International Mon-



208 EVALUATION OF ADJUSTMENT PROGRAMS
etary Fund (IMF) is in place to handle the macroeconomic issues or when
the World Bank is assured that the government is pursuing a program
that addresses short-term stabilization issues as well as medium-term
development objectives. In Ghana, for example, a strong IMF-supported
stabilization program provided the appropriate environment for five suc-
cessful sectoral adjustment operations. Similarly, in Morocco an IMF
standby arrangement and an extended Fund facility provided an umbrella
under which five sectoral loans could proceed. At first, budgetary con-
straints slowed the implementation of trade liberalization, but later ef-
forts to mobilize public resources made possible considerable progress
in trade reform.
Despite the World Bank's general intention that sectoral reforms be
supported by appropriate stabilization programs, there have been some
worrisome exceptions. Turkey achieved remarkable adjustment during
five years of SALs and IMF standby arrangements. Subsequently, however,
Turkey received three sectoral loans despite growing budget deficits and
high real interest rates, which dampened private investment in tradable
goods and threatened the viability of firms and financial institutions. A
number of financial sector and agricultural sector loans have been given
to countries without the macroeconomic framework to ensure the real
exchange rates, interest rate and credit conditions, and price stability
needed for the program's success (for example, Argentina, Ecuador,
Kenya, Mexico, and Turkey). The case of Mexico illustrates the problems
of trying to achieve stabilization and structural adjustment objectives
simultaneously. Although the heterodox anti-inflationary policy adopted
in December 1987 reflected the government's pragmatism, the use of
price controls created a dilemma because distorted relative prices will
still have to be corrected when controls are lifted.
Tranche-Release Conditions
The trend throughout the World Bank toward greater use of tranche-
release conditions shows up in the fifteen-country sample as well. All
SALS have included tranche-release conditions except for those in Thai-
land. The only SECALS not tranched in recent years have been the export
development loans for Mexico (two) and Pakistan, which have been small
or repeat operations. SALs have averaged eleven or twelve conditions per
loan, 91 percent of which were tranche-release conditions (the remainder
were other dated or undated covenants). SECALS have averaged nine to
ten conditions, only 75 percent of which were tranche-release conditions.
Thus, on average, the number of conditions in SALS and SECALS does not
appear to have been excessive, although in some cases a considerably
larger number of conditions has been specified (for example, Chile, C6te
d'lvoire, Jamaica, and Philippines).



The Design and Implementation of Conditionality 209
Although there appears to be no general movement toward greater use
of first-tranche conditions in the fifteen countries, such front-loading has
shown up in loans for Colombia (trade and export diversification, 1985),
Ghana (imports for reconstruction, 1985, and the industrial sector,
1986), Pakistan (the energy sector, 1985), and Zambia (agricultural re-
habilitation, 1985, industrial reorientation, 1986, and recovery credit,
1986; all fiscal years).
The importance that the World Bank places on the implementation of
tranche conditions is attested to by the fact that tranche-release delays
have occurred in nearly three-quarters of all adjustment loans as a result
of insufficient progress in fulfilling conditions, although almost all
tranches have eventually been released. These delays reflect the difficulties
in implementing complex and politically sensitive reforms and the World
Bank's expectation that borrowers will achieve a reasonable record of
compliance. The importance of tranche conditions is also reflected in
implementation progress on the loans for the fifteen countries: full im-
plementation for 71 percent of first-tranche conditions, 62 percent of
second- (and third-) tranche conditions, and 43 percent of other legal
conditions.2 Despite the seriousness with which the World Bank views
tranche conditions, several tranche releases have been granted after rather
questionable performance. Examples are the SALS for Bolivia, Costa Rica,
Guyana, and Yugoslavia. The second SAL for Kenya is an example of
tranche-release despite little or no progress toward freeing up grain mar-
keting or liberalizing the trade regime-both important conditions.
Nonetheless, the message is clear: implementation is better for a small
number of key conditions in each loan, and it is better for tranche con-
ditions than for other conditions. Any improvement in implementation
is likely to depend on greater selectivity in setting conditions and on
greater care in monitoring progress. This would involve increased reliance
on tranche-release conditions, limiting conditions to a reasonable number
of key concerns, setting realistic schedules for implementation, and then
taking a stricter attitude about fulfillment. There would still need to be
some flexibility to waive or alter conditions in light of unforeseen changes
in economic or political circumstances, however.
Greater consideration of actions already taken by governments appears
desirable, whether actions are taken before disbursement or as a con-
dition of presenting the loan to the World Bank's Board of Executive
Directors for approval. For governments, such conditionality, especially
relating to prior actions, reduces their vulnerability to charges of yielding
to outside pressure. For the World Bank, such actions dispel uncertainty
by demonstrating a government's commitment and capacity to implement
the loan. Increased reliance on front-loading would be particularly de-
sirable in countries where there are vocal opposition parties or no track
record of success. Limiting conditionality to a small number of agreed-



210 EVALUATION OF ADJUSTMENT PROGRAMS
on concerns also seems essential if the World Bank is to take a stricter
position on fulfillment. The gains from requiring strict adherence to con-
ditionalities should outweigh by a large margin any losses resulting from
shortening the list of conditions. When the list of conditions is long, it
has proved difficult to delay tranche disbursements when most conditions
have been fulfilled, even though some of the missing conditions have
supported objectives essential to the program.
Assessment of Government Commitment
Programs have been better implemented where the government has
played the leading role in analyzing the issues, designing the program,
and formulating the actions. Nonetheless, there is often a need for sub-
stantial World Bank assistance in the design of adjustment programs,
and the Bank's experience with programs in other countries can almost
always be of help. Serious problems arise, however, if an adjustment
program is viewed as the World Bank's rather than the country's pro-
gram. Progress in implementation has been stronger where governments
have "owned" the program and hence were committed to carrying it
through. The ownership of and commitment to any economic reform
program appear to depend on several factors:
* An understanding and acceptance of the program by political leaders,
senior officials, and technical experts
* A clear assignment of responsibilities to various implementing agencies
* An institutional capacity to carry out the program
* An institutional capacity for economic and financial analysis and the
formulation of policy advice
* A longer-term program (say, three to five years) of economic objectives
and policy goals that provides the context for the adjustment measures
supported by a particular SAL or SECAL.
The first three factors determine the will and the capacity to implement
the program; the last two affect the sustainability of the adjustment effort.
Since adjustment is a continuing process, sustaining policy changes de-
pends on the ability to make midcourse corrections and to place those
changes in an appropriate context of continuing medium-term adjust-
ment. Because the extent to which these factors apply in a particular
country varies significantly, the relative roles of governments and the
World Bank, and of technical assistance and country economic and sector
work, have also varied.
The degree of involvement of the government and the World Bank
(and usually the iMF) in the design of adjustment programs has varied
by region as well. The government's role has tended to be greatest in
Asia and Latin America, somewhat less in the EMENA region, and weakest



The Design and Implementation of Conditionality 211
in Sub-Saharan Africa. Government involvement also tends to be weakest
at the beginning of adjustment, when time pressures are the heaviest.
Typically the country is in urgent need of resources, and specific knowl-
edge of what reforms are both desirable and feasible is often poor. In
these circumstances there has been a tendency to accept commitments to
actions that are undesirably vague and inadequately thought through. If
the government is really to own the program, more time will have to be
allowed for the policy dialogue, for the careful design of measures to fit
local conditions, and for the formation of a consensus among the nation's
political leaders and officials.
As the adjustment proceeds, and especially if enough progress is gen-
erated to justify a series of loans, this problem is likely to be reduced by
a continuing dialogue, supplemented by increased policy work within the
country (aided by technical assistance, as needed) and by a better focused
World Bank program for economic and sectoral work. Among the most
successful adjusting countries (Chile, Korea, Thailand, and Turkey), sub-
stantial progress was possible from the inception because governments
had their own ongoing programs and medium-term frameworks. In Thai-
land this was aided by a well-designed and relevant economic and sectoral
work program that contributed quickly to the design of action programs.
In other countries studies in the context of the early adjustment loans
and a redirection of economic and sectoral work were needed to assist
in the design of many policies (as, for example, in C6te d'Ivoire, Ghana,
and Pakistan).
TIhat this dialogue in the context of adjustment lending works is evi-
denced by the growing consensus among Sub-Saharan countries and the
World Bank about the essential problems and specific remedies. Issues
that were contentious at the start of the 1980s, such as exchange rates,
agricultural price reforms, tax policies, and the respective roles of the
public and private sectors, have become decidedly less so. But this process
takes time and a willingness of World Bank staff and government officials
to learn from experience and adopt a pragmatic, open-minded approach.
What is still needed for Sub-Saharan Africa is greater sensitivity in the
World Bank to the possibility of slow supply responses to price changes,
to the importance of nonprice factors in resource allocation, and to con-
flicts between the objectives of changing exchange rates and reducing
budget deficits (as, for example, in Zambia). It is also evident that the
emphasis on building institutions in Sub-Saharan Africa is justified, given
the need to implement and sustain policy reform over the longer term.
The Quality of World Bank Advice
The World Bank has now had considerable experience with adjustment
lending. Since 1980 it has learned a great deal about the formulation of



212 EVALUATION OF ADJUSTMENT PROGRAMS
advice, appropriate coverage, the need for accurate assessment of gov-
ernment commitment, realistic timetables for implementation, and the
desirability of flexibility in designing conditionality and interpreting its
implementation. This experience has allowed the World Bank to improve
its advice.
Some early SALS caught the World Bank quite unprepared to give spe-
cific advice. For example, the first Pakistan SAL called for tariff reform
but without specifying what was to be done or what the effects might
be on government revenues, production, and imports. The government
was unwilling to adopt a tariff reform program without a clearer un-
derstanding of its likely effects in these areas. Later, as the result of a
program of economic and sectoral work, the World Bank was able to
give more concrete suggestions about export incentives and quantitative
restrictions, which the government accepted. In Malawi progress in re-
ducing the taxation of agricultural export crops was spotty early on,
picking up momentum only after a method was introduced for analyzing
the effects of price changes on the production of food and export crops.
The number of conditions in adjustment loans has varied considerably,
and many loans have been heavily loaded with actions to be carried out
by governments. This has often led to a lack of focus and excessive com-
plexity and has caused difficulties in implementation, even in successful
adjusting countries such as Korea and Thailand. Also evident has been
a tendency to set unrealistic timetables, especially for operations involv-
ing institutional change or political processes, such as tax reforms and
the restructuring, liquidation, or divestiture of public enterprises.
The Economic Environment
Although there can be short-term tradeoffs between stabilization and
growth, the two reinforce each other in the long run. Desirable adjust-
ment measures are likely to be sustained and effective only if reasonable
economic growth is maintained. A favorable economic environment
makes the implementation of adjustment programs easier for several rea-
sons. In an expanding economy it is easier to generate resources for
financing investment, government expenditures, and imports. Private in-
vestment is likely to be more responsive to changes in market signals.
Resources are likely to flow more smoothly and quickly between sectors.
More generally, the desired reallocation of resources must be accom-
panied by new investments, which are unlikely in an environment of
economic recession and pessimistic expectations. Painful policy changes
are likely to be more acceptable if people think the pain will not last
long. And governments can devote more attention to medium-term struc-
tural adjustment efforts if they do not have to dissipate their energies



The Design and Implementation of Conditionality 213
fighting such short-run problems as severe foreign exchange shortages,
budget imbalances, or runaway money supplies.
Several countries have survived the crises of the 1980s with very re-
spectable growth rates-examples are Korea, Pakistan, Thailand, and
Turkey. Each of these countries quickly restored favorable export growth
rates through changes in exchange rates and other export incentives.
Pakistan and Turkey were also helped, at least in the early years, by large
inflows of foreign capital and remittances from migrant workers. All four
countries implemented extensive reform programs. Even so, there were
notable lapses, including continuing large budget deficits in Turkey and
Pakistan, and a failure to complete import liberalization programs in
Pakistan and Thailand.
For countries that began from a position of stagnation and massive
distortions, the implementation of adjustment programs has been more
difficult. Zambia's adjustment program followed a prolonged period of
stagnation and a record of poor economic management. Inflows of IMF
and World Bank resources could not offset the hardships inflicted by the
end of price controls and subsidies and the introduction of a foreign
exchange auction. The failure of the program to generate any visible
benefits quickly allowed political opposition to overwhelm it. Reforms
in Ghana also followed a long period of falling incomes, mismanaged
budgets and public enterprises, and badly distorted price incentives. The
Economic Recovery Program, announced in April 1983, emphasized sub-
stantial cuts in budgetary deficits through both expenditure reduction
and revenue mobilization, a massive realignment of the exchange rate
with a view toward eventually establishing a foreign exchange auction,
and a phased removal of price controls. Complementing these reforms,
which emphasized reduced absorption and increased domestic produc-
tion, was a large inflow of resources from the IMF and World Bank to
provide the necessary inputs to increase production. Positive results, pro-
duced fairly quickly, inspired confidence in the government's manage-
ment and were important to the program's success.
Summary and Conclusions
First, about 60 percent of the conditions in SALs and SECALs have been
fully implemented. Extending the review beyond the loan disbursement
period yields a somewhat higher rate of success in implementation. This
suggests that reversals of policy have been relatively minor and that
progress continues to be made after loan completion in implementing
changes that take more time. Second, implementation varies significantly
among policy areas: it is best for price changes and the reallocation of
public expenditure and worst for institutional reforms. Third, imple-
mentation performance is better for a small number of key conditions



214 EVALUATION OF ADJUSTMENT PROGRAMS
(about four or five) in adjustment loans than for conditions overall. (Key
conditions are expected to have a large impact on stabilization or ad-
justment within a short period.) Fourth, the major factors affecting im-
plementation appear to be the overall macroeconomic policy framework,
tranche-release conditions, government commitment, the quality of
World Bank advice, and the economic environment.
Given the high economic and social costs of rapid adjustment to ex-
ternal shocks, quick-disbursing adjustment loans are likely to remain an
important part of World Bank lending for the foreseeable future. Learn-
ing from experience and adapting and changing structural adjustment
procedures to increase their effectiveness thus continue to be essential.
The analysis in this chapter gives important indications of what works
and what does not, of what needs changing and what should not even
be attempted in the context of adjustment lending. More broadly, it
shows a greater need for selectivity in adjustment lending to ensure that
loans will be effective and that they do not serve merely to postpone
adjustment by relieving resource constraints.
Greater selectivity has a number of dimensions:
* More concern with developing a consensus within governments about
the appropriate reform strategy and with ensuring government com-
mitment to that strategy
* More time spent ensuring that external and internal financing for the
adjustment program are adequate and that programs will not have to
be abandoned because of lack of resources
* More careful selection of a small set of important conditions that are
likely to have a major impact on adjustment within a short time.
In all, greater selectivity would assist the World Bank by increasing
the effectiveness of its programs and reducing the risk to its portfolio. It
will also assist developing countries by ensuring realistic levels of external
and internal financing in support of their adjustment efforts.
Notes
1. See notes to table 10-1 for country groupings and the countries included
in each group. These fifty-one loans constitute 42 percent of the number and 60
percent of the value of SALs and SECALs approved during fiscal 1980-87. The
conditions examined were those legally required as part of agreements with the
World Bank.
2. If the conditions on which progress was substantial are added to the total,
the figures become 88 percent of first-tranche conditions, 85 percent of second-
tranche conditions, and 73 percent of other legal conditions.
References
World Bank. 1986. "Structural Adjustment Lending: A First Review of Expe-
rience." Report 6409. Operations Evaluation Department, World Bank, Wash-
ington, D.C.



The Design and Implementation of Conditionality 215
. 1988a. Adjustment Lending: An Evaluation of Ten Years of Experience.
Policy and Research Series 1. Washington, D.C.
. 1988b. "Interim Report on Adjustment Lending." Report R88-15.
Washington, D.C.
Comments
Elliot Berg
THE EFFECTIVENESS of policy-based lending can be evaluated by three kinds
of criteria: process (the quality of the dialogue), implementation (meeting
conditions and other proposed reforms), and results in meeting either
intermediate targets (such as for investment rates, exports, and deregu-
lation of markets) or ultimate goals (such as for output and income dis-
tribution). Judgments about the effectiveness of policy reforms in any of
these dimensions are always difficult because of problems of definition
and measurement. In these comments I define successful implementation
as the induced adoption of policy reforms that are both durable and
additional (in other words, beyond what would have been adopted any-
way).
William McCleary focuses on the second type of criteria-success in
implementation. His chapter provides a detailed and yet admirably con-
cise overview of the content of conditionality in World Bank adjustment
loans and assesses the degree to which loan conditions have been im-
plemented. He notes that implementation has been slow in reforms of
the industrial sector, taxes, and public enterprises and mentions that
tranche disbursements were delayed in three out of four loans because
of inadequate progress in meeting conditions. The overall assessment,
however, is strongly positive. Implementation has been generally satis-
factory and improves over time. Backsliding is apparently not much of
a problem. Implementation is even better for key conditions than it is
overall.
That the explicit conditionality in the World Bank's adjustment lending
has been so effectively implemented is a rather surprising finding. Com-
pliance with International Monetary Fund (IMF) performance criteria has
been much less impressive; according to some counts, the compliance
rate in the early 1980s- was something like 20 percent (see, for example,



216 EVALUATION OF ADJUSTMENT PROGRAMS
Spraos 1986). Moreover, successful implementation is not what earlier
episodes in conditional lending lead us to expect.
The main effort in conditioned policy lending occurred in Latin Amer-
ica in the 1960s under the Alliance for Progress. Most of the condition-
ality involved stabilization measures familiar in IMF standby agreements.
Few traces of the exercise in conditionality were visible by the end of the
decade. One recent assessment offers the following observation: "the
Alliance assumption that the offer of external assistance can provide an
effective inducement for a country to adopt basic policy changes and
reforms still appears to retain considerable vitality in U.S. assistance
strategies and negotiations. Since the Alliance experience can be inter-
preted as an overwhelming repudiation of the general efficacy of this
assumption, its continued vitality is striking" (Heller and Wionczek
1988).
Some attempts at policy-based lending also took place in South Asia
in the 1960s; the United States and the World Bank tried to condition
aid to India on reforms in agriculture and liberalization of the trade
regime, including a devaluation of the rupee. One participant-observer
argues that at least the timing of the devaluation was affected by the
conditioned policy loan and that the tilt of long-term trade policy was
thereafter in the direction of liberalization (Lewis 1989). This is a de-
batable conclusion, and in any case the experience was marked by mis-
understandings on all sides and caused much bitterness among Indians;
policy-based lending was never tried in India again.
In Pakistan the U.S government, between 1963 and 1966, gave con-
ditioned loans tied to import liberalization. For various reasons the im-
plementation was on-again, off-again until 1967, when it was finally
abandoned. The trade regime was more restrictive in 1967 than it had
been in 1963. These and other experiences with explicit conditionality
(including the World Bank's own experience with covenants and con-
ditions in project lending) indicate a generally poor track record.
In addition to the sparsity of historical precedent for effective condi-
tionality, there is a more important reason why the good results reported
by McCleary are surprising. The general environment within which con-
ditions are set and monitored and the kind of behavior that they inspire
are such that highly imperfect implementation is to be expected.
Before taking up this point in detail, one methodological point is worth
making. Of the more than fifty countries that had received adjustment
loans by 1989, fifteen are analyzed by McCleary. Of these, five are from
Sub-Saharan Africa, four each from Asia and Latin America, and two
from the Middle East and North Africa. But the proportion of total
adjustment loans (in number) going to Sub-Saharan Africa-about
half-is much greater than the proportion in the study sample. It is not
certain that this makes much difference in the experience with condi-



The Design and Implementation of Conditionality 217
tionality, but it is possible, even likely, that it does. Adjustment programs
in Sub-Saharan Africa tend to be less home grown, more heavily con-
ditioned, and more heavily laden with institutional conditionality than
is usual in other regions. African economies are more open and more
commodity dependent. All of this should make programs there at once
more difficult and more fragile than in other regions. The bias in the
sample may therefore yield a more upbeat assessment of the effectiveness
of conditionality than reality warrants.
Intuition, fed by experience, suggests that explicit conditionality is a
highly implausible instrument of policy reform and adjustment from
which little should be expected. This is due to the nature of the envi-
ronment in which the conditioned lending takes place. The relevant ele-
ments in this environment are well known.
1. Many of the intellectual or analytical underpinnings of 1980s-style
adjustment lending are contested-such fundamentals as the feasibility
of export-led growth, the efficacy and beneficence of deregulated mar-
kets, and the desirability of market-determined interest rates or exchange
rates. It is true, as McCleary notes, that there is more consensus on these
ideas now than there was ten years ago. But as indicated by the Economic
Commission for Africa's recent attack on liberalization ideas, profound
disagreements remain. Similar disagreements also exist on less cosmic
matters, such as the size and timing of exchange rate adjustments, the
desirability of border pricing for basic foods, and the degree and the
speed of reducing protection for the industrial sector. Since consensus is
lacking in many countries on important loan conditions, implementation
will tend to lack conviction and programs will be easily diluted or de-
railed.
2. Political and bureaucratic consensus is even more uncommon than
intellectual or analytical agreement. Agreements negotiated and signed
by ministers of finance or planning are implemented by sectoral minis-
tries. Sometimes these ministries are only perfunctorily consulted; often
they are in deep disagreement with the spirit and the particulars of the
reform program. This seems to be especially likely for ministries of ag-
riculture.
3. Conditions related to institutional reforms are almost sure to lag
and are in any case prone to cosmetic compliance. This may have some-
thing to do with lack of commitment or political will, but more often it
is a simple result of administrative failures that are due to limited ca-
pacity, that is, to factors inherent in underdevelopment.
4. Explicit conditionality coexists uneasily and may indeed be incom-
patible with the notion of local "ownership" of adjustment programs-
that is, with a government's responsibility for and commitment to ad-
justment.



218 EVALUATION OF ADJUSTMENT PROGRAMS
5. Monitoring is difficult. There are multiple conditions, many more
than the ten or so that are sometimes regarded as "legal" (those found
in loan agreements). Many-especially the "process" conditions-are so
imprecise as to be essentially unmonitorable. Moreover, in developing
countries the administrative systems are normally diffuse and opaque.
Combined with infrequent and sometimes hurried supervision, this means
that it is often difficult for lenders to know whether or in what sense
implementation has in fact occurred. Deliberate or accidental slippage is
possible at every stage. The World Bank's capacity to supervise all this
is extremely limited. Administrative opacity may even prevent confir-
mation that conditions for effectiveness have been satisfied, as happened
in one West African country recently when lack of implementation was
not discovered until months after disbursement had begun.
6. The threat of loan suspension, rarely very credible to begin with, is
likely to diminish rather quickly. Both parties are anxious to continue
disbursement, performance is rarely all bad, and assessment of good and
bad implementation is usually subjective. Nowadays, adjustment loans
are usually part of an exercise to close the resource gap, with the IMF
frequently in the wings making ominous noises about the need to continue
spending if the program is not to unravel. In some countries, programs
seem to continue in the face of repeated foot-dragging and surface com-
pliance with conditions. The low and diminishing credibility of the threat
of program cancellation should encourage the reduction of reform efforts.
Given the circumstances outlined above, one should expect imple-
mentation of explicit conditionality to be often perfunctory and super-
ficial. This does not mean that governments will try to avoid imple-
menting the whole slate of conditions. In fact, many of the conditions
are superfluous in the sense that the government in question would have
taken the actions even without conditionality. It does mean, however,
that probably in a substantial number of cases conditionality is acceptably
implemented only at the expense of watering down the original require-
ments or because the World Bank accepts pro forma compliance. Thus,
for example, a condition for tranche release may be the adoption of an
action plan for reforming public enterprises. If implementation is nil or
partial, World Bank staff will settle for the completion of terms of ref-
erence for a study to recommend an action plan. Or tranche release may
require economic evaluation of a certain number of projects in the public
investment program. But delays and risks of nonperformance induce
World Bank staff and management to accept perfunctory reviews of
projects as meeting the condition.
The fact that such dilution of content and pro forma compliance may
be common raises the question of how to interpret the figures on im-
plementation assembled by McCleary. The observation that 60 percent
of conditions are fully implemented and more than 80 percent "sub-



The Design and Implementation of Conditionality 219
stantially" implemented takes on a different meaning if the yardstick for
measuring compliance is rubbery and if many conditions are superfluous.
Given not only the ability but the strong incentive to bend the yardstick,
one should hesitate to conclude that a 60 percent implementation rate
indicates success. When the superfluity of many conditions is also taken
into account, the 60 percent compliance rate looks even less impressive.
One should also hesitate to interpret the delays in tranche release as
evidence of "the importance that the World Bank places on the imple-
mentation of tranche conditions." It is at least as plausible to interpret
it as evidence that weak compliance is widespread and as an indicator
that the World Bank and its partners needed time to work out a mutually
acceptable combination of improved performance, softening of content,
and surface compliance.
Thus far the effectiveness of conditionality has been considered only
in terms of the implementation record and its interpretation. There is
another side of the question that is much more important: the propensity
for explicit conditionality to undermine the quality of policy dialogue
and the durability of policy reforms.
First of all, conditionality gives the impression of being imposed from
outside, even when it is not. It causes liberalization ideas and policies to
be identified with outsiders. Local critics attack the "World Bank's aus-
terity program" or the "World Bank's liberalization scheme"; foreign
reporters write about it the same way. This may have a slight facilitating
effect since the authorities know they can blame the World Bank or IMF
if things go wrong. But the overall effect is to discourage the growth of
local "ownership," discredit the policy ideas at issue, and delay the
growth of political responsibility.
More important, explicit conditionality tends to distort the policy dia-
logue and can poison it altogether. It tends to replace the joint search
for acceptable policies, which is what the dialogue should be about, with
a process of game playing aimed at signing an agreement and completing
disbursements. To address the real obstacles to reform-for example,
local skepticism about the technical suitability of proposed reforms, or
concern that minority groups may benefit excessively, or aversion to
political risk-the policy dialogue should be conducive to joint problem
solving and friendly persuasion, rather than simply to document signing.
Effective dialogue of this kind requires open and experimental attitudes,
a noncontentious environment, and extensive opportunity for genuine
exchange of ideas and consensus building around specific issues and pro-
posals.
No doubt in some cases the policy dialogue in adjustment lending
approaches this ideal. But the conditionality factor discourages effective
dialogue. Instead, it tends to encourage a game in which differences are
reconciled more on paper than in reality and agreements are framed so



220 EVALUATION OF ADJUSTMENT PROGRAMS
as to meet conflicting needs-the country's need for understated, flexible
conditions that will not involve rigid, risky, or excessively difficult com-
mitments, and the World Bank's desire for conditionality that is as firm
and explicit as possible. This does not necessarily rule out true dialogue
and spirited debate over substantive issues; there has been more such
debate in the past decade than ever before, stimulated by policy lending.
But in terms of solving problems and changing minds the impact is much
diluted by the conditionality-ridden atmosphere.
To the extent that explicit conditionality leads to this kind of game
playing, it is counterproductive. It diverts attention and energy from real
problem solving to contract maintenance. Along with the related matter
of the rubbery yardstick for evaluating the extent of implementation, it
implies a considerably less sanguine view of the effectiveness of explicit
conditionality than that put forward by McCleary-and by the authors
of the earlier report on adjustment lending (World Bank 1988).
Of course, the validity of this view depends on how widespread have
been the tendencies to dilute conditionality and distort the dialogue. My
view is that these tendencies are common and probably dominant. But
this could easily be wrong because it is based on observation that is too
distant or biased, in the sense of being based on too few or too unrep-
resentative a sample of countries. There is nonetheless enough evidence
to suggest that the tendencies in question are not rare and could be wide-
spread. The recommendations presented by McCleary-to reduce the
number of conditions, to front-load more of them, to build more internal
consensus, to take more time, to monitor conditipnality more closely,
and to take a firmer line on compliance-are a recognition that problems
of the sort outlined above do exist. Just how general or pervasive they
are will not be known until there are more in-depth country case studies.
In any case, to neglect these seamier sides of the conditionality process,
as do McCleary's chapter and the World Bank's 1988 report on adjust-
ment lending, may encourage unwarranted complacency.
Even if it were granted that explicit conditionality has proved to be
inappropriate, ineffective, and counterproductive, better alternatives
would have to be found before explicit conditionality could be aban-
doned. To consider the options here would take this discussion too far
afield. But the search for alternatives is not likely to be vigorous so long
as it is generally believed that there is not much basically wrong with
the way explicit conditionality works in adjustment lending. The first
task should be to take a hard look at that belief.
References
Heller, Jack, and M. S. Wionczek. 1988. "The Assumptions of the Alliance." In
L. R. Scheman, ed., The Alliance for Progress: A Retrospective. New York:
Praeger.



The Design and Implementation of Conditionality 221
Lewis, John. 1989. "Policy Based Assistance: An Historical Perspective on AID'S
Experience and Operations in India/Asia in a Bilateral and Multilateral Con-
text." Washington, D.C.: U.S. Agency for International Development.
Spraos, John. 1986. IMF Conditionality: Ineffectual, Inefficient, Mistargeted.
Princeton Essays in International Finance 166. Princeton, N.J.: Princeton Uni-
versity.
World Bank. 1988. Adjustment Lending: An Evaluation of Ten Years of Ex-
perience. Policy and Research Series 1. Washington, D.C.



11 Macro Performance under
Adjustment Lending
Riccardo Faini
Jaime de Melo
Abdel Senhadji-Semlali
Julie Stanton
THE PROMOTION of both stabilization and growth represents a significant
challenge for the design of adjustment policies. Despite the relatively short
time that has elapsed since the advent of adjustment programs, an eval-
uation of them is clearly important-but it is fraught with difficulties.
To begin with, any assessment of World Bank-supported adjustment
programs must recognize that most of them were undertaken jointly with
the International Monetary Fund (IMF). Any assessment must also rec-
ognize that performance will be influenced by the external environment:
countries that faced a more unfavorable environment would be expected
to show less improvement than other countries. Furthermore, it would
be desirable to control for the adjustment policies that might have been
undertaken in the absence of World Bank and IMF adjustment lending.
Finally, one should recognize that participation in World Bank and IMF
programs is not independent of a country's own policies.
Taking into account all these difficulties with the quantitative assess-
ment of adjustment lending is well beyond the scope of the preliminary
study undertaken here, which focuses on a simple statistical analysis of
macroeconomic performance. In a companion paper we attempt to tackle
some of these difficulties by postulating a simple but specific model of
how World Bank and IMF lending and the external environment interact
(Faini and others forthcoming). Here, we rely on a simpler control-group
analysis, which is more restrictive in its assumptions and therefore less
satisfactory in controlling for the methodological problems mentioned
above. We use several approaches to check for the robustness of our
results, however.
In the following section we briefly discuss the pitfalls of alternative
statistical approaches in assessing the effectiveness of adjustment pro-
grams. This discussion indicates the limitations of our statistical evalu-
ation. Next we provide summary statistical evidence on the environment
222



Macro Performance under Adjustment Lending 223
under which adjustment lending took place, emphasizing the relative
roles of commercial and official multilateral funds and the size of external
shocks. The following sections present our statistical analyses. A first-
cut statistical evaluation compares the relative performance of countries
that received a high amount of World Bank and IMF adjustment loans
with countries that received a low amount. Then thirty countries that
received their first World Bank-supported adjustment loan by 1984 are
compared with sixty-three countries that had not. Finally, we look at
tradeoffs between growth and improvement in the balance of resources
for the same group of thirty recipient countries and compare these trade-
offs with those for nonrecipient countries.
Limitations and Implementation of the Methodology
Two issues in any assessment of adjustment lending are the choice of
evaluation method and the choice of indicators. A discussion of the se-
lection of method is necessary because none of the available methods
provides a good prediction of the counterfactual, that is, of what policies
would have taken place and consequently what the performance would
have been in the absence of adjustment lending. A discussion of the choice
of indicators is useful since the intermediate objectives of adjustment
lending are many, and there are interactions between indicators.
Alternative Approaches and Selection of Indicators
Adjustment effectiveness can be measured in several ways.' One way,
the before-after approach, defines the effectiveness of adjustment as the
difference in the values of a set of indicators before and after adjustment
was initiated. Another way (which is often used by the Operations Eval-
uation Department of the World Bank) measures effectiveness in terms
of the difference between the target values specified in the adjustment
programs and the values that are realized.
Each approach has its advantages and disadvantages. The before-after
approach is simple to apply, but it is subject to serious biases because it
assumes that all factors (besides the adjustment program) that affect per-
formance, such as terms of trade variations and movements in interest
rates, do not change between the pre- and postadjustment periods. In the
actual-versus-target approach, one gets no information about how ad-
justment lending affects a country's performance. Furthermore, just as
the situation prevailing before a country enters into an adjustment lending
program is not likely to be a good predictor of what would have happened
in the absence of an adjustment loan, so too loan targets may be achieved
because of favorable exogenous developments, irrespective of the policies



224 EVALUATION OF ADJUSTMENT PROGRAMS
adopted. Thus neither of the two approaches provides an adequate view
of the costs and benefits of the adopted policy course.
The control-group approach, with its several variants, is designed to
overcome the shortcomings described above. The simplest variant as-
sumes that adjustment loan recipients and nonrecipient comparators face
the same exogenous factors and the same initial conditions. The perfor-
mance of the nonrecipient group serves as an estimate of what the per-
formance of the adjustment loan group would have been in the absence
of adjustment programs. Because of its simplicity and ease of application,
this is the approach we use in the statistical analysis of performance
indicators for adjustment loan recipients.
This control-group approach still does not provide a good basis for
the counterfactual, however. It does not take into account the fact that
changes in the external environment may not be the same for both groups
of countries. Nor does it identify what the policy choices would have
been in the absence of adjustment lending. In our companion paper (Faini
and others forthcoming) we recognize these complications and use a mod-
ified control-group approach in which we postulate a simple reaction
function that policy choice in the absence of adjustment lending depends
on initial conditions.
Moreover, it may be argued that changes in performance attributed
to an adjustment lending program may reflect the country-specific con-
ditions that led countries to participate in the program.2 For example,
countries with large external disequilibria had to cut public investment
expenditures and also had a greater probability of entering into an ad-
justment lending program. This possibility is not usually adequately cap-
tured in statistical evaluations of effectiveness. As shown by Goldstein
and Montiel (1986), if macroeconomic performance is positively cor-
related with the country-specific conditions that led countries to partic-
ipate in an adjustment lending program (sample-selectivity bias), then
the control-group estimate of the effects of adjustment lending will ov-
erstate the true effects of the program.
Other issues arise because of the choice of indicators for measuring
performance. We used nine indicators in four areas: growth, external
balance, internal balance, and external debt. Growth indicators are av-
erage real growth rates for gross domestic product (GDP) and exports,
and the investment/GDP ratio. Indicators of external balance are average
values of the real effective exchange rate index (an increase in the value
of the index signifies a real devaluation) and the current account deficit
as a fraction of GDP. Indicators of internal balance are the government
budget surplus or deficit as a fraction of GDP, and inflation. Indicators
of external debt are gross external debt and total debt-service payments
(interest and principal); both are given as fractions of exports of goods
and nonfactor services.



Macro Performance under Adjustment Lending 225
Obviously there are interdependencies among this group of indicators.
To take these interdependencies into account one would have to specify
an appropriate structural model as given in Khan and Knight (1981) or
Bourguignon, Branson, and de Melo (1989). Such an approach would
lend itself to counterfactual simulation analysis but would be more de-
manding of the data, especially for a large sample of heterogeneous coun-
tries. Furthermore, the nine indicators selected here do not assess the full
set of targets under adjustment lending (for example, public sector and
financial sector reforms are not covered). The choice of indicators was
dictated by the desire to have indicator values for as large a group of
countries as possible. Because of this focus on breadth rather than depth,
the results presented here are best viewed as complements to the more
thorough country studies of the effectiveness of adjustment lending (see
Part III of this volume).
Implementation
Subject to these caveats, our objective was to assess the effects of ad-
justment lending programs on macro performance. Of the forty-eight
countries that received World Bank structural adjustment loans or sec-
toral adjustment loans, only thirty received their first adjustment loan
before 1985. Our evaluation covers this subgroup of thirty recipients for
which two to three years of data are generally available after receipt of
the first adjustment loan. This group of thirty countries is compared with
a control group of sixty-three countries that had not received their first
adjustment loan before 1985, that participated in IMF adjustment pro-
grams only, or that did not participate in adjustment programs supported
by either the IMF or the World Bank. We also rely on a two-way clas-
sification of countries by income level (low income and middle income)
to control for the influence of institutional characteristics on program
effectiveness.
Obviously this definition of the control group of nonrecipients of ad-
justment lending is not very satisfactory. Furthermore, most of the thirty
countries that participated in World Bank-supported adjustment pro-
grams also participated in IMF-supported adjustment programs.3 We
therefore decided to evaluate both the effectiveness of World Bank-sup-
ported adjustment programs and the joint effectiveness of World Bank-
IMF-supported adjustment programs. In this regard, it was not easy to
decide on a before-after cutoff point when World Bank and IMF programs
were not initiated in the same year. We selected 1982 as the cutoff point,
that is, as the first year for the assessment of World Bank-IMF programs,
because this was the year in which adjustment lending intensified.
In our examination of the joint effectiveness of World Bank-IMF ad-
justment programs, we assess countries that received a nonnegligible



226 EVALUATION OF ADJUSTMENT PROGRAMS
amount of World Bank and IMF financial support. In a first-round eval-
uation, we look for robustness by defining the control group as low-
credit (LC) recipient countries. This control group is constructed so that
it includes one-third of the ninety-three countries in our sample. Values
of performance indicators for low-credit countries are then compared
with the corresponding values for high-credit (HC) recipients, also defined
so as to include one-third of the countries in the sample. Slicing the sample
into terciles and omitting from the comparisons the middle tier is a crude
way to add some robustness to the comparisons of the average value of
indicators before and after adjustment lending. The first-round evalua-
tion relies on nonparametric statistics (the number of countries in the
high-credit group that improved their performance relative to themselves
and to the low-credit control group) and is therefore similar to previous
assessments (for example, Balassa 1988).
The second-round evaluation relies on tests of significance of changes
in mean values (corrected for unequal variances and differences in sample
size). Two tests are performed. The first is for change in mean perfor-
mance among high-credit recipients after adjustment lending. For each
of the nine indicators, x, we test the null hypothesis   C = f   where
subscripts 1 and 2 refer to the before and after periods. The second test
is for effectiveness, and for indicator x we test the null hypothesis
PI.2 - PH  =  L       .LC _   xL
In assessing the effectiveness of only World Bank-supported lending
for the group of thirty countries defined above, we use the same set of
performance indicators and the same combination of nonparametric and
parametric statistics. The main difference in these tests is that the cutoff
point is now endogenous, that is, it is determined individually for each
country according to the year of its first adjustment loan.
The External Environment
In this section we ask two questions: (1) Was adjustment assistance under
IMF and World Bank auspices directed mostly to countries that faced a
more unfavorable external environment? (2) Was there complementarity
or substitutability between commercial bank lending and IMF and World
Bank lending? The relative importance of private and official funds was
examined to determine whether official lending served as a catalyst for
private funding, which is one of its objectives.
To measure the significance of the deterioration in the external envi-
ronment, we quantify the impact of the external disturbances associated
with declining terms of trade and rising real interest rates. External dis-
turbances are measured over the adjustment loan period of 1982-86,



Macro Performance under Adjustment Lending 227
while 1976-81 is taken as the base period.4 The formula for computing
the welfare costs of external disturbances is as follows:'
(11-1)  aw/w =         - R)(D/Y)1 + (PX2/PX1 - 1)(X/Y)1
- (PM2!PM1 - 1)(MIY)l = -RIR + TOT
where subscripts 2 and 1 refer to the periods 1982-86 and 1976-81,
respectively, a bar over a variable means an average value over the rel-
evant period, and the variables are defined as follows:
R = average real interest rate (the deflator is the U.S. GDP deflator
and the nominal interest rate is the weighted interest on con-
cessional and commercial debt)
Y = average real GDP measured in dollars
PX, PM = export and import price indices measured in dollars and de-
flated by the U.S. GDP deflator
X, M = exports, imports
D = gross outstanding debt, net of reserves
RIR = real interest rate index
TOT = terms of trade index.
In equation 11-1 the first term measures the welfare effect of higher-
than-expected interest payments, and the remaining terms measure the
effect of changes in the terms of trade. The choice of periods implies that
the welfare measure is expressed as a percentage of the average value of
GDP during 1976-81. Because cutoff points for the measurement of ex-
temal shocks are debatable, however, we also report calculations that
use 1978-81 as the base period.
Estimates of the size of the external shock are shown in table 11-1.6
The choice of base period does not significantly affect the estimated im-
pact of higher real interest rates, but use of the base period 1976-81
results in a significantly higher estimate for the impact of declining terms
of trade than does use of the base period 1978-81. In general, the es-
timated magnitude of external shocks is greater for recipients of World
Bank adjustment loans than for the control group.
When countries are classified by intensity of World Bank-IMF lending
(high- and low-credit groups), the terms of trade shock is higher for low-
credit countries. The measure of World Bank-IMF lending intensity is the
average net IMF credit plus World Bank structural and sectoral adjustment
loans during 1982-86, expressed 'as a percentage of average 1982-86
GDP. The cutoff points for high-credit and low-credit countries are chosen
so that the sample of ninety-three countries is divided into three groups.
The cutoff points are World Bank-IMF credit of more than 0.7 percent
of GDP for the high-credit group and less than 0.02 percent for the low-
credit group. Countries that fall in the middle are excluded from the
comparisons.



228 EVALUATION OF ADJUSTMENT PROGRAMS
Table 11-1. External Shocks as Percentages of Average GDP for
Sample Groups of Countries, 1978-81 and 1976-81
+ RIR'                    -TOT b
Country group (number)        1978-81       1976-81      1978-81       1976-81
All recipients (30)             0.011        0.010         0.036        0.052
All intensive recipients (12)c  0.011        0.010         0.045        0.065
Nonrecipients (63)              0.006        0.006         0.033        0.049
High-credit (31)                0.012        0.010         0.022        0.037
Low-credit (31)                 0.003        0.003         0.042        0.063
Low-incomed
Recipients (12)               0.015        0.012         0.067        0.091
Nonrecipients (24)            0.010        0.008         0.016        0.037
Middle-income
Recipients (18)               0.009        0.009         0.017        0.028
Nonrecipients (39)            0.004        0.004         0.041        0.054
Note: A positive value indicates a worsening of the external environment in 1982-86 compared
with the base period. Number in parentheses indicates the number of countries in each category.
See table 11-6 for country classifications.
a. RIR: real interest rate, as measured by (R2 - RB(D/Y)I, the first term of equation 11-1.
b. TOT: terms of trade index as measured by (FM2/PM, - 1)(V/7Y), - (PX2/PX1 - 1)
(R/Y)1, the third and second terms respectively of equation 11-1 (reversed to show a sign change).
Note that the column heading has a negative sign. The numbers are reported as positive ratios
to GDP whereas equation 11-1 defines a deteriorating external environment as a negative number.
c. Intensive recipients are countries that received three or more adjustment loans (see table 11-
6).
d. For definitions of low- and middle-income, see Wortd Bank (1989), chap. 2.
Source: World Bank data.
The correlation between the estimated size of the external shocks and
the availability of World Bank, IMF, and private credit is shown in table
11-2. The correlation between the size of the external shock and the
intensity of World Bank and IMF credit is in all three cases insignificant,
which suggests that World Bank and IMF lending was not targeted to
countries facing the most unfavorable external environment. As expected,
there is a strong positive correlation between World Bank and IMF credit,
reflecting the fact that much of World Bank adjustment lending took
place simultaneously with IMF stabilization programs. The significantly
negative correlation between World Bank and IMF adjustment lending
and net private credit supports the widely held view that World Bank
and IMF credit served as a substitute for rather than a supplement to
private credit. This result suggests that the joint involvement of both
institutions was not sufficient to serve as a catalyst for inducing com-
mercial bank lending.



Macro Performance under Adjustment Lending 229
Table 11-2. Correlation Matrix of External Shocks and External
Financing, 1982-86: Pearson Correlation Coefficients
(credit flows as a ratio of average GDP for the period)
Net
World Bank    private
IMF'     World Bankb      and IMF      credit
External shocks'
Number of observations          91          91              91          91
Correlation coefficient      - 0.06        0.03         -0.01       -0.05
Significance level             0.56        0.75           0.90        0.62
IMF financing
Number of observations                       93             93          93
Correlation coefficient                    0.29           0.78      -0.19
Significance level                        0.005         0.0001        0.06
World Bank financing
Number of observations                                      93          93
Correlation coefficient                                   0.83      -0.23
Significance level                                      0.0001        0.03
World Bank and IMF financing
Number of observations                                                  93
Correlation coefficient                                             - 0.26
Significance level                                                    0.01
Note: Mean values of variables:
Net private credit  -0.002
External shock        0.05
Net IMF credit        0.003
World Bank credit     0.003
World Bank-IMF credit  0.006
a. Net IMF credit.
b. Structural and sectoral adjustment loans.
c. Calculated and expressed as in table 11-1 (+RIR -TOT).
Source: World Bank data.
Indicators of Performance
The relative performance of high-credit and low-credit countries in each
income group (low and middle income) is summarized in table 11-3. The
measure of relative performance is the change in performance in high-
credit countries in 1982-86 compared with 1978-81 relative to the
change in performance in low-credit countries between the same two
periods. For example, the entry " 10 + " in the third row under the column
heading "Low income" indicates that ten of sixteen high-credit countries
in the low-income classification improved their GDP growth when com-
pared with the change for the low-credit control group. The plus sign



Table 11-3. Performance Indicators for Recipients of World Bank and IMF Adjustment Loans Relative to Nonrecipients,
by Income Group
Low-income                                        Middle-income
Relative to                                        Relative to
own             Relative to                        own             Relative to
Indicator                         Number        performance'       nonrecipientsb     Number        performance       nonrecipientsb      Row sum
High-credit countries (HC)           16                                                  15                                                  31
Low-credit countries (LC)            10                                                  21                                                  31
GDP growth                           10+             -                  -                10+             *                                   20
Investment/GDP                        5-             -                  -                 2-             *                   t                 7
Export growth                        14+             -                   *                9 +            -                  -                23
Real exchange rate                   12 +            -                  -                13 +                                                25
Current account surplus/GD'          12+             -                   *               11 +            *                  *                23
t    Budget surplus/GDP                     7-             -                  -                 8 +           -                   -                15
(    Inflation                             10-             -                  -                10-            -                   -                20
External debt/exports                 9-             -                  -                 6 -                                                 15
Debt service/exports                 11+             -                   *               10+             -                  -                21
Share showing improvement          0.63                                                0.59
* Denotes 10 percent (or more) significance level of test of equality of means.
- Denotes statistical insignificance.
Note: The numbers in the table show for each indicator how many high-credit recipient countries in each classification improved during 1982-86 compared with 1978-
81. The plus and minus signs show the direction of change of the average value of an indicator in comparison with the average value of the same indicator for the low-
credit comparators. For five indicators-GDP growth, investment/GDP, export growth, current account surplus/GDP, and overall budget surplus/CDP-a positive difference
is an improvement (shown by a plus sign). For example, if the average GDP growth of high-credit recipients in a subgroup was 0.2 percentage points less than that of low-
credit countries during 1978-81 and 0.1 percentage points less during 1982-86, the difference (0.1) is positive, and the relative performance of high-credit recipients
improved. For three indicators-inflation, external debt/exports, and debt service/exports-a positive difference is a worsening (shown by a minus sign). For the real
exchange rate, a greater real depreciation between periods than that of the comparators is an improvement.
a. pH I # p2 (see the discussion under "Implementation" in the text).
b. pHC I     # ,l o  ,c -  sL,c (see the discussion under "Implementation" in the text).
Source: World Bank data.



Macro Performance under Adjustment Lending 231
indicates that change in GDP growth was on average higher in the high-
credit group during 1982-86 than during 1978-81 relative to the change
for the low-credit group.
The nonparametric statistics in table 11-3 suggest a mild improvement
in the performance of high-credit countries relative to that of low-credit
countries. If an equal weight is given to each indicator, 63 percent of
high-credit countries in the low-income classification and 59 percent in
the middle-income classification improved when compared with corre-
sponding countries in the low-credit group. Interestingly, the pattern is
very similar across indicators for both income classifications.
On the positive side, the patterns suggest relative improvements in
high-credit countries in GDP growth, the current account and exports (via
a larger real exchange rate devaluation than in low-credit countries), and
debt service. On the negative side, the patterns suggest relative worsening
in the ratio of real investment to real GDP and of the budget deficit to
GDP.7 Although the worsening fiscal position is an unambiguous sign of
a deteriorating situation that was unintended by the adjustment pro-
grams, it could be argued that the falling share of investment in GDP was
intended and reflects a cut in overly ambitious public investment pro-
grams. Unfortunately, the available data do not allow for a breakdown
of investment into its public and private components.
Even though the value of each performance indicator represents an
average over several years, there is much variance in the calculated means,
especially among low-income countries. As the second column in table
11-3 indicates, none of the changes in means is significant for the high-
credit recipients in the low-income classification. Since most of the coun-
tries in this classification are Sub-Saharan African countries, in which
performance varied considerably and for which data are unreliable, this
should come as no surprise. For the middle-income countries, however,
the changes in performance are statistically significant for approximately
half the indicators. Note the significant improvement in GDP growth dur-
ing 1982-86 compared with 1978-81 and the worsening investment
performance.
Relative to low-credit comparators, high-credit recipients among low-
income countries significantly improved their performance in terms of
export growth, debt service, and the current account. This relative im-
provement may not be a sign of sustainable recovery, however, since
investment dropped. High-credit recipients among middle-income coun-
tries improved upon their own earlier performance for half the indicators.
Relative to low-credit comparators, however, their performance showed
no significant improvement (except for the current account and real ex-
change rate) and a significant worsening for investment. Here, too, in
spite of a mild relative improvement, there are no signs of a sustainable
recovery.



232 EVALUATION OF ADJUSTMENT PROGRAMS
As mentioned previously, this simple control-group approach suffers
from several shortcomings. In the companion paper using the same data
set (Faini and others forthcoming), we specify a simple statistical model
for a subset of the indicators used here. We postulate that changes in the
value of indicator i for country j, yij, depends on a vector of (unobserv-
able) autonomous policy changes, xi, on changes in the external envi-
ronment, SH,, and on participation (coNj = 1 if participation, otherwise
zero). Specifying autonomous policies as an adjustment process toward
the long-run values of the dependent variable allows us to write the model
as:
(11-2)   Ay,j = f3o, + (yl I),3lj + SHij2; + CON 3j + fi-
In equation 11-2 A refers to the difference between the post- and pread-
justment periods, and y'i - 1 is a vector of first-period averages of selected
indicators. In our sample, the control group is composed of thirty-two
countries that did not receive any IMF or World Bank loan during 1982-
86 and nine countries that received their first adjustment loan only in
1985 or 1986. This gives forty-one countries in the control group and
fifty-two countries that received IMF and World Bank adjustment lending.
In the latter group, only two countries received their first adjustment
credit in 1984, so that the statistical results relate to the performance of
fifty countries which received adjustment credits before 1984 and hence
were carrying out policy reforms whose performance-enhancing effects
were supposed to last throughout the period (until 1986).
Equation 11-2 was estimated for four indicators (GDP growth, invest-
ment share in GDP, inflation, and the current account) for the same sample
of countries as in this paper. In the estimation, period averages were for
1978-81 and 1982-86. The main results of Faini and others (forthcom-
ing) can be summarized as follows:
- Countries that faced the most unfavorable external environment fared
worse on each of the four indicators.
- Countries that participated in World Bank-IMF programs had a loss
in growth similar to that of the control group.
• Controlling for autonomous policy changes and for the external en-
vironment, countries that participated in World Bank-IMF programs
had investment rates similar to those of the control group.
* The current account of countries that participated in World Bank-
IMF programs improved more than the current account of the control
group (a result similar to one in Khan 1990).
Thus when one controls for the external environment and for auton-
omous policies that would have been pursued in the absence of adjust-
ment programs, one finds that growth performance was not statistically
different for countries that participated in World Bank-IMF adjustment



Macro Performance under Adjustment Lending 233
programs. Since the comparisons are for 1982-86 as opposed to 1978-
81, program effects are measured for a period of up to five years. This
result contrasts with the negative effect of IMF adjustment programs found
by Khan (1990), who used a similar methodology but allowed for pro-
gram effects to last only two years. Since structural adjustment is a long-
run proposition and adjustment lending is a short-run tool, it should
come as no surprise that when relevant factors are taken into account,
growth is not higher in countries with adjustment loans.
Performance of Adjustment Loan Recipients
The remainder of the chapter examines the interactions between World
Bank and IMF adjustment programs. The analysis therefore deals only
with the group of thirty countries that received World Bank structural
or sectoral adjustment loans before 1985. For this group of thirty coun-
tries, enough time has elapsed for a preliminary before-and-after analysis.
The focus here is on how the thirty recipients of adjustment loans com-
pare with nonrecipient countries. Nonrecipient countries are now defined
as countries that did not receive an adjustment loan from the World Bank
before 1985. We use the same low-income and middle-income classifi-
cations as before, but we do not control for IMF participation or for
lending intensity. We use the same set of indicators as before and the
same averaging method, but the average values for each performance
indicator are now taken over the three years preceding and the three
years following receipt of the first World Bank adjustment loan. As in
Balassa (1988), the year of receipt of the first loan is excluded from the
comparisons.
Although we are not controlling for IMF participation, we do note that
the thirty countries that received World Bank adjustment loans before
1985 account for about 84 percent of IMF adjustment loan commitments
and 59 percent of net IMF credit disbursements through 1986. The com-
parisons presented here should therefore be viewed as complementary
to, and a check for, the comparisons reported previously for combined
World Bank-iMF adjustment lending.
Table 11-4 gives the average value of the performance indicators for
the group of thirty recipient countries and the comparator group of sixty-
three nonrecipient countries. Both groups are further divided into the
low-income and middle-income classifications used earlier. Comparing
the average values for the indicators for the recipient and nonrecipient
groups generally reveals poorer performance for loan recipients in the
years preceding adjustment lending. With the exception of exports, the
growth indicators, on average, show no improvement. As before, there
is a marked decline in investment as a share of GDP. The indicators of
external balance show an improvement, however, which reflects the fact



Table 11-4. Average Values of Performance Indicators for Recipient and Nonrecipient Countries
Low-incomea                                               Middle-income"
Three years before            Three years after            Three years before             Three years after
Loan          Non-           Loan           Non-           Loan           Non-           Loan           Non-
Indicator               recipients     recipients     recipients     recipients     recipients     recipients     recipients      recipients
GDP' growth                0.030          0.032          0.023          0.029          0.023          0.046          0.022          0.024
Investment/GDP             0.19           0.21           0.18           0.19           0.25           0.28           0.19           0.25
~ Export growth                -0.02           0.06         -0.00            0.01           0.03           0.05           0.06           0.03
a    Real exchange rate          1.06          1.00           1.94           1.04           1.02           1.00           1.09            1.01
Current account
SurplUS/GDP             -0.09          -0.08          -0.07          -0.07          -0.07         -0.05           -0.05          -0.05
Budget surplus/GDP        -0.08         -0.04          -0.08          -0.04          -0.06          -0.06          -0.06          -0.06
Inflation                  0.26           0.14           0.33           0.13           0.26           0.20           0.38           0.32
External debt/exports      3.96           2.03           4.45           3.03           1.35           1.06           2.10           1.68
Debt service/exports       0.21           0.10           0.23           0.16           0.28           0.17           0.31           0.22
Note: The year of receipt of the first loan is excluded from the calculations, which are for three-year periods preceding and following the year of receipr of the first
loan.
a. Twelve recipients of adjustment loans and twenty-four nonrecipients.
b. Eighteen recipients of adjustment loans and thirty-nine nonrecipients.
Source: World Bank data.



Macro Performance under Adjustment Lending 235
that loan recipient countries generated the necessary noninterest current
account surplus to service the larger interest payments on their external
debt.
As has been analyzed elsewhere (for example, Cohen 1988 and Rodrik
1988), the burden of external debt servicing has been accompanied by
a deterioration (inflation) or stagnation (budget deficit) in the indicators
of internal balance. And for both low-income and middle-income ad-
justment loan recipients, external debt indicators deteriorated. The com-
bination of stagnating or deteriorating indicators for internal balance
and external debt raises questions about the long-run sustainability of
the achievements of adjustment programs to date.
Table 11-5 reports values of performance indicators for adjustment
loan recipients in the same format as table 11-3. A comparison of the
two tables shows similar results for middle-income countries. For low-
income countries, however, four of the nine indicators have different signs
in the two tables. As in table 11-3, tests of the changes in means for a
country's own performance show that the changes are insignificant for
low-income recipients. When the same comparisons are made for coun-
tries that received three or more adjustment loans, the share of indicators
showing improvement rises from 53 percent to 62 percent for the low-
income group and from 55 percent to 67 percent for the middle-income
group.
Tradeoffs between Growth and External Adjustment
For most countries, foreign borrowing was closed off after 1981, which
implied that the gap between absorption and income had to be reduced.
Corden (1989) refers to the resulting reduction in absorption as the pri-
mary (or inevitable) cost of reducing a current account deficit. The goal
of adjustment lending in the medium term was to raise capacity utilization
and to shift resources into the production of tradables at levels higher
than would have been attainable in the absence of adjustment lending.
This goal was to be achieved, in particular, through institutional reforms
and reforms of incentive structures, which would reduce the extent of
relative price rigidities. Since relative price rigidities cause a secondary
cost of adjustment, one can say that adjustment lending was intended to
reduce the secondary cost component of adjustment by increasing supply
responsiveness.8
In this section we look for evidence of whether the adjustment loan
recipients were able to achieve external adjustment at a lower cost of
forgone growth than did nonrecipients; that is, whether the change in
growth and the change in the current account are different for adjustment



Table 11-5. Performance Indicators for Recipients of World Bank Adjustment Loans Relative to Nonrecipients,
by Income Group
Low-income                                   Middle-income
Row sum
Relative to                                    Relative to                                   for
own           Intensive                       own            Intensive       Row       intensive
Indicator                         Number       performance'      recipientsb     Number       performances      recipientsb     sum       recipients
Loan recipients (AL)                 12                                5           18                                 7           30           12
Nonrecipients (NAL)                 24                                24           39                                39           63           63
GDP growth                            5-            -                  3           11 +            -                  5           16            8
Investment/GDP                        5 +           -                  2             6-                               2           11            4
Export growth                         8 +           -                  4            9 +            -                  5           17            9
Real exchange rate                   9 +            -                  5           11 +            -                  6           20           11
*    Current account surplus/GDP           6 +            -                  3           15 +            -                  6           21            9
ON   Budget surplus/GDP                    6 +            -                  3            6 +            -                  4           12            7
Inflation                            3-             -                  1           13-             -                  6           16            7
External debt/exports                 7+            -                  4             9_                               3           16            7
Debt service/exports                  8 +           -                  3             9 +           -                  5           17            8
Share showing improvement          0.S3                             0.62          0.55                             0.67         0.54         0.65
* Denotes 10 percent (or more) significance level of a test of equality of means for the recipient and nonrecipient results.
- Denotes statistical insignificance.
Note: The numbers in the table show for each indicator how many recipient countries in each classification improved relative to nonrecipient comparators in the three-
year period of the first loan compared with the three-year period before the loan. The year of receipt of the first loan is excluded from the calculations. The plus and
minus signs show the direction of change in the average value of an indicator for recipients compared with its average value for the nonrecipient comparators (a plus sign
indicates an improvement). Indicators and interpretation of results are the same as in table 11-3.
a. Only the test AL.` # FA` is reported here; the test for effectiveness cannot be performed because averaging over the control group (nonrecipient countries) is done
repeatedly.
b. Refers to the twelve countries that received three or more adjustment loans (see table 11-6).
Source: World Bank data.



Macro Performance under Adjustment Lending  237
loan recipients. To test this hypothesis we estimated the following re-
duced form equation:
(11-3)     AGDPKI = aO + a1D + a2 A(RESBAL/GDPK)i
+ a3D A (RESBAL/GDPK)i + Ei
where A is the percentage change between the three-year average after
receipt of the first adjustment loan and the three-year average before
receipt of the loan, A GDPK iS the change in real GDP, i = 1, . . . 93
countries, D = 1 for the thirty adjustment loan recipients (AL) and 0 for
the sixty-three nonrecipient countries (NAL), and A (RESBAL/GDP) is the
terms-of-trade-adjusted change in the resource balance over GDP. The
estimation is for the data appearing in figure 11-1. For countries that
did not receive a World Bank adjustment loan, 1982 was selected as the
year marking the break in three-year periods for measuring the change
in GDP.
After correcting for heteroskedasticity, the results are as follows (t
values in parentheses):
AGDPKi = -0.01 + 0.02 D -0.19 ARESBAL,
(11-4)             (-2.69)  (1.94)  (-3.03)
-0.71 D * ARESBALi.
(-0.31)
F = 10.8
The intercept dummy is of the expected sign (an upward shift for re-
cipient countries). The slope dummy, however, is of the wrong sign (that
is, a steeper resource balance to GDP tradeoff for recipient countries) but
insignificant.
Because of suspected multicollinearity between the slope and intercept
dummies, we exclude the slope dummy, which results in a further im-
provement:
(11-5)     AGDPKi = -0.01 + 0.02D- 0.20 ARESBALi.
(-2.71) (2.13) (-3.50)
F = 14.5
The regression results do not directly support the hypothesis that ad-
justment lending helped reduce the secondary burdens of adjustment since
the tradeoff between growth and improvement in the external balance
is the same for both groups of countries. However, the greater improve-
ment in the trade balance among recipients of World Bank adjustment
loans is consistent with the results in table 11-4, which shows these coun-
tries had a higher initial debt burden and consequently a need for greater
improvement in their current account. And the higher growth among



238 EVALUATION OF ADJUSTMENT PROGRAMS
Figure 11-1. Tradeoff between GDP and Resource Balance for
Recipient and Nonrecipient Countries
AADJ RB/GDP
20-              Malawi'
C6te d'lvoire.
Mauritius
10Niei
Mexico *Niea
Philippines .   x
Bolivia    /        Turkey  * Uruguay
NAL MY  /  *      y
NAL LY    Korea
-Yugoslavia
* Uganda
*Ghana.
-10* Senegal
,   I  _              I       IX   _
-10              0              10             20
AGDP
AQP = Percentage point improvement in the GDP growth
rate three years after the loan compared with three
years before.
AADJ RB/GDP =Percentage point improvement in terms-of-trade-
adjusted resource balance as a fraction of GDP three years
after the loan compared with three years before.
NAL LY = Average for low-income countries that did not receive
adjustment loans.
NAL MY =Average for middle-income countries that did not re-
ceive adjustment loans.
Source: World Bank data.



Macro Performance under Adjustment Lending 239
loan recipients is consistent with an increased availability of foreign
funds, which reduces the foreign exchange constraint.
Conclusions
In this chapter we used a simple control-group approach to assess quan-
titatively the economic performance of countries that received adjustment
lending from the World Bank and the IMF. The statistical evaluation
addressed three questions: (1) Was there complementarity between
World Bank-IMF lending and private lending? (2) Did recipients of ad-
justment loans improve their performance relative to their own past per-
formance? (3) Did recipients improve their performance relative to a
control group of nonrecipients?
With respect to the first question, there was a negative correlation
between the intensity of World Bank-IMF lending and private lending
during 1982-86. World Bank-IMF lending thus appears to have been a
substitute for rather than a complement to private lending.
With respect to their own past performance, countries with adjustment
loans improved their external position, a reflection of the fact that they
generated a sufficient surplus in their trade balance to service their ex-
ternal debt. Fiscal (and inflation) indicators deteriorated, however, in-
dicating that macroeconomic imbalances remained and casting some
doubt on the sustainability of adjustment. Growth rates also fell, a re-
flection of deteriorating terms of trade and the difficulties in achieving
the required reduction in absorption.
With respect to performance relative to comparators, nonparametric
statistics for the selected group of nine economic indicators show that,
overall, adjustment loan recipients improved relative to nonrecipients.
However, the percentage of indicators showing an improvement (the
average of the number of countries showing improvement in all indi-
cators) varied between 53 and 63 percent depending on classification.
For the group of twelve countries that received three or more adjustment
loans, the share of indicators showing an improvement varies between
62 percent for the low-income group and 67 percent for the middle-
income group. Statistical tests of difference of means for individual in-
dicators suggest that these improvements were often statistically insig-
nificant. In sum, the evidence suggests that some mild relative improve-
ment occurred.
Appendix: Data Sources and Country Classification
All data were extracted from the World Bank's economic and social data
base (BESD) and Andrex data base. Data in constant dollars were obtained
by using the exchange rate conversion factor from the World Bank's



240 EVALUATION OF ADJUSTMENT PROGRAMS
Table 11-6. Classification of the Ninety-Three Sample Countries
Credit                                  Credit
intensity           Country             intensity             Country
L        Algeria                        H         Madagascar
H        Argentina                      H         Malawi (1981, 4)
Bangladesh                    L         Malaysia
Barbados                      H         Mali
L        Benin                          L         Malta
Bolivia (1980, 2)             H         Mauritania
L        Botswana                       H         Mauritius (1981, 2)
Brazil (1983, 3)                        Mexico (1983, 2)
Burkina Faso                  H         Morocco (1984, 4)
Burundi                       L         Myanmar
L        Cameroon                       L         Nepal
H         Central African Republic      L         Nicaragua
L         Chad                          H         Niger
H         Chile                                   Nigeria (1983, 2)
L        China                                    Pakistan (1980, 4)
Colombia                      H         Panama (1983, 2)
Congo                         L         Papua New Guinea
H         Costa Rica (1983, 2)          L         Paraguay
H         Cote d'lvoire (1981, 3)                 Peru
L        Cyprus                                   Philippines (1980, 3)
H        Dominican Republic                       Portugal
H         Ecuador                       L         Rwanda
L        Egypt                          H         Senegal (1980, 2)
L        El Salvador                    L         Seychelles
L        Ethiopia                                 Sierra Leone (1984, 1)
Fiji                          L         Singapore
Gabon                         H         Somalia
Gambia                        L         South Africa
H         Ghana (1983, 5)               L         Sri Lanka
L         Greece                                  Sudan (1980, 2)
L        Guatemala                      L         Syria
H         Guinea                                  Tanzania (1981, 2)
H         Guinea-Bissau (1984, 1)                 Thailand (1982, 2)
L         Guyana (1981, 1)              H         Togo (1983, 2)
Haiti                         L         Trinidad and Tobago
Honduras                      H         Tunisia
L        Hong Kong                                Turkey (1980, 8)
H         Hungary                       H         Uganda (1983, 1)
India                         H         Uruguay (1984, 1)
Indonesia                     L         Venezuela
L        Israel                                   Yemen Arab Republic
H        Jamaica (1979, 6)                        Yemen, People's Dem. Rep.
Jordan                                  Yugoslavia (1983, 2)
H        Kenya (1980, 3)                H         Zaire
Korea (1981, 3)               H         Zambia (1984, 4)
L         Lesotho                       H         Zimbabwe (1983, 1)
H        Liberia
H = high-credit; L = low-credit.
Note: Numbers in parentheses are the date of the first World Bank adjustment loan and number
of adjustment loans; countries with three or more loans are the intensive recipients.



Macro Performance under Adjustment Lending 241
annual Atlas. In the calculation of external shocks (equation 11-1), terms
of trade indices were obtained by dividing the current values of exports
and imports (expressed in dollars) by the constant values. Similar results
were obtained when the terms of trade indices were calculated from cur-
rent and constant local currency values from national accounts data.
To calculate the effective interest rate on external debt, we applied the
London interbank offer rate (LIBOR) + 1 to the share of total non-
concessional debt and the implicit interest rate from interest payments
on concessional debt. For World Bank-IMF lending, we constructed two
variables, one based on gross IMF credit, another on net IMF credit (IMF
purchases less IMF repurchases). In both cases, World Bank adjustment
loan credits are the sum of structural and sectoral adjustment loan com-
mitments. Results based on gross iMF credit are not reported because they
are extremely close to those obtained for net credit.
Table 11-6 describes the sample of ninety-three countries and the clas-
sifications used in the text (low- and middle-income classifications are
from World Bank 1989, chap. 2). For the thirty World Bank adjustment
loan recipients, the year of receipt of the first loan is indicated in paren-
theses, followed by the total number of adjustment loans received. The
subset of intensive recipients consists of twelve countries that received
three or more adjustment loans.
Notes
1. The following discussion draws on Faini and others (forthcoming).
2. This is known as sample-selectivity bias. For a further discussion, see Gold-
stein and Montiel (1986), Khan (1990), and Faini and others (forthcoming).
3. The group of thirty adjustment loan recipients accounts for 84 percent of
the commitments of adjustment lending and for 59 percent of net IMF credit
disbursements through 1986.
4. The year 1982 is chosen as the breakpoint to correspond with the break-
point selected for measuring performance.
5. The formula is derived from a two-period maximization by firms and house-
holds under assumptions of perfect competition and wage-price flexibility (see
Dornbusch 1986, pp. 354-56).
6. Note that the definitions of both RIR and TOT are dependent on the signs
used. Specifically, RIR is interpreted as a worsening if it is positive (higher interest
rates) and an improvement if negative. Conversely, for TOT, which is defined as
X - M, a positive number indicates improvement and a negative number a
worsening. Thus in equation 11-1 improvement in the external environment is
a positive number and a worsening is a negative number, whereas for convenience
the opposite sign convention is used in table 11-1.
7. It should be noted, however, that the fiscal data are very weak and sparse.
8. Orthodox stabilization programs usually concentrated on expenditure-re-
ducing policies to reduce external imbalances. By contrast, the World Bank-iMF-
supported programs initiated in the early 1980s have placed greater emphasis
on expenditure-switching and, particularly, on output-augmenting policies as a
component of adjustment. Hence the characterization of these programs as "ad-



242 EVALUATION OF ADJUSTMENT PROGRAMS
justment with growth." World Bank-IMF-supported adjustment programs were
supposed to increase the responsiveness of demand and supply to relative price
shifts so as to increase the effectiveness of expenditure-switching policies. This
objective was to be achieved by greater reliance on the price mechanism and by
a reduction in distortions.
References
Balassa, Beta. 1988. "Quantitative Appraisal of Adjustment Lending." PPR Work-
ing Paper 79. Policy, Planning, and Research, World Bank, Washington, D.C.
Bourguignon, Frangois, William H. Branson, and Jaime de Melo. 1989. Ma-
croeconomic Adjustment and Income Distribution. OECD Development Center
Technical Paper 1. Paris: Organisation for Economic Co-operation and De-
velopment.
Cohen, Daniel. 1988. "The Management of the Developing Countries' Debt:
Guidelines and Applications to Brazil." World Bank Economic Review 2 (1)
January:77-104.
Corden, W. Max. 1989. "Macroeconomic Adjustment in Developing Countries."
World Bank Research Observer 4 (1):51-64.
Dornbusch, Rudiger. 1986. The Effects of OECD Macroeconomic Policies on
Non-Oil Developing Countries. World Bank Staff Working Paper 793. Wash-
ington, D.C.
Faini, Riccardo, Jaime de Melo, Abdel Senhadji-Semlali, and Julie Stanton. Forth-
coming. "Growth-Oriented Adjustment Programs: A Statistical Analysis."
World Development.
Genberg, Hans, and Alexander K. Swoboda. 1987. "The Medium-Term Rela-
tionship between Performance Indicators and Policy: A Cross Section Ap-
proach." EPD Discussion Paper EPD-01. Economic Projections Department,
World Bank, Washington, D.C.
Goldstein, Morris, and Peter J. Montiel. 1986. "Evaluating Fund Stabilization
Programs with Multicountry Data: Some Methodological Pitfalls." IMF Staff
Papers 33 (2) June:314-44.
Khan, Mohsin S. 1990. "The Macroeconomic Effects of Fund-Supported Ad-
justment Programs: An Empirical Assessment." IMF Staff Papers 37 (2)
June:195-231.
Khan, Mohsin S., and Malcolm Knight. 1981. "Stabilization Programs in De-
veloping Countries: A Formal Framework." IMF Staff Papers 28 (1):1-53.
Rodrik, Dani. 1988. The Welfare Economics of Debt Service. NBER Discussion
Paper 2655. Cambridge, Mass.: National Bureau for Economic Research.
World Bank. 1989. Adjustment Lending: An Evaluation of Ten Years of Ex-
perience. Policy and Research Series 1. Washington, D.C.



Macro Performance under Adjustment Lending 243
Comments
Peter J. Montiel
THERE IS A LONG TRADITION of using multicountry samples to assess the
macroeconomic effects of International Monetary Fund (IMF) financial
programs. The most recent survey I have seen is in a paper by Mohsin
Khan (1990). This chapter by Riccardo Faini and associates is the first
I have seen that attempts to apply methodologies from this literature to
an analysis of the macroeconomic effects of World Bank adjustment lend-
ing, and I agree with the authors that this is a useful complement to
country studies.
Since my previous work in this area concerned the methodological
pitfalls of using multicountry samples to assess the macroeconomic effects
of IMF-supported financial programs (Goldstein and Montiel 1986), my
comments on this chapter will focus on this set of issues. I will take the
specific questions addressed by the authors one at a time and provide
alternative interpretations of the evidence, based largely on methodo-
logical quibbles.
Faini and his associates construct a data set consisting of eleven years
(1976-86) of observations on nine macroeconomic variables for a group
of ninety-three developing countries. They then divide this sample in two
ways. First they rank the countries according to their use of World Bank
and IMF adjustment loans (relative to GDP) during 1982-86 and classify
the upper tercile into a group denoted the high-credit (HC) countries and
the lower tercile into a group denoted the low-credit (LC) group. The
former is the "treatment" group while the latter is the control group. A
second grouping of thirty countries that entered into World Bank ad-
justment loan arrangements before 1985 is the treatment group, while
the remaining sixty-three countries represent the control group.
The authors pose two questions as they begin their analysis: (1) Did
high-credit countries face a worse external environment than low-credit
countries during the 1982-86 period? (2) Were World Bank and IMF
loans complementary to commercial bank loans during this period? The
reason for the first question seems clear: the authors want to measure
the effects of adjustment lending by comparing pre- and post-1982 ma-
croeconomic outcomes, so they want to make sure that external factors
are not responsible for any observed differences in performance between
high-credit and low-credit countries. The second question relates to the



244 EVALUATION OF ADJUSTMENT PROGRAMS
authors' first exercise in measuring the effects of loans-in other words,
the response of commercial lending is one of the macroeconomic out-
comes they are interested in.
To answer the first question, the authors use a measure of external
shocks that seems a reasonable approximation of the welfare effects on
developing countries of changes in real interest rates and the terms of
trade:
AwIw = -(R2 - R)(D/Y)l + (PX2/PX1 - 1)(Xly)
- (PM2/PM1 - 1)(M/Y)l = -RIR + TOT
where the variables are defined in chapter 11. Does the measure Awlw
adequately summarize changes in the external environment? I suspect
not. Most economists would undoubtedly agree that the most significant
shock to the external environment facing developing countries in 1982
was the debt crisis. The changed availability of external financing is not
represented in this equation, which contains no mechanism to capture
the effect of credit rationing. Unfortunately, this means that, contrary to
the authors' conclusion, the external environment may well have been
systematically different for high- and low-credit countries.
With regard to the second question, the authors find a negative (al-
though not significant) correlation between World Bank-IMF lending and
net private credit and conclude that adjustment lending from these in-
stitutions did not serve as a catalyst for private lending during the 1982-
86 period. This is surprising since this period was precisely the heyday
of the active involvement of the Bretton Woods institutions in securing
private lending in support of adjustment programs. I think the problem
could be the use of cross-section tests. For example, low-income countries
tend to rely more on official financing than on private lending relative
to middle-income countries, and this country-specific characteristic could
account for the negative correlation observed in cross-section. Since a
number of countries would tend to rely preponderantly on either official
or private lending, these two variables would tend to be negatively cor-
related across countries. As shown in table 11-3, for example, half of
the high-credit countries proved to be in the low-income group while
only a third of the low-credit countries were in that group.
In the main part of the chapter the macroeconomic experience of coun-
tries that received adjustment lending is evaluated with the use of both
simple before-after and control-group approaches. The main conclusion
is that adjustment lending seemed to improve growth, but at the expense
of lower investment and a somewhat worse fiscal situation. The inter-
pretation offered is that growth may have improved because of the greater
availability of imports and that lower investment may have been due to
the weeding out of inefficient investment.



Macro Performance under Adjustment Lending 245
Although I am in strong agreement with the authors that the availa-
bility of imports may have been an important contribution to the positive
short-run impact of adjustment lending on growth, particularly in the
credit-rationed 1982-86 period, in my role as devil's advocate I would
like to offer an alternative interpretation based on sample-selectivity bias.
Sample-selectivity bias can occur in control-group studies such as this
whenever the criterion for selection into the treatment group (high credit
in this case) and control group (low credit) is nonrandom. Given non-
random selection, sample-selectivity bias will exist if the criterion that
determines selection is related to the determinants of post-treatment per-
formance. Bias results because the differences in post-treatment perfor-
mance that are due to the way the groups were chosen will tend to be
wrongly attributed to the treatment itself.
Is there reason to suspect nonrandom selection of countries to the
treatment and control groups in this case? A priori, I would say yes. I
would expect two kinds of countries to appear in the high-credit group
during the 1976-86 period. First would be low-income countries that
never had much access to commercial financing and have always been
heavy official borrowers. Some evidence on this was mentioned above.
Second would be countries that relied quite heavily on market borrowing
before 1982 but then became active IMF and World Bank borrowers be-
cause of the debt crisis.
These speculations are borne out by the numbers. Using World Eco-
nomic Outlook (IMF 1988) classifications, I found the composition of the
high-credit and low-credit groups shown in table 11C-1, which indeed
seems to suggest that selection was nonrandom, as one would expect.
Next, is there reason to believe that the criterion of selection for the
treatment group was related to post-1981 economic performance? Again
I would have to say yes. As I mentioned earlier, the main external shock
confronting these countries during the treatment period (post-1981) was
the debt crisis itself. This shock had a very severe effect on the highly
indebted countries but relatively little effect on small, low-income official
borrowers with no access to the private international capital market. As
Table 11 C-1. Composition of the High-Credit and Low-Credit
Groups by Financial Criteria
Criterion                           High-credit           Low-credit
Highly indebted countries               6                     1
Market borrowers                        2                    10
Diversified borrowers                   6                     9
Official borrowers                     17                    11
Small, low-income countries          12                     5



246 EVALUATION OF ADJUSTMENT PROGRAMS
Table 11 C-2. Growth and Investment of Alternative Developing
Country Groups, before and after the Debt Crisis
Before
Group and indicator              1982          1982-86          Difference
All developing countries
Growth                          5.2             3.1              -2.1
Investment/GDP                 26.5            23.8              -2.7
Highly indebted countries
Growth                          5.4             1.3              -4.1
Investment/GDP                 24.3            18.1              -6.2
Small, low-income countries
Growth                          3.4             3.1              -0.3
Investment/GDP                 20.2            17.2              -3.0
Note: All figures are unweighted arithmeric averages. The pre-1982 figures cover 1970-81 in
the case of growth, and 1980-81 for investment.
Source: IMF (1988).
is well known, the highly indebted countries suffered a sharp slowdown
in growth after 1981 and an even more severe collapse in investment.
Growth and investment in the small low-income countries did not move
nearly as much. Table 11C-2 provides the numbers.
Rather than adjustment lending resulting in relatively better growth
despite the relatively poorer performance of investment, my alternative
interpretation is the following: the criterion for selection was such that
the treatment group (the high-credit countries) contained many small,
low-income official borrowers for which the debt crisis resulted in rela-
tively better growth and a smaller number of highly indebted countries
for which the debt crisis meant a much worse investment performance
than average. Thus average growth was better for the high-credit coun-
tries as a group, and a number of countries in this group showed relatively
better performance on this criterion than the low-credit group. Average
investment was relatively worse for this group, however, because of the
magnitude of the investment collapses among the highly indebted coun-
tries.
References
Goldstein, Morris, and Peter J. Montiel. 1986. "Evaluating Fund Stabilization
Programs with Multi-Country Data: Some Methodological Pitfalls." IMF Staff
Papers 33 (2) June:314-44.
IMF. 1988. World Economic Outlook. Washington, D.C.: International Mone-
tary Fund.
Khan, Mohsin S. 1990. "The Macroeconomic Effects of Fund-Supported Ad-
justment Programs: An Empirical Assessment." IMF Staff Papers 37 (2)
June:195-231.



12 The Social Costs of Adjustment
Elaine Zuckerman
UNTIL RECENTLY, World Bank-supported adjustment programs have fo-
cused narrowly on economic problems, with little attention given to social
consequences. They have emphasized short-term restoration of balance
of payments equilibrium and financial stability (as have International
Monetary Fund stabilization programs). Their objectives have also in-
cluded improvements in the incentive system and restoration of medium-
to long-term aggregate growth. They have fallen short of expectations
in stimulating growth (World Bank 1988a), however, generally because
many adjusting countries, burdened by overwhelming debt, have had to
implement austerity measures that leave little room for growth (Fischer
1986).1
By the mid-1980s it was recognized that adjustment experiences, com-
bined with the legacy of inappropriate past policies, were negatively af-
fecting human welfare.2 UNICEF'S seminal book Adjustment with a
Human Face (Cornia, Jolly, and Stewart 1987) raised the development
community's consciousness about the need to cushion the poor from the
fiscal cutbacks under adjustment.3 The data needed to demonstrate that
social welfare is deteriorating are scarce, but it has become evident that
many countries with adjustment programs have reduced public expend-
iture on social services, cut back staffing in the public sector, and ex-
perienced higher unemployment. Real wages have eroded dramatically
in many adjusting countries, and social welfare has suffered.4 Today,
adjustment programs are gradually being broadened to emphasize con-
cerns with growth and equity. But more needs to be done.
Poverty Groups at Risk
Since 1987, World Bank staff guidelines concerning adjustment lending
and social welfare have required that policy framework papers for low-
income countries include "a brief description and assessment .. . of the
social impact of the government's intended adjustment program" and
that the president's report on structural adjustment loans and credits
(SALS) analyze the "short-term impact of the adjustment programs on the
urban and rural poor, and measures proposed to alleviate negative ef-
fects." An increasing number of these reports now estimate the expected
247



248 EVALUATION OF ADJUSTMENT PROGRAMS
impact on the "new poor"-those impoverished by adjustment programs
in the short term, such as retrenched public sector employees. Many SALS
require compensation for the new poor, such as severance payments or
retraining, as a condition of loan disbursement. A few recent SAL reports
have gone even further, analyzing the social impact on both the new and
the old poor and requiring measures to counter the negative effects of
adjustment operations on both groups (for example, in Chile).' With
lenders' assistance, some countries (for example, Bolivia and Ghana) have
initiated programs to alleviate adverse impacts on the poor during the
adjustment process through complementary compensatory operations.
Although not all adjustment programs incorporate measures to protect
the poor or undertake systematic analysis of the likely impact, there is
an accelerating trend in this direction.
Recent World Bank symposiums on poverty and adjustment and ad-
justment lending have shown that confusion and controversy surround
the question of which category of the "poor" should be considered during
adjustment.6 Three broad groups are candidates. (Note that females tend
to be overrepresented in all of these groups.7)
First are the new poor-the direct victims of adjustment, such as re-
trenched civil servants and public and private enterprise employees who
were laid off because of austerity measures or shifts in production. They
include former middle-class professionals who were better off before the
adjustment and who, although they have become poorer, are not nec-
essarily poor (although some of them may be).
Second are the borderline poor-low-income, vulnerable groups that
are most affected by changes in the availability and price of major items
of consumption, especially food, and by cutbacks in social programs.
Recessions and adjustment hit these groups severely. The borderline poor
tend to include some of the elderly, infants, landless and poor farmers,
and lactating and pregnant women.
Third are the extreme or structural poor.8 Some members of this group
are directly affected by adjustment, such as those in cities who are subject
to increased food prices, while others, usually those in rural areas, are
little affected because they live outside the influence of production forces
and public services. This is particularly the case in Africa and parts of
Asia, where 80 to 90 percent of the poor live in rural areas, producing
much of what they consume, and seldom benefiting from government
services. This is less true of Latin America, where large portions of the
poor are urban dwellers who receive some subsidies and services.
In general, the poorest 10 to 20 percent of a population often reside
in rural areas and suffer from malnutrition, illiteracy, disease, short life
expectancy, and high infant mortality rates (Lipton 1988). Traditional
investment projects have seldom reached them, either because they are
too weak and poor to contribute the required energy and assets or because



The Social Costs of Adjustment 249
projects were not directed to them. Since most of the extreme poor in
developing countries are potentially productive, singling them out for
special assistance during adjustment programs, by increasing their access
to employment and assets, could contribute to economic growth.9
Following much debate at a 1988 symposium on poverty and adjust-
ment, developing country administrators reached a strong consensus that
measures to help the poor under adjustment should cover all the poor,
not just the new or borderline poor (World Bank 1988c). Some devel-
opment agency officials felt, however, that the complexity of adjustment
programs precludes attention to issues such as long-term structural pov-
erty.
Data Approaches and Methodologies for Measuring Social
Welfare
Little hard data exist yet to demonstrate that adjustment adversely affects
the poor or other socioeconomic groups. Still, an emerging literature,
pioneered by Cornia, Jolly, and Stewart (1987), compassionately argues
that the poor are worse off because of adjustment. For several reasons,
the arguments so far are based largely on anecdotal evidence. First, many
adjustment programs are still in the early stages and their medium- and
long-term effects cannot yet be determined. Second, in trying to account
for the primary causes of deteriorating social welfare, it is difficult to
separate the influence of adjustment measures from the negative legacy
of inappropriate past policies. Third, accurate longitudinal data on wel-
fare distribution are woefully inadequate.
What data can be used to assess the impact of adjustment on social
welfare? A variety of biological and socioeconomic measures are em-
ployed, but they must be viewed cautiously. Most lack a sufficient level
of disaggregation, and differences among and within income groups are
obscured by the presentation of only national per capita averages. Meth-
ods of data collection may be suspect, and even relatively reliable country
data may not be comparable across countries. All these caveats should
be kept in mind during the following description of popular data ap-
proaches.
Basic Indicators
Time-series data covering nutrition, infant and child mortality rates, life
expectancy, disease, anthropometry, school enrollment, and poverty cut-
off lines are popularly used to measure poverty. The accompanying tables
present cross-country data for per capita caloric intake as a percentage
of daily requirement (1970-85) and infant mortality rates and life ex-



250 EVALUATION OF ADJUSTMENT PROGRAMS
pectancy data (1970-87) for the twenty-three countries reviewed for an
earlier report on adjustment lending (World Bank 1988a).
Data on per capita caloric intake as a percentage of daily requirement
(table 12-1) are available only through 1985, but interesting trends are
observable. (Annual fluctuations of a few percentage points are normal
in all countries.) Overall, standards for caloric intake were maintained
or increased everywhere over the entire period except in five of the
twenty-three sample countries, where significant drops occurred (Ghana,
Guinea, Peru, Zaire, and Zambia; Zambia began and ended the period
at 92 percent of the required caloric intake, but it had improved to 102
percent in 1977 before dropping back to 92 percent). These countries
either postponed adjustment indefinitely (Peru), began undertaking it
only after protracted delays (Ghana and Guinea), or failed to sustain it
(Zaire and Zambia). In addition, Kenya's per capita caloric intake dipped
slightly from 97 percent to 95 percent of daily requirement, dropping
further below a nutritionally acceptable level. Except for these countries
plus Bolivia, Guinea-Bissau, and Pakistan, all the other adjustment coun-
tries studied achieved nutritionally adequate levels of per capita caloric
intake. The latter three countries improved their intake over the period
but still remained below the accepted standard. Sub-Saharan Africa as a
whole fared worse than other regions. Thus per capita caloric intake
appears to reflect both the state of the economy and the government's
degree of commitment to sustained macroeconomic adjustment.
All countries made steady overall progress in reducing infant mortality
rates and increasing life expectancy (tables 12-2 and 12-3). Although
progress has been gradual for some and more dramatic for others, reces-
sions and adjustment did not impede it. These gains are attributable
largely to medical advances. The level of achievement differs vastly from
country to country, however. Of the twenty-three countries surveyed,
Costa Rica and Jamaica scored best, both with an infant mortality rate
of 18 per thousand in 1987 (down from 60 and 40, respectively, in 1970)
and a life expectancy of 74 years (up from 67 years in both countries in
1970). Malawi had the worst infant mortality rate (150, down from 193),
despite steady progress, and Guinea and Guinea-Bissau had the lowest
life expectancy at 43 and 40 years, respectively (up from 36 years for
both). No correlation exists between achievements on these two indi-
cators and per capita caloric intake as a percentage of daily requirement.
Malawi, for example, made excellent progress in increasing caloric con-
sumption, reaching 104 percent of intake in 1987. Regionally, Sub-
Saharan Africa again lags seriously behind in both these indicators.
To conclude, per capita caloric intake declined during adjustment, but
the other two most popular basic health indicators, infant mortality rate
and life expectancy, improved notably despite recession and adjustment.
This is because nutritional status is more sensitive to economic changes



Table 12-1. Per Capita Caloric Intake as a Percentage of Daily Requirement in Twenty-Three Developing Countries,
1970-85
Country       1970  1971   1972   1973   1974   1975  1976  1977   1978  1979   1980  1981   1982   1983   1984   1985
Bolivia       83.2   82.9   83.6   81.5   83.2   84.4  85.4   85.4   85.5   86.4  87.1   88.2   90.2   84.7   89.8   90.8
Brazil       102.8  104.0  103.0  102.9  104.3  104.2  105.0  105.3  105.9  109.5  110.3  109.5  110.5  109.6  110.2  111.2
Chile        109.5  109.9  112.4  106.1  109.1  103.7  104.7  103.6  105.5  107.6  107.8  109.4  107.9  107.4  106.6  104.3
Colombia      92.4   95.6  96.1   97.8   98.2  100.2  105.0  106.0  105.9  105.2  108.5  110.3  109.7  110.9  110.9  111.6
Costa Rica   107.2  109.9  110.6  111.6  111.8  115.0  116.3  116.3  115.8  116.8  117.0  117.2  118.5  120.8  125.1  125.3
Cote d'lvoire   103.4  102.6  101.5   99.7  100.4  99.6  100.0  101.1  105.2  108.1  113.1  112.4  112.6  109.5  108.4   99.9
Ghana         96.4   96.5   94.8   95.6   95.6  94.0  92.9   85.5   84.2  80.4  78.1   74.3   69.7   65.4   76.0   77.6
Guinea        84.8   80.4   79.4   83.3   85.1   83.9  82.1   82.5   83.7  78.4   76.9   74.3   75.7   74.2   74.8   74.9
Guinea-Bissau   83.7  82.2   79.7   79.5   76.8   78.2  77.9  76.4   78.1   78.2   80.0   86.3   85.5   83.1   84.2   89.7
N   Jamaica       113.4  117.7  119.6  115.2  116.4  119.6  120.3  118.9  119.4  115.0  115.4  114.1  112.7  114.6  115.4  115.1
Kenya         96.9   95.7   98.5   97.3   97.2   96.2  97.5   97.5   97.1  95.1   94.7   93.7   94.3   91.4   92.7   95.4
Korea        108.5  109.4  112.9  115.1  114.9  118.7  118.3  117.7  120.4  121.4  119.6  119.4  120.6  120.0  120.9  119.4
Malawi        99.9  104.4  101.6  106.4  105.6  106.3  108.5  108.6  109.9  106.8  105.7  107.2  105.4  104.4  105.5  104.1
Mexico       116.3  115.6  117.3  119.0  119.9  122.7  121.3  123.2  126.0  128.2  131.2  133.7  134.6  134.7  136.4  134.2
Morocco      100.2  101.7  105.0  105.3  105.3  105.9  107.8  112.4  112.7  113.6  112.9  111.7  113.3  109.7  110.7  112.8
Pakistan      87.8   88.6   88.4   87.4   89.7  90.3  94.2  92.9   92.4  95.8   94.8   97.8   98.2   96.0   93.5   94.4
Peru          97.8   98.7   96.2   95.3   97.4  96.7  96.0  94.0   93.1   93.2  92.1   92.9   94.3   91.1   92.4   90.2
Philippines   91.5   90.8   89.3   90.7   94.4  92.8  96.3   99.0  100.8  104.6  104,3  103.7  103.1  103.4  103.6  100.0
Senegal       98.9   97.4   95.9  94.0   95.3   95.1  95.5   98.5   95.8  100.1  100.6  100.5   99.1   94.9   98.4  101.6
Thailand     101.8  101.8  101.9  105.1  106.8  107.0  107.5  106.9  108.0  108.5  107.0  109.5  108.6  110.8  110.9  108.1
Turkey       111.9  113.7  115.0  114.7  115.0  117.6  118.8  119.2  121.6  122.7  123.1  123.7  124.9  125.2  125.7  127.7
Zaire        102.8  101.7  102.3  103.7  103.4  103.1  103.1  101.0   96.2  95.3   95.7   96.5   97.5   97.1  97.0   96.9
Zambia        92.2   97.1   98.3   95.3   98.7  101.0  101.6  101.8   96.1   96.6   96.4   93.2   91.4   91.1   92.5   92.0
Source: World Bank data.



Table 12-2. Infant Mortality Rates in Twenty-Two Developing Countries, 1970-87
(deaths per thousand live births)
Country         1970     1971     1972    1973    1974       1975    1976    1977      1978    1979    1980       1981    1982    1983       1984    1985'   1986b    1987
Bolivia         153.4    152.2    151.0   148.4   145.8   143.2   140.6   138.0   135.2   132.4   129.6   126.8   124.0   121.2    118.4   115.6    112.8    110.0
Brazil           94.6     92.8     91.0    88.6    86.2       83.8    81.4    79.0      77.4    75.8    74.2    72.6    71.0         69.4     67.8    66.2    64.6       63.0
Chile            82.2     73.9     76.5    65.8    65.2    57.5    56.6    50.1         40.0    37.8    33.0       26.9    23.5    21.6       19.6    19.5      19.8     20.0
Colombia         79.0     77.0     75.0    72.0    69.0       66.0    64.5    63.0      61.3    59.7    58.0       55.5    53.0    51.6       50.2    48.8    47.4       46.0
Costa Rica       61.5     56.6     54.4    44.8    37.5       37.8    33,1    27.9      23.8    23.2    20.1       17.9    19.3    18.6       18.9    18.6    18.3       18.0
C6te d'lvoire   134.6    131.8    129.0   126.4   123.8    121.2   118.6   116.0    113.8   111.6   109.4   107.2   105.0   102.8    100.6    98.4    96.2               94.0
Ghana           110.0    108.5    107.0   106.0   105.0    104.0   103.0   102.0    101.2   100.4    99.6          98.8    98.0    95.8       93.6    91.4    89.2       87.0
Guinea          185.4    183.2    181.0   179.0   177.0    175.0   173.0   171.0    168.6   166.2   163.8    161.4   159.0   156.2   153.4   150.6    147.8    145.0
Jamaica          39.6     37.8     36.0    33.8    31.6       29.4    27.2    25.0       24.2    23.4    22.6      21.8    21.0    20.4       19.8     19.2     18.6     18.0
Kenya           102.0    100.0     98.0    96.0    94.0       92.0    90.0    88.0      86.4    84.8    83.2       81.6    80.0    78.4       76.8    75.2      73.6     72.0
Korea            51.4     49.2     47.0    44.6    42.2       39.8    37.4    35.0      34.0    33.0    32.0       31.0    30.0      n.a.     n.a.    27.0    25.0        n.a.
Malawi          193.4    192.2    191.0   188.2   185.4    182.6   179.8    177.0    174.2   171.4   168.6    165.8   163.0   160.4    157.8    155.2   152:6    150,0
Mexico           73.0     71.0     69.0    67.2    65.4       63.6    61.8    60.0      58.6    57.2    55.8       54.4    53.0    51.8       50.6    49.4    48.2       47,0
Morocco         128.4    125.2    122.0   119.6   117.2    114.8   112.4   110.0    107.4   104.8   102.2          99.6    97.0      n.a.     n.a.    90.0      85.0      n.a.
Pakistan        142.4    141.2    140.0   138.2   136.5    134.7   133.0   131.2    129.0   126.7   124.5    122.3   120.1   117.8    115.6   115.0    111.0              n.a.
Peru            116.4    113.2    110.0   109.0   108.0    107.0   106.0   105.0    103.8   102.6   101.4    100.2    99.0    96.8            94.6    92.4    90.2       88.0
Philippines      66.4     65.2     64.0    62.0    60.0       54.0    56.0    54.0      53.4    52.8    52.2       51.6    51.0      n.a.     n.a.    48.0    46.0        n.a.
Senegal         164.4    163.2    162.0   160.4   158.8    157.2   155.6   154.0    151.6   149.2   146.8    144.4   142.0   139.0    136.0   133.0   130.0    127.0
Thailand         88.0'     n.a.     n.a.     n.a.    n.a.     n.a.     n.a.     n.a.     n.a.     n.a.     n.a.    n.a.     n.a.     n.a.     n.a.    43.0    41.0        n.a.
Turkey          133.6    133.8    134.0      n.a.    n.a.     n.a.     n.a.     n.a.     n.a.     n.a.     n.a.    n.a.     n.a.     n.a.     n.a.    84.0    79.0        n.a.
Zaire           131.0    129.0    127.0   125.0   123.0    121.0   119.0   117.0    115.0   113.0   111.0    109.0   107.0   105.2    103.4   101.6    99.8              98.0
Zambia          106.0    103.0    100.0    98.8    97.6       96.4    95.2    94.0      92.8    91.6    90.4       89.2    88.0    86.4       84.8     83.2    81.6      80.0
n.a. Not available.
a. Data for Korea, Morocco, Philippines, Thailand, and Turkey from World Bank (1987).
b. Data for Korea, Morocco, Pakistan, Philippines, Thailand, and Turkey from World Bank (1988d).
c. Data are for 1964 (World Bank 1987).
Source: World Bank data.



Table 12-3. Life Expectancy in Twenty-Three Developing Countries, 1970-87
(years)
Counrny        1970    1971    1972    1973    1974    1975    1976    1977    1978    1979    1980    1981    1982            1983    1984      198s    1986    1987
Bolivia        46.1    46.4    46.7    47.1    47.5    47.9    48.3    48.6    49.1    49.5    49.9    50.3    50.7            51.4    52.1    52.7    53.4    54.0
Brazil         59.0    59.4    59.8    60.2    60.6    61.0    61.4    61.8    62.1    62.5    62.8    63.1    63.4    63.8    64.1              64.5    64.9      6s.3
Chile           62.4    63.0    63.6    64.3    65.0    65.7    66.4    67.2    67.9    68.7    69.s    70.2    71.0    71.1    71.2    71.3    71.5    71.6
Colombia       59.1    59.4    59.8    60.2    60.6    61.0    61.4    61.8    62.2    62.6    63.1    63.5    64.0    64.3    64.7    65.0    6s.4               65.7
Costa Rica      67.1    67.6    68.1    68.4    68.7    69.0    69.4    69.7    70.4    71.0    71.7    72.4    73.0    73.2    73.3    73.4    73.5    73.7
C6te d'lvoire   44.5    45.0    45.5    46.0    46.5    47.0    47.5    48.0    48.5    49.0    49.5    50.0    50.5    50.9    51.4    51.8    52.3               52.7
Chana           49.4    49.7    50.0    50.2    50.4    50.6    50.8    51.0    51.2    51.4    51.6    51.8    52.0    52.5    52.9    53.4    53.9    54.3
Guinea          36.2    36.5    36.9    37.2    37.6    37.9    38.2    38.6    38.9    39.2    39.6    39.9    40.2    40.7    41.2    41.7    42.3              42.8
Guinea-Bissau    35.5    35.5    35.6    35.7    35.8    35.8    35.9    36.0    36.3    36.7    37.0    37.3    37.7    38.0    38.4    38.7    39.0    39.4
Jamaica         67.2    67.5    67.8    68.0    68.3    68.5    68.7    69.0    69.7    70.5    71.3    72.1    72.9    73.1    73.2    73.3    73.4    73.6
Kenya           50.0    50.5    51.0    51.5    52.0    52.5    53.0    53.5    53.9    54.4    54.9    55.4    55.9    56.3    56.7    57.1    57.5    57.9
Korea          59.9    60.7    61.4    62.3    63.1    63.9    64.7    65.5    66.0    66.4    66.8    67.3    67.7    68.1    68.4    68.8    69.1    69.5
Malawi          40.4    40.7    41.0    41.3    41.6    41.9    42.1    42.4    42.8    43.1   .43.4    43.7    44.1           44.4    44.8    45.1    45.4    45.8
Mexico          61.7    62.2    62.6    63.2    63.7    64.3    64.8    65.4    65.8    66.2    66.6    67.0    67.4    67.6    67.9    68.1    68.4    68.6
Morocco        51.9    52.4    52.9    53.5    54.1    54.6    55.2    55.8    56.2    56.7    57.2    57.7    58.2    58.6    59.0    59.5    59.9    60.4
Pakistan        46.1    46.3    46.5    46.8    47.1    47.4    47.7    48.0    48.4    48.8    49.2    49.6    50.0           50.4    50.8      51.2    51.5     51.9
Peru            53.9    54.7    5s.6    55.8    56.1    56.4    56.7    57.0    57.3    57.6    57.9    58.3    58.6           59.1    59.5      60.0    60.4    60.9
Philippines     57.2    57.5    57.9    58.3    58.7    59.1    59.s    59.9    60.2    60.6    61.0    61.4    61.8    62.2    62.6    63.1    63.5    63.9
Senegal         42.6    42.9    43.2    43.5    43.9    44.2    44.6    44.9    4s.0    45.1    45.2    45.2    45.3           45.8    46.2    46.7    47.1    47.6
Thailand        58.4    59.0    59.6    59.9    60.2    60.6    60.9    61.2    61.5    61.9    62.2    62.5    62.9    63.2    63.5    63.8    64.1    64.4
Turkey          56.2    56.8    57.5    58.1    58.7    59.3    59.9    60.5    61.0    61.5    62.0    62.5    63.0    63.4    63.8    64.2    64.6    64.9
Zaire           45.2    45.6    46.0    46.4    46.8    47.2    47.6    48.0    48.4    48.8    49.2    49.6    50.0           50.5    51.0      51.5    52.0    52.6
Zambia          46.5    46.9    47.3    47.7    48.1    48.5    48.9    49.3    49.6    49.8    50.1    50.3    50.6    51.1    51.6    52.1    52.6    53.1
Source: World Bank data.



254 EVALUATION OF ADJUSTMENT PROGRAMS
than are infant mortality and life expectancy and is therefore the most
important indicator of social welfare (Carlson 1988).1o Nevertheless, as
national averages these basic indicators mask inequalities among and
within income groups just as do per capita GDP data.
Public Sector Expenditures
Some commentators believe that a country's commitment to improving
basic needs is demonstrated by its expenditures on health, education,
housing, water supply, and other amenities. These expenditures are fre-
quently regressive, however (that is, they are spent on sophisticated urban
hospitals, higher education, and other services largely for higher-income
groups), and therefore they need to be disaggregated by beneficiary group
before their impact on the poor can be evaluated. Of the World Bank
reviews of public expenditure and investment, which often precede or
accompany adjustment operations, only a handful have examined the
effects on income distribution, although most have examined social sector
expenditures. "
In 1988 equity issues were more systematically addressed in reviews
of public social expenditures in Argentina, Brazil, Jamaica, and the Phil-
ippines. They confirm that (except in Jamaica, which has relatively eq-
uitable spending) expenditures are grossly mistargeted to higher-income
and urban groups and that social sector bureaucracies have excessively
high administrative costs. Public social expenditures in these countries
in 1985-87 ranged from 18 percent of GNP in Brazil to 13 percent in
Jamaica to only 3.6 percent in Argentina and the Philippines.12 The
World Bank is preparing public social expenditure reviews for other bor-
rowing countries as well. These reviews are a valuable source of infor-
mation on income distribution and provide a rationale for countries with
budgetary constraints to redirect inequitable spending toward the poor.
What is more, they have demonstrated that social bureaucracies are ex-
tremely inefficient and wasteful. Each of the 1988 reviews of social ex-
penditure will be followed up with World Bank loans for projects and
policy reform to help the countries achieve greater equity and efficiency
in their social expenditures.
Household Surveys
The most promising data approach relies on household surveys, which
disaggregate data to the household level, thereby differentiating among
income groups. It must be cautioned, however, that past household sur-
veys have not adequately captured intrahousehold consumption patterns
(Glewwe and van der Gaag 1988).13 In many countries women and chil-
dren, especially girls, are fed less nourishing food than men, girls receive



The Social Costs of Adjustment 255
less education than boys, and women have less access than men to health
care and productive assets within the same household (Dwyer and Bruce
1988). Such inequitable distribution within the household is not captured
by the traditional nondisaggregated household model (Gillespie 1989).
Given intrahousehold inequalities, individuals rather than households
should be the recipients of economic outlays (Dwyer and Bruce 1988).
Some household surveys are being refined to account for gender and for
other intrahousehold differences in consumption. Since it is unlikely that
nutritional status within the family could be captured directly, anthro-
pometric studies serve as a proxy.
Fairly comprehensive household surveys such as the World Bank Living
Standards Measurement Study (LSMS) collect data on consumption, in-
come, anthropometric measures, basic needs, and other aspects. They
furnish basic social indicators on consumption (nonfood, food, home-
produced, purchased, prices), employment (self or wage), income, assets,
transport, education (enrollment, attainment), health (status, use of med-
ical care), nutrition, water supply, fertility, migration, and so on. Much
of the expenditure data are disaggregated to the individual level, as are
anthropometric data.
Ideally, the longitudinal data of these surveys would be analyzed be-
fore, during, and after adjustment in order to monitor the impact on
income groups. This is already being attempted for C6te d'lvoire (Kanbur
1988a, Glewwe and de Tray 1988) and hypothetically (in that little ad-
justment has occurred) for Peru (Glewwe 1988b and 1988c), based on
the LSMS data available for these countries (see appendixes A and B to
this chapter). Data from other household surveys have also been used to
estimate the impact of adjustment on poverty, including a recent study
on South Asia (Srinivasan and Bardhan 1988) and one on the Philippines
(Blejer and Guerrero 1988). LSMS data are becoming available for Bolivia,
Ghana, Jamaica, and Mauritania. LSMS surveys are also being launched
in other countries.
In a few years, the Social Dimensions of Adjustment project in Sub-
Saharan Africa will start yielding LSMs-type data for some twenty-five
participating countries. The project questionnaires are being designed to
capture intrahousehold gender differences.
Analytical Methodologies
Mapping the impact of adjustment programs on social welfare is difficult.
Some feel that reliable analytical methodologies to do so have yet to be
formulated. Kanbur (1988a) has attempted to quantify the impact of
adjustment on income groups with the use of a household survey. A
variety of other methodologies employ household and other available
data. Another approach uses economywide models such as computable



256 EVALUATION OF ADJUSTMEN-T PROGRAMS
general equilibrium models, which simulate the distributional impact of
adjustment (de Melo and Robinson 1982; Bourguignon, Branson, and
de Melo 1989; Narayana, Parikh, and Srinivasan 1987), and input-out-
put models such as social accounting matrices (Pyatt and Round 1985),
which claim not only to organize data but to analyze the distribution of
social welfare as well."4 A useful complement to economywide models
is partial analysis, which can be used to evaluate specific target groups,
regions, and sectors.
Although much sectoral analysis has been undertaken, evaluation of
experience with targeted programs, asset distribution, rural public works,
and other schemes is still needed for many countries. All of the above
methodologies can be used complementarily to contribute to our under-
standing. Furthermore, judgments concerning the political economy are
essential for all countries.
Adjustment Measures Affecting Social Welfare
Several common features of adjustment programs particularly affect the
distribution of social welfare, some positively, others negatively. These
features and their distributional consequences are listed below."5
1. Currency devaluation and export promotion. These measures in-
crease the price of tradable goods relative to nontradable goods and thus
generally benefit exporters, including farmers, processors, and manufac-
turers. As the price of nontradable goods falls in relative terms, farmers
growing nonexportable crops and others engaged in nontradable activi-
ties (such as construction, utilities, services, and informal sector activities)
are hurt. Consumers, including many of the poor, are hit with higher
prices for food and other goods.
2. Liberalization of prices and lifting of subsidies. The impact of these
measures on the poor is severe if the general subsidies on food, public
transport, energy, water, or other goods they consume disappear. Sim-
ilarly, the removal of subsidies on farmers' inputs such as fertilizer may
hit the rural poor. The removal of other subsidies, such as those on
gasoline or middle-class housing, will affect the nonpoor more and the
poor less. If the poor do consume such "luxury" items as gasoline or
electricity, however, they may spend a larger portion of their budget on
them. To the extent that poor farmers are engaged in the production of
tradables and not merely in subsistence agriculture, they would benefit
from increased producer prices.
3. Reduction of public sector staffing and cuts in public sector salaries.
Civil servants and public enterprise employees who find themselves out
of work and income as a result of staffing cutbacks or liquidation of
public enterprises tend not to be destitute, and many may well earn in-
come from other sources. An increasing number of adjustment loans



The Social Costs of Adjustment 257
include compensation for this group. It may take the form of full or partial
salary for a fixed term (usually six months to two years), retraining pro-
grams, job creation schemes such as emergency work and food-for-work,
or, occasionally, resettlement incentives to relocate in rural areas (this is
being tried in Guinea-Bissau and Mauritania, for example), a difficult
transition for most urbanites. Assessing the impact of adjustment on laid-
off employees who find new jobs (for example, in the informal sector)
requires comparing their new net income with their old. Salary cuts,
although not as drastic as job loss, also lower the standard of living.
4. Cuts in public sector social services or an increase in cost-recovery
measures. All income groups may be hit by such measures, but especially
the poor beneficiaries who lack the savings to pay for their own health,
education, and other services. Expenditure cuts particularly constrain
operation and maintenance budgets. Since recurrent costs usually con-
stitute the bulk of social sector spending, the poor suffer most to the
extent that they cannot afford alternative services.
5. Targeted social sector expenditures to protect and assist the poor.
Such measures are a rare feature within structural adjustment programs
(a SAL to Chile for health and social security is a rare example) but several
sectoral adjustment loans (SECALS) target the poor (education SECALs in
Ghana and Morocco and an agricultural SECAL in Mexico are examples).
In the future, there are likely to be more social sector SECALs and more
investment loans associated with SALS and directed toward the poor.
Solutions to Protect the Poor
Adjustment programs can further impoverish various income groups, but
there is considerable scope for policies and programs that benefit the
poor without undermining adjustment. Adjustment programs should be
designed to ensure that, at least in the short and medium term, the in-
comes and standard of living of the poor do not deteriorate further. In
the long term, the poor should be brought into the growth process by
redistributing assets, generating employment, and developing human re-
sources.
It is also important to keep in mind that some adjustment measures
alleviate poverty for some groups while exacerbating it for others. For
example, deregulating producer prices on agricultural products benefits
the farmers who market their crops, while it worsens the position of the
landless and poor urban dwellers who must pay the resultant increased
food prices. The interests of net sellers and net purchasers of food are
diametrically opposed, and there is a tradeoff between producer poverty
groups and consumer poverty groups (Besley and Kanbur 1987). More
generally, adjustment policies benefit those engaged in tradable activities
while those engaged in nontradable activities suffer. In the ideal long-



258 EVALUATION OF ADJUSTMENT PROGRAMS
term scenario, the dislocated would switch jobs and benefit from ad-
justment.
Short-Term Measures
Short-term measures, although no substitute for more important and
fundamental long-term antipoverty measures, are needed to ease the
hardship temporarily during adjustment. They include measures to ensure
the continued output of subsistence crops (rather than the single-minded
promotion of export crops) and to protect outlays to the poor in social
sectors such as primary education, health care, nutrition, and slum up-
grading. Adjustment programs typically raise agricultural producer prices
and reduce food subsidies, thereby causing higher food prices that hurt
the urban poor and rural landless. These groups can be partly protected
in several ways. Producer prices for certain food crops can be increased
more slowly. Carefully targeted food subsidies can be introduced or can
replace existing general or mistargeted subsidies. Such solutions entail
both a budgetary cost and an efficiency cost (because of the "under-
pricing" of the targeted subsidized commodity). The efficiency cost may
be justified by the social welfare benefits, while the budgetary problem
is not necessarily insuperable. There might be scope for raising taxes
among the better-off as a means of paying for popular poverty programs.
For example, when Tunisia attempted to reduce public expenditures on
food subsidies by doubling the price of wheat products, widespread pro-
tests that began in one of the poorest regions of the country forced a
change in approach. The government decided instead both to introduce
the price increase gradually (partly by halving the size of subsidized bread
loaves) and to recoup the budgetary outlay through additional excise
taxes. Options that target food subsidies specifically to the poor are also
being considered.16 Mexico, for example, has eliminated global food
subsidies in favor of targeted programs for the poor based on nutritionally
superior but socially inferior foodstuffs; Morocco is attempting to do the
same. Neither country could sustain the cost of global subsidies."1
For the new and existing poor, short-term solutions include transitory
compensatory measures such as emergency public works or food-for-
work programs that provide temporary employment. Severance pay for
retrenched workers is becoming a common feature of adjustment pro-
grams. Targeted food programs based on food stamps or rations are also
recommended to guarantee a minimal nutritional level. These compen-
satory programs are designed to ensure that the basic needs of the new
poor are maintained.
Multisectoral, quick-disbursing compensatory programs are also be-
coming popular (these are discussed in detail in Zuckerman 1989). Ex-
amples are Bolivia's Emergency Social Fund, Chad's proposed Social De-



The Social Costs of Adjustment 259
velopment Action Program, Ghana's Program of Actions to Mitigate the
Social Costs of Adjustment, Guinea's Socioeconomic Development Sup-
port Project, Guinea-Bissau's proposed Social and Infrastructure Relief
Project, Guyana's proposed Social Impact Amelioration Programme,
Haiti's proposed Economic and Social Fund Project, and Madagascar's
Economic Management and Social Action Project. These programs sug-
gest that labor-intensive, community-executed public works projects to
rehabilitate or create social infrastructure may constitute appropriate
types of intervention, but they need testing over time. These compen-
satory programs were originally designed as temporary measures (for
example, Bolivia's is a three-year program and Ghana's a two-year pro-
gram) on the assumption that the beneficiaries will be absorbed into the
sectors that are spurred into new growth by the reallocation of resources
that accompanies adjustment. Given the protracted experience with ad-
justment in most cases, however, it is clear that temporary compensatory
solutions are insufficient to protect and assist the vulnerable and extreme
poor. Some of these programs will need to be extended or made per-
manent.
For the borderline poor and extreme poor, it is recognized that greater
public social expenditures are required in both the short and long runs.
Even when budgetary constraints require that total social expenditure be
cut back, programs for the vulnerable and poor must be protected
through better targeting. For example, in Korea subsidized health pro-
grams aimed at the poor were expanded during the adjustment period
and other poverty programs were maintained. Unlike many other coun-
tries, Korea managed to alleviate poverty even during adjustment. In
many countries, however, social expenditures are not well targeted; they
reach the better-off (for example, through expenditures for higher edu-
cation and curative health care) and fail to benefit the poor. This mis-
targeting has exacerbated already highly regressive income distribution
patterns in many countries.
Even under conditions of severely constrained resources, savings could
be realized that would help to free resources for use in programs to assist
the poor during adjustment. Three methods are described briefly here.
First, targeting available resources to the poor provides enormous po-
tential for improving the equity of social expenditures without the need
for additional resources. Two complementary methods to accomplish this
are cost recovery and the shift of some service supply functions to the
private sector; both are based on the principle that those who can afford
to pay should do so. If at least some of the costs of public social services
that are provided to the better-off could be recovered, these resources
could be used to fund social services directed specifically to the poor.
Reducing and redirecting public expenditures is, of course, an admin-
istratively and politically complex effort and may need to be undertaken



260 EVALUATION OF ADJUSTMENT PROGRAMS
gradually to avoid serious social unrest. A different approach is to shift
the provision of social services for the better-off from the public to the
private sector, thereby enabling more public expenditures to be directed
specifically to the poor. Jamaica, for example, is privatizing some hospital
services while maintaining essential public health services for the needy.
Another advantage to this approach, as experience has shown, is the
efficiency gains that can be realized from the increased use of private
services among those who can afford them.
Second, administrative reform of social services can reap important
gains in efficiency. Most social sector bureaucracies in developing coun-
tries are extremely inefficient. If these operations were streamlined and
improved, considerable savings could be realized and applied to programs
for the poor. Social security reform is a special need of Latin American
countries, where expensive, unsustainable programs drain resources.
Third, taxation reform, a typical adjustment measure, could be de-
signed to assist the poor. Equity objectives, which should be central to
taxation reforms, are usually neglected. Progressive taxation can be in-
stituted through more effective tax collection and revision of the tax code
(often done to eliminate exemptions).," In Haiti and the Philippines, for
example, poor administrative and collection practices result in regressive
taxation. Underreporting and evasion are common in numerous coun-
tries. If supported by strong political will, tax reform can play a role in
redistributing income; for example, China's central government redis-
tributes nationally collected taxes to the poorest regions (although there
is inevitable leakage to the rich within those regions).
Long-Term Measures
Over the long term, the focus must be on bringing the vulnerable and
extreme poor into the production system as contributors to growth. Ad-
justment programs must be designed from the outset to restructure pol-
icies in the productive and social sectors so that they will reach the poor.
The adjustment process in most countries has proved to be longer and
more arduous than anticipated, often necessitating follow-up operations
to maintain the reform momentum. Thus it is not sufficient to think
simply in terms of short-term transitional measures.
Experience highlights that participation by poverty groups at the local
community level is a crucial ingredient for success (Lewis 1988). The
human capital of the poor should continue to be nurtured through the
targeted provision of education, training, clean water, health care, family
planning services, and social security. Their physical capital should be
developed through special credit schemes, such as the Grameen (Rural)
Bank program in Bangladesh; a reduction in regulatory barriers in the
informal sector, where the poor often find employment; and the redis-



The Social Costs of Adjustmnent 261
tribution of assets such as land, livestock, and tools in rural areas and
of jobs in urban areas. As the poor become more self-supporting, their
need for secondary income transfers is reduced (Addison and Demery
1988). Only policies that involve the poor in the process of economic
growth by increasing their access to productive assets and employment
will reduce poverty.
In summary, it is clear that there is technical scope for governments
to promote policies and programs to protect and assist the poor without
undermining adjustment. Both short- and long-term measures are nec-
essary. Short-term compensatory schemes, which are designed as an ad-
dition to adjustment programs, alleviate hardship only temporarily. What
is also needed is to design adjustment programs from the outset that aim
to increase the incomes, employment, and productivity of the poor.
Institutional Constraints
Programs targeted to the poor require strong institutions and staff with
strong administrative skills. In many developing countries, weak insti-
tutions are a greater constraint on such programs than is a lack of funds.
Indeed, donors are eager to finance compensatory programs as a way of
assisting the poor during adjustment, but weak institutions impede the
implementation of programs. The delay in implementing the Ghanaian
Program of Actions to Mitigate the Social Costs of Adjustment was at-
tributable mainly to weak institutions. Bolivia's short-term Emergency
Social Fund established a separate organization to circumvent the cum-
bersome functioning of its normal bureaucracy. This may not be a suit-
able option for other countries, however, particularly for those that ur-
gently need to strengthen their institutions and make them more
sustainable.
Not all countries have the technical skills for the data collection and
monitoring required to identify those who qualify for targeted programs.
Socioeconomic information on all the poor is especially difficult to collect
and update because the poor population tends to shift. To simplify this
process one can target economically distressed areas (rather than partic-
ular individuals) with social services and subsidize foods that are con-
sumed primarily by the poor and considered to be inferior goods by the
rich.19 Targeting programs too narrowly should be avoided, however,
because of higher administrative costs and greater leakages (Kanbur
1988b).
Political Sustainability
Although lending agencies and academics should continue to insist on
protecting the poor during adjustment, only strong and sustained political



262  EVALUATION OF ADJUSTMENT PROGRAMS
commitment on the part of policymakers will effect lasting changes. Key
measures to assist the poor during adjustment include shifting public
spending on social services away from the rich and middle classes to the
poor and improving the efficiency of social expenditures. These are highly
controversial measures in all countries because the middle classes tend
to dominate the political process while the poor have little political clout.
These political facts limit the scope and speed of reform.
Political opposition or support may make or break an adjustment pro-
gram. For example, during the Zambian attempt at structural adjustment
the removal of general food subsidies caused hardship to many more
people than expected. Not only did political support erode as the an-
ticipated benefits of reforms failed to materialize, but also social unrest
and riots undermined the program. Attempts to remove food subsidies
also led to riots in Algeria, Egypt, Morocco, Sri Lanka, and Tunisia. By
contrast, a fairly stable political situation coupled with positive economic
growth rates in Pakistan during the 1980s permitted a more successful
implementation of adjustment measures. In Chile the military junta,
which has been in power longer than any regime in the country's modern
history, has been able both to implement and to reverse adjustment re-
forms. In addition, Chile has targeted social programs since the 1920s,
and this long tradition facilitated the retargeting of resources to the poor
under the budgetary constraints of adjustment and made these programs
politically acceptable.
Popular social programs to protect the poor during adjustment also
contribute to the sustainability of adjustment programs, as Bolivia's ex-
perience with the Emergency Social Fund is proving. Many community-
based and grass-roots organizations have participated in and supported
the program, as well as other recent compensatory schemes. To gain
popular support for adjustment programs, some countries have mounted
propaganda efforts. Guinea-Bissau's president and cabinet tour the coun-
tryside to explain the rationale for reforms and short-term sacrifices, and
Guyana's political leaders convene rallies and distribute literature to
drum up popular support for the Economic Recovery Program.
Lending Instruments
Several types of World Bank lending instruments are available to assist
the poor during adjustment. Poverty alleviation measures can be incor-
porated into SAL operations; for example, a SAL in Chile includes con-
ditions for social sector reform, while many other SALS include provisions
for severance pay and retraining of retrenched workers. Sectoral adjust-
ment lending operations such as education SECALS in Ghana and Morocco
and the agricultural SECAL in Mexico can also incorporate measures to
alleviate poverty. Several adjustment loans for human resources or social



The Social Costs of Adjustment 263
sectors are under consideration for countries such as Brazil, Jamaica,
Senegal, and Venezuela. Some of these SECALS are hybrid loans combining
features of both investment and adjustment operations. Another ap-
proach is through parallel sector investment projects, such as Bolivia's
Emergency Social Fund and Ghana's Program of Actions to Mitigate the
Social Costs of Adjustment, which provide quick responses. Comple-
mentary technical assistance operations such as Argentina's Social Sector
Management Technical Assistance Project are another way to assist the
poor during adjustment.
Each country's circumstances and needs should govern the selection
of the appropriate option. Some argue that SALs are inappropriate for
most countries because they tend to load too many reforms onto one
adjustment operation (the "Christmas tree" approach). Chile, because
of its long tradition of targeted social programs, could more readily in-
corporate poverty alleviation measures into its adjustment program than
could most developing countries. Other countries might be better assisted
through pure or hybrid SECALS or investment loans that do not involve
competing policy reforms in the same package, as does a SAL. Although
it is doubtful that quick-disbursing SECALS can directly alleviate long-term
poverty, they can be a catalyst (for example, by accelerating expenditure
shifts toward the poor), and they can mitigate the social costs of ad-
justment. Whatever the option selected, all SAL policy reforms should be
designed so as to protect the poor from bearing the costs of adjustment
and to assist them in becoming productive participants in the nation's
economy.
Conclusions
It has become clear that the poverty-related aspects of adjustment are
not simply a short-term problem. Adjustment programs have taken
longer to implement and have proved far more arduous than originally
expected. Ideally, adjustment will eventually spur long-term growth and
an improved standard of living for the poor.20 In the meantime, however,
adjustment programs must be designed to protect and assist the poor.
Measures to mitigate the social costs of adjustment are feasible and need
not dilute the main objectives of adjustment.
Short-term, targeted interventions can ensure that the social welfare
of needy groups is maintained at an acceptable standard. Public expen-
ditures on social sectors should be redirected from the better-off to the
poor. Private social services and programs with greater cost recovery are
options for those who can afford to pay. Social bureaucracies need to
be made more efficient in order to achieve needed savings. Compensatory
programs are desirable complements to adjustment programs to mitigate
their social costs, but they are insufficient to ensure sustained benefits



264 EVALUATION OF ADJUSTMENT PROGRAMS
for the poor. Some adjustment measures, such as less contractionary
policies and progressive taxation, can also help the poor in the short
term. In the long term, however, the extremely poor must become in-
volved in the growth process through employment generation, human
resource development, and the redistribution of assets such as land. These
measures would contribute to adjustment's principal objective of real-
izing long-term growth and obviate the need for program targeting, which
should be used only in the short term. Involving the poor in the growth
process will also make adjustment more sustainable."
Appendix A. C6te d'Ivoire: A "Factual" Case
A prerequisite to an analysis of the impact of adjustment on the poor in
any country is an understanding of its poverty groups. Kanbur (1988a)
used LSMS data for C6te d'lvoire to disaggregate households into five
occupational types, for each of which he constructed a poverty profile.
According to his results, growers of food crops had the highest incidence
of poverty (49.5 percent of them were poor), followed by export croppers
(households that devote at least half their area to major export crops;
36.4 percent), those employed in the informal sector (19.3 percent), those
in the formal sector (6.1 percent), and finally those in the government
sector (3 percent). Conversely, government employees had the highest
per capita expenditures and food croppers the lowest.
Kanbur also established a poverty line based on per capita household
income. About 30 percent of Ivoirians fell below this line, and another,
"hard core" poverty line cut off the poorest 10 percent of the population.
He disaggregated his poverty groups into five regions: Abidjan, other
urban areas, west forest, east forest, and the savanna. He distributed the
sample households by ratio of agricultural to nonagricultural income, by
area cropped, and by type of crop; he crossed these with the socioeco-
nomic groupings listed above. Kanbur classified these categories by status
of the household head. The incidence of poverty in the savanna was over
60 percent, whereas in Abidjan it was only around 5 percent. Per capita
income and expenditure data support this finding. As for the hard-core
poor, 26 percent of those in the savanna and none in Abidjan ranked in
this group. Over 80 percent of the savanna population were food crop-
pers.
Kanbur used data from Glewwe (1988a) to examine some basic needs:
education, health, and housing were disaggregated by region and socio-
economic group. He found that the poorest groups were also worse off
in basic needs. For example, literacy and school attendance rates were
lowest in the savanna (11 percent) and highest in Abidjan (62 percent);
the percentage of sick people who consulted health personnel was sig-
nificantly lower for the poor than for the nonpoor; and although most



The Social Costs of Adjustment 265
poor owned their housing, they had less access to toilets, piped water,
electricity, and other amenities.
Next, Kanbur examined the three Cote d'Ivoire structural adjustment
loans of 1981, 1983, and 1986 to analyze the effects on poverty of ma-
croeconomic adjustments during the 1980s. None of these SALs explicitly
considered the potential effect on income distribution in any detail. The
only major reform of social policy was the elimination of the urban
housing subsidies, which had gone primarily to the better-off. The fol-
lowing were implications of the adjustment reforms:
* Many government sector employees were retrenched and lost their
income.
* Increased prices for tradables benefited export croppers and those em-
ployed in the formal private sector to process and manufacture for
export.
* Food croppers and informal and government workers were adversely
affected because they were involved in nontradable activities. To the
extent that demand for food increased on the part of the gainers from
adjustment, food croppers earned extra income; but this increase may
have been offset by the reduced expenditures of losers.
* One SAL recommended raising the domestic price of rice to interna-
tional levels. This would have had a positive effect on rice farmers
who are poorer than average but would not have affected other pov-
erty groups very much because little rice is consumed by the poor.
After analyzing patterns of income distribution from 1980 to 1985,
Kanbur found that real per capita income fell by 2.6 percent a year. This
dramatic decline understates the corresponding increases in poverty,
which rose by 4.8 percent a year, and in hard-core poverty, which rose
by 7.9 percent a year. He also examined shifts in sectoral output and
found that tradables gained, particularly in processing, while services,
construction, and utilities declined in value added. He concluded that
food crop farmers, already the poorest group, were hit hardest by the
relative drop in the price of domestic goods, while export crop farmers,
poorer than average, did best although their incomes also declined.
Among food croppers the incidence of absolute poverty increased from
42 percent to 50 percent during this period and hard-core poverty in-
creased from 13 percent to 20 percent.
Kanbur recommended that (1) food croppers receive compensatory
protection during the adjustment of prices in favor of traded goods, (2)
proposals to cut producer prices for exports in order to augment gov-
ernment revenue be reconsidered because export croppers are still very
poor, (3) housing subsidies be removed as proposed in the SALS since they
do not benefit the poor, and (4) the government direct expenditures on
housing, health, education, and other amenities to the poor.



266 EVALUATION OF ADJUSTMENT PROGRAMS
Appendix B. Peru: A Counterfactual Case
In April 1988 the Ghanaian undersecretary for finance and economic
planning remarked that "adjustment is necessary to mitigate the cost of
not adjusting." He concluded that adjustment, with all its welfare and
other costs, is inevitable sooner or later for all countries, and that if later
rather than sooner, the costs will have escalated substantially.
What are the social costs for a country that delays adjusting? The case
of Peru, a nonadjuster, is considered here.22 Peru was selected, first,
because it has persisted in delaying adjustment even though many coun-
tries with similar problems are undertaking adjustment programs and,
second, because the Peru Living Standards Survey (part of the World
Bank's Living Standards Measurement Study) has already yielded data.23
Peru's recent macroeconomic situation has shown worsening fiscal and
external imbalances. Its macroeconomic policies have been characterized
by the application of selective price controls, overvalued and multiple
exchange rates, restrictions on manufactured imports, agricultural sub-
sidies, subsidized credit, negative real interest rates, unilateral limitation
of debt-service payments, fiscal stimulation of consumption, and broad-
scale state economic intervention.24
Problematic sectoral policies in Peru are not unlike those in countries
with adjustment programs; Peru, however, is doing less to push the econ-
omy back on track. Sectoral policies with social welfare consequences in
Peru have included disincentives to the production of tradables, partic-
ularly the country's key exports (mining and petroleum products, which
accounted for 64 percent of exports in 1984); ill-advised public invest-
ments, such as massive irrigation projects with low returns that resulted
in a 16 percent decline in agricultural GDP per capita during 1970-84;
underinvestment in education, an inefficient education bureaucracy, and
poor educational services, which contributed to rising dropout and repeat
ratios, particularly at the primary level and in rural areas; and, until
recently, expenditures on public health weighted heavily toward urban,
curative, hospital-based health care at the expense of primary health care
(which in 1984 received only 5 percent of the recurrent budget), despite
the fact that in 1982 children less than five years old accounted for 47
percent of all deaths in Peru.
To assess how the poor have fared without adjustment in Peru, Glewwe
(1988b) first identified the Peruvian poor. At the time of his study they
resided primarily in rural areas: although 44 percent of all Peruvians
lived in rural areas, 70 percent of the poorest 30 percent of the population
were found there, and 83 percent of the poorest 10 percent. Land reform
in the 1970s eliminated most large-scale private agriculture but did not
improve the lot of most of the rural population. Most farmers, as well
as the urban poor, were worse off in the 1980s than in the 1960s (Webb



The Social Costs of Adjustment 267
forthcoming). Although this was true in most Latin American countries
as a result of the depression of the early 1980s (World Bank 1986),
regardless of whether adjustment was undertaken, extreme poverty and
inequality were particularly widespread in Peru. Of twelve countries ex-
amined in Latin America, Asia, and Africa during the 1960s and 1970s,
Glewwe found that Peru had the highest Gini coefficient (that is, the most
unequal per capita income distribution). Although it is difficult to es-
tablish a correlation between poverty and the absence of adjustment in
Peru, its deteriorating nutritional record may well reflect such a rela-
tionship.
Glewwe (1988c) hypothesized the effects on the poor if adjustment
policies were implemented. He estimated that adjustment measures
would include a currency devaluation, which would raise producer prices
for exports such as coffee that is grown by poor farmers and the price
of imported food such as rice; reductions in social service expenditures,
which would hit the poor's education and health care; and increased
charges for utilities, such as electricity and water services, public trans-
port, and petroleum products, which would be regressive (although the
poor have less access to utilities than the rich, they spend a much larger
proportion of their total budget on them, as is also the case with public
transport and petroleum products). Even if Peru were to undertake ad-
justment, many of the poor would suffer from such cutbacks in expend-
iture. Glewwe therefore recommended targeted programs to protect the
poor.
Notes
1. It may be too early to assess the long-term impact on growth since ad-
justment programs are still in their early stages.
2. Recession, economic mismanagement, lack of effective social policies, and
external shocks are some of the other influences that are responsible for dete-
riorating social welfare.
3. This new concern also extends to the IMF. A review of the implications for
poverty of IMF-supported adjustment programs argues that stabilization and ad-
justment have important short-run distributional implications, including negative
effects on certain vulnerable groups (Heller and others 1988). It recommends
that compensatory measures to protect these groups become part of IMF stabi-
lization packages.
4. The country studies prepared for the World Bank adjustment lending report
show that wages dropped significantly in some countries (World Bank 1988a).
5. Chile's six-decade-long tradition of social welfare interventions has been
the impetus for its antipoverty interventions during adjustment.
6. This controversy is reflected in the literature: for example, Huang and Ni-
cholas (1987), on the one hand, favor the new poor while Cornia, Jolly, Stewart
(1987), on the other, favor the old as well as the new poor.
7. Lack of data makes it difficult to illustrate the share of the adjustment
burden borne by females. The same applies to other vulnerable groups within



268 EVALUATION OF ADJUSTMENT PROGRAMS
the household, such as children and the elderly. Intrahousehold disaggregation
of data is needed to enable an analysis of adjustment's impact on the social
welfare of these groups. Appropriate policy responses could then be designed for
individuals, not just income groups.
8. The extreme or structural poor are also known in the literature as the
chronic and ultrapoor (Lipton 1988). In this chapter the term extreme poor is
generally used.
9. Lipton (1988) distinguishes between the poorest 10 percent of the popu-
lation in developing countries, who are a potential resource, and the same decile
in developed countries-the old, ill, handicapped, and drug and alcohol ad-
dicted-who are not potentially productive.
10. Lipton (1988) uses nutritional status to define the poor. He distinguishes
between the "ultrapoor"-those at significant risk of income-induced under-
nutrition-and the "poor"-those with sufficiently low income to be at risk of
hunger but not of undernutrition.
11. Before 1988, only the 1985 review of Ecuador's public investment dis-
aggregated social sector expenditures by income group (de Melo 1988). These
reviews generally cover a multitude of sectors, such as agriculture, industry, en-
ergy, transport, telecommunications, housing, water, education, and health.
12. Including the private sector, Brazil spent 25 percent of GNP on social ex-
penditures in 1986. Brazil still has strikingly low social welfare indicators, how-
ever.
13. Deaton (1987) attempted to use the results of the Living Standards Mea-
surement Study in C6te d'Ivoire to isolate gender differences. He found that the
allocation of adult goods is heavily biased toward adult males and that women,
old men, and particularly old women appear to have much less access to goods.
14. Social accounting matrices will be one of the methodologies employed for
some of the countries in the Social Dimensions of Adjustment project to track
income distribution changes before, during, and after adjustment in Sub-Saharan
countries.
15. This list was compiled by examining SAL documents and a systematic sum-
mary of conditionality by William Steel.
16. Tunisian food subsidies are not targeted specifically to poor groups. Al-
though they benefit all sections of the population including the poor, it is the
middle- and upper-income groups and the urban population that benefit most.
There is scope for specifically targeting the Tunisian poor if political and ad-
ministrative constraints can be surmounted.
17. The limited successful experiences with reducing food subsidies indicates
that geographical and commodity targeting, with their inevitable leakages, work
better than screening programs based on income or nutrition level. Many at-
tempts to reduce food subsidies have resulted in riots and political instability.
These aspects of targeting food subsidies are discussed in detail in Berg (1987)
and Pinstrup-Andersen (1988).
18. Bela Balassa has pointed out to me that there are tradeoffs in instituting
progressive taxation: it may have adverse effects on efficiency and may lead to
the emigration of professional, skilled, and technical workers.
19. In some countries it may be difficult to identify inferior but nutritionally
adequate food consumed by the poor and disdained by the rich. For example,
this was possible in Morocco, but it might be difficult in West Africa.
20. The standard of living of the poor will improve with growth only to the
extent that the benefits of growth trickle down to them. Even so, it is now rec-
ognized that growth and its trickle-down effects alone are insufficient to eliminate
poverty (World Bank 1988b).



The Social Costs of Adjustment 269
21. It has been argued that adjustment programs in several countries have not
been sustained at least in part because of the social costs (World Bank 1988a).
22. A recent IMF paper (Heller and others 1988) concluded that the counter-
factual case can be an elusive and possibly misleading yardstick for comparison.
It prefers a before-after approach for countries that have tried adjusting.
23. Data from the Peru Living Standards Survey became available in the fall
of 1986. It is uncertain when the survey will be repeated, but the intention is to
do so about every three years.
24. Although these policies contribute to macroeconomic imbalances, those
that assist the poor could be selectively maintained and redirected through careful
targeting, such as subsidized credit, within the context of overall trimming.
References
Addison, Tony, and Lionel Demery. 1988. "The Economics of Poverty Allevia-
tion under Adjustment." Paper prepared for the World Bank Symposium on
Poverty and Adjustment. Overseas Development Institute, London.
Berg, Alan. 1987. Malnutrition: What Can Be Done? Baltimore, Md.: Johns
Hopkins University Press.
Besley, Timothy, and S. M. Ravi Kanbur. 1987. "Food Subsidies and Poverty
Alleviation." All Souls College, Oxford University, and University of Warwick,
United Kingdom.
Blejer, Mario I., and Isabel Guerrero. 1988. "The Impact of Macroeconomic
Policies on Income Distribution: An Empirical Study." IMF Working Paper WP/
88/57. International Monetary Fund, Washington, D.C.
Bourguignon, Francois, William H. Branson, and Jaime de Melo. 1989. Ma-
croeconomic Adjustment and Income Distribution. OECD Development Center
Technical Paper 1. Paris: Organisation for Economic Co-operation and De-
velopment.
Carlson, B. A. 1988. Monitoring Human and Social Indicators in the Adjustment
Process. New York: UNICEF.
Cornia, G. A., Richard Jolly, and Frances Stewart, eds. 1987. Adjustment with
a Human Face. Oxford, U.K.: Clarendon Press for UNICEF.
Deaton, Angus. 1987. The Allocation of Goods within the Household: Adults,
Children, and Gender. LSMS Working Paper 39. Washington, D.C.: World
Bank.
de Melo, Jaime, and Sherman Robinson. 1982. "Trade Adjustment Policies and
Income Distribution in Three Archetype Developing Economies." World Bank
Reprint Series 271. Washington, D.C.: World Bank.
de Melo, Martha. 1988. "Public Investment/Expenditure Reviews: The Bank's
Experience." Public Economics Division, Country Economics Department,
World Bank, Washington, D.C.
Dwyer, Daisy Hilse, and Judith Bruce. 1988. A Home Divided: Women and
Income in the Third World. Stanford, Calif.: Stanford University Press.
Fischer, Stanley. 1986. "Issues in Medium-Term Macroeconomic Adjustment."
World Bank Researcb Observer 1 (2) July.



270 EVALUATION OF ADJUSTMENT PROGRAMS
Gillespie, Nancy. 1989. "Women in Development Issues for Economic and Sector
Analysis." Women in Development Division, World Bank, Washington, D.C.
Glewwe, Paul. 1988a. The Distribution of Welfare in C6te d'Ivoire in 1985. LSMS
Working Paper 29. Washington, D.C.: World Bank.
1988b. The Distribution of Welfare in Peru in 1985-86. LSMS Working
Paper 42. Washington, D.C.: World Bank.
. 1988c. The Poor in Latin America during Adjustment: A Case Study of
Peru. LSMS Working Paper 56. World Bank, Washington, D.C.
Glewwe, Paul, and Dennis de Tray. 1988. The Poor during Adjustment: A Case
Study of C6te d'Ivoire. LSMS Working Paper 47. Washington, D.C.: World
Bank.
Glewwe, Paul, and Jacques van der Gaag. 1988. Confronting Poverty in De-
veloping Countries: Definitions, Information, and Policies. LSMS Working
Paper 48. Washington, D.C.: World Bank.
Heller, Peter S., A. Lans Bovenberg, Thanos Catsambas, Ke-Young Chu, and
Parthasarathi Shome. 1988. The Implications of Fund-Supported Adjustment
Programs for Poverty: Experiences in Selected Countries. Washington, D.C.:
International Monetary Fund.
Huang, Yukon, and Peter Nicholas. 1987. Protecting the Poor during Periods
of Adjustment. Development Committee Pamphlet 13. Washington, D.C.:
World Bank.
Kanbur, S. M. Ravi. 1988a. "Poverty and the Social Dimensions of Structural
Adjustment in Cote d'Ivoire." World Bank, Washington, D.C.
. 1988b. "The Principles of Targeting." Paper prepared for the World
Bank Symposium on Poverty and Adjustment. Warwick, U.K.: Development
Economics Research Center, University of Warwick.
Lewis, John Prior, and Norman Thomas Uphoff, eds. 1988. Strengthening the
Poor: What Have We Learned? New Brunswick, N.J.: Transaction Books.
Lipton, Michael. 1988. The Poor and the Poorest: Some Interim Findings. World
Bank Discussion Paper 25. Washington, D.C.: World Bank.
Narayana, N. S. S., K. S. Parikh, and T. N. Srinivasan. 1987. "Policies and
Impacts (Pandits): Analysis with a General Equilibrium Model for India." In-
dira Gandhi Institute of Development Research, Bombay, India.
Pinstrup-Andersen, Per. 1988. Food Subsidies in Developing Countries: Costs,
Benefits, and Policy Options. Baltimore, Md.: Johns Hopkins University Press.
Pyatt, Graham, and Jeffery I. Round, eds. 1985. Social Accounting Matrices: A
Basis for Planning. Washington, D.C.: World Bank.
Srinivasan, T. N., and P. K. Bardhan, eds. 1988. Rural Poverty in South Asia.
New York: Columbia University Press.
Webb, Richard Charles. Forthcoming. "The Political Economy of Poverty, Equity
and Growth: Peru 1947-85."
World Bank. 1986. Poverty in Latin America: The Impact of Depression. Wash-
ington, D.C.: World Bank.
-    . 1987. World Development Report 1987. New York: Oxford University
Press.



The Social Costs of Adjustment 271
-1988a. Adjustment Lending: An Evaluation of Ten Years of Experience.
Policy and Research Series 1. Washington, D.C.
. 1988b. "Report of the Task Force on Poverty Alleviation." Report R88-
183. World Bank, Washington, D.C.
. 1988c. Targeted Programs for the Poor during Structural Adjustment:
A Summary of a Symposium on Poverty and Adjustment, April 1988. Wash-
ington, D.C.
. 1988d. World Development Report 1988. New York: Oxford University
Press.
Zuckerman, Elaine. 1989. "Compensatory Programs: Redressing Social Costs of
Adjustment." Country Economics Department, World Bank, Washington,
D.C.
Comments
Dennis de Tray
THE WELFARE CONSEQUENCES of adjustment programs have generated con-
siderable debate. Elaine Zuckerman adds to this discussion by setting
out the issues related to poverty and adjustment and by presenting some
of the findings of country studies that review adjustment programs and
their distributional impacts. Since these country studies are vignettes,
often of programs that are still developing, I will restrict my comments
to general issues of adjustment and the poor.
Breaking down the topics discussed in chapter 12 in the following three
overlapping but quite distinct sets of issues helps to focus and clarify the
discussion:
* Reduction of structural poverty-especially issues related to reducing
poverty over the long term
* Protection of the poor during adjustment-ensuring that those who
are least able to weather the economic fluctuations that may be part
of the adjustment process are somehow protected
* Sharing of the social burden-providing a "soft landing" for the non-
poor who are affected by adjustment policies.
Two other issues covered in this chapter would also benefit from
greater clarity of analysis: the questions of whether changes in the welfare
of the poor are the consequences of adjustment programs or the legacy



272 EVALUATION OF ADJUSTMENT PROGRAMS
of past failures to adjust, and regional differences in the nature of poverty,
especially as they relate to adjustment, in Sub-Saharan Africa and Latin
America.
Clarity in discussing these different aspects of poverty is especially
important in a policy context. First, different policies are needed to deal
with each of the three sets of issues listed above. Second, conclusions
about the social costs of adjustment may be quite different, depending
on how blame is distributed between initial conditions (the consequences
of past policies) and the adjustment policies themselves.
Reducing structural poverty-increasing the capacity of individuals,
especially the poor, to earn income-is a central mandate of the World
Bank. In this sense, a concern for the poor underlies many World Bank
lending operations, both policy-based loans and project loans. It is de-
batable, however, whether reducing structural poverty ought to figure
prominently in a discussion of adjustment policies. Improving a country's
ability to address structural poverty in the future through a healthier,
faster growing economy may be one of the main motivations for ad-
justment, but reducing structural poverty is seldom in itself a driving
force in getting countries to adjust in the first place.
Countries under adjustment are often short of resources and of the
capacity to implement programs. Asking them to add to the already large
demands on their economic systems the task of developing programs to
reduce poverty permanently may be very dangerous indeed. Pressure to
develop such programs may endanger the adjustment process and pro-
duce little sustained improvement in the welfare of the poor.
One useful way of deciding whether a particular poverty-reduction
issue belongs in a discussion of adjustment policies is to determine
whether the poverty is to be reduced through consumption transfers or
through programs to enhance productivity. Although temporary transfer
programs may be needed under adjustment, it is difficult to argue that,
in the context of adjustment lending, countries should focus more atten-
tion than usual on the longer-run issue of enhancing the earning capacity
of the poor. This is not to argue that these longer-run issues are unim-
portant. They are, as I said, at the top of the World Bank's list of concerns.
In the design of adjustment programs, however, the emphasis should be
on the immediate problems associated with adjustment and the poor and
not on longer-run issues associated with reducing structural poverty.
The second element of this three-part taxonomy-and the one that I
believe is in most people's minds when they discuss issues of adjustment
and welfare-is protecting the poor during adjustment. In thinking about
the consequences of adjustment policies for the poor, it is useful to dis-
tinguish the immediate effects from medium- and long-term effects.
The immediate effects depend mainly on what the poor have been doing
during the preadjustment period. What were they consuming? Were they



The Social Costs of Adjustment 273
working, and if so, at what? How much of their consumption was self-
produced and how much was purchased through the market? How much
of their consumption was subsidized and how much was purchased at
market prices? What use were they making of government-provided
goods and services? In the past, rhetoric has often outstripped fact in
assessing the immediate consequences of adjustment programs or policies
for the poor. As this chapter points out, however, in a small but growing
number of countries, household surveys are producing information that
provides a good starting point for determining when and where special
programs are needed to offset the harmful effects of adjustment on the
poor.
These first-round effects of adjustment policies on the poor differ sig-
nificantly by region. For prototypical African countries, one fact domi-
nates the assessment: the overwhelming majority of the poor in most
African countries live in rural areas. Rural households in Africa tend to
produce much of what they consume. Furthermore, most government
services and subsidies in Africa are consumed by or benefit an urban
middle class. Thus adjustment programs that cut subsidies, reduce gov-
ernment services, or increase user fees for water and electricity and the
like often will not directly affect the majority of the poor. By contrast,
in Latin America substantially larger portions of the poor live in urban
areas, although even there the negative consequences of adjustment for
the poor are often overstated.
Whether in Africa or Latin America, however, there can be no denying
that adjustment policies generally hit the urban poor the hardest. This
fact carries with it an important, optimistic message for the design of
poverty-related adjustment polices: there is some hope of being able to
deal with the problem since well-targeted programs to reach the urban
poor are administratively more feasible and much less costly than pro-
grams aimed more generally at the poor.
What of the medium- and long-term effects of adjustment programs
on the poor? These are much more contentious and problematic issues-
and indeed the subject of considerable ongoing research. Often at the
crux of the issue is the change in relative prices of tradables and non-
tradables and the extent to which the poor are involved in the con-
sumption and production of tradables and nontradables. In C6te d'lvoire,
for example, adjustment policies have increased rural incomes by in-
creasing the earnings of rural export crop producers. This change has
increased the relative income disparity between export crop producers
and the poor food crop producers. I believe, however, that eventually
the result will be a richer rural community that benefits, in absolute terms,
even the rural poor who do not produce cash crops.
The third category of welfare-related issues, sharing the social burden,
concerns those who are the victims of government efforts to reduce fiscal



274  EVALUATION OF ADJUSTMENT PROGRAMS
deficits by reducing the size of the public sector and aggregate demand.
Although the effects of these actions are certainly a function of the ad-
justment program, they are not a poverty problem in the conventional
sense. The experience of several countries has shown, however, that the
costs associated with this aspect of adjustment are often large enough to
bring adjustment programs to a halt.
What is required for this group is a "soft landing" in recognition of
the need for (and obligation of) society to distribute adjustment costs
broadly. The key characteristic of policies addressing this problem is that
they should ease the cost without stopping the process of adjustment.
These are short-run costs, and the programs to deal with them must be
temporary. If the programs are allowed to drag on, they will exacerbate
the very problems the adjustment program was designed to solve. Since
protests from this group of middle-income (and sometimes lower-middle-
income) individuals often topple adjustment programs, there can be no
disagreement that this is an important policy issue. But for real adjust-
ment to take place, those earning rents from a highly distorted economy
must eventually "lose."
The distinctions implied by this taxonomy of adjustment issues related
to the poor are important for focusing policies correctly. It is all too easy
to mislay the blame for the long-standing problems of the poor at the
feet of adjustment policies. In many countries with active adjustment
programs the subsidies and employment policies of the government were
in any case not sustainable. Without adjustment, the existing system
would have collapsed, through either runaway inflation or a complete
breakdown of the economic system. At issue here is the counterfactual.
Adjustment policies may be implemented during a period of declining
well-being for the poor, but it is quite another matter to argue that the
two are causally related. This debate is more than semantic. To judge
the distributional impacts of adjustment programs and, therefore, to as-
sign blame for declines in the well-being of different segments of society,
one must have some sense of the negative pull of long-standing but ul-
timately unsustainable policies.
As I have already indicated, it is important in discussing adjustment
and the poor to limit generalities, or at least to qualify them geograph-
ically. For most of the African continent, the issues of poverty are dom-
inated by medium- and long-term development issues. The question is
how best to increase the productivity and earning power of the poor-
through investment in human capital or through improved access to in-
come-generating assets? In Latin America the major issue concerns the
overall decline in public subsidies and the provision of public sector goods
and services and how to shift the composition of the remaining expen-
ditures toward more targeted, more efficient programs for the really poor.



The Social Costs of Adjustment 275
In thinking about adjustment and the poor it helps to divide the issues
along the lines described above; appropriate policies differ for each of
the subcategories, as do the effects on the poor and the links to adjust-
ment. Clear thinking about adjustment and the poor as distinct from
debates about broad poverty-related issues will, in the end, produce better
focused, more efficient, and more sustainable policies for dealing with
the poor during episodes of adjustment. Focusing on adjustment-related
poverty will also give a more realistic sense of the implications for both
the administration and the financing of programs. Many very competent
people are challenging the barriers to knowledge in this area, and I am
confident that we will soon be on much firmer factual ground than we
now are when we discuss the welfare effects of adjustment programs.






PART III
Country Experience with
Adjustment Lending in the 1980s






13 Introduction to Part III
PART III EXAMINES the experiences of nine countries in the four operational
regions of the World Bank: C6te d'Ivoire, Ghana, and Zambia (Africa);
Indonesia and the Republic of Korea (Asia); Pakistan and Turkey (Eu-
rope, Middle East, and North Africa); and Chile and Mexico (Latin
America). All these countries except Indonesia have received two or more
adjustment loans from the World Bank since 1986.
Tables 13-1 and 13-2 present selected economic indicators for eighty-
eight developing countries and for the nine countries whose experience
is reviewed in the following chapters. As a group, the countries selected
for detailed review shared the deterioration in economic performance
that the group of eighty-eight developing countries also experienced in
the 1980s. The turnaround by the mid-1980s, however, was stronger for
the nine countries than for the rest. The broad conclusion of the analysis
of comparative behavior presented in chapter 11 was that adjustment
lending was marginally successful overall. The positive outcome asso-
ciated with adjustment lending was even stronger when the focus was
narrowed to only those countries that were intensely involved in ad-
justment lending. The World Bank's 1988 report on adjustment lending
also suggested that the positive association between performance and
adjustment lending increases from mild to strong when the focus shifts
from all loan recipients to the intensive recipients.'
What are some of the findings or implications of the case studies as a
group? First, externally funded programs, whether by the World Bank
or the imF, are sometimes seen as inflexible and unresponsive to local
conditions. Lack of government commitment to reforms may spring from
inappropriate design. A government may grudgingly accept the program
in order to get access to the external resources, but there may not be true
commitment to the program. Nonetheless, the studies show that several
adjusting countries, including Ghana, Korea, and Turkey, have to varying
degrees been able to adapt and improve externally funded programs to
meet their specific needs.
This issue touches on the "ownership" of the program. Ownership
appears to require an enthusiasm for or at least a recognition of the need
279



Table 13-1. Selected Economic Indicators for Eighty-Eight Developing Countries, 1965-72 to 1987
(unweighted average percent)
Period average                                Recent experience
Indicator                                   1965-72  1973-77  1978-81  1982-88'  1982   1983   1984   1985   1986   1987   1988'
Growth rate (constant prices)
GDP (market prices)                          5.4        5.1       3.8        2.5       1.8    1.2    2.4    3.1    3.2    2.3    3.3
Exports (goods and nonfactor services)       7.4        5.2       6.4        3.4     - 1.8    3.5    6.0    1.4    4.0    5.3    5.5
Imports (goods and nonfactor services)       6.6        8.8       6.2        0.7     -3.7   -4.5      1.3    3.7    0.2    4.0    4.2
Share of GDP (current prices)
Exports (goods and nonfactor services)      21.9       24.9      25.6       24.8      24.4   24.7   26.0   25.7   24.5   24.3   24.2
Imports (goods and nonfactor services)      25.5       31.9      35.5       32.5      35.5   33.2   32.5   32.3   31.2   31.3   31.4
tj     Resource balanceb                           -3.6      - 7.0      - 9.9     - 7.7    -11.1   - 8.5   - 6.5  -6.6   -6.7   -7.0   - 7.2
Debt ratio'
External debt/exports (goods and services)  106.3     107.7     143.4      277.6     199.6  230.3  234.5  282.6  333.3  341.7  321.2
Debt service/exports (goods and services)   11.6       11.5      15.1       21.8      18.4   19.1   19.1   22.0   24.5   22.9   26.5
Prices
Consumer prices (percentage change)          5.9       21.4      19.5       28.7      20.9   28.1   29.3   32.9   24.9   26.5   38.5
Real exchange rated (1980 = 100)             n.a.      n.a.      99.8       97.6     107.4  106.1  105.7  104.5   91.2   85.2   83.4
Terms of trade' (1980 = 100)               130.4      120.4     102.1       92.1      89.3   92.0   93.3   90.1   95.5   89.2   95.1
n.a. Not available.
a. Preliminary estimates.
b. Resource balance is the difference between exports and imports of goods and nonfactor services.
c. Public and private, medium- and long-term debt outstanding and disbursed; data for 1965-72 refer to 1970-72 only.
d. Increase indicates real appreciation.
e. Ratio of export to import price indices of merchandise goods; increase indicates an improvement.
Source: World Bank and fMF data.



Table 13-2. Selected Economic Indicators for the Nine Case Study Countries, 1965-72 to 1988
(unweighted average percent)
Period average                                Recent experience
Indicator                                   1965-72  1973-77  1978-81  1982-88a   1982   1983   1984   198S   1986   1987   1988'
Growth rate (constant price)
GDP (market prices)                           6.0       4.6       4.9        3.2      -0.9     1.1    4.7    4.2    4.7    4.3    4.0
Export (goods and nonfactor services)         8.5       6.5       6.1        7.2       10.0    1.8    5.7    1.0   13.0   15.4    3.2
Import (goods and nonfactor services)         8.2       8.3       6.6        2.3    -15.1  -4.1       5.6    4.9    4.1   15.0    5.8
Share of GDP (current prices)
Export (goods and nonfactor services)        18.3      22.4      21.0       25.0      20.7   23.1   25.1   25.3   26.7   27.5   26.3
Import (goods and nonfactor services)        19.0      23.6      24.4       24.6      23.6   24.4   23.9   24.3   26.0   25.7   24.4
Resource balanceb                           -0.7      -1.2      -3.4         0.4      -2.9   -1.3      1.2    1.0    0.7    1.8.    1.9
Current account balance                     -2.3      - 3.9      -5.8      -3.8       -6.5   - 4.7   - 2.7   -4.1   - 3.7   -1.6   -3.0
t-i    Gross domestic investment                    18.8      21.0      21.7        18.4      19.0   18.2   17.7   18.5   18.1   18.7   18.9
o0
Debt ratio'
External debt/GDP                           26.8       31.2      33.6       62.9      42.7   51.S   55.8   66.7   77.6   80.7   65.4
External debt/exports (goods and services)  201.6     170.2     182.4      243.2     199.4  227.9  219.8  251.0  282.8  278.9  242.8
Debt service/exports (goods and services)    19.0      18.3      26.1       29.8      28.4   30.1   27.7   30.0   31.3   30.2   30.6
Interest/exports (goods and services)         5.8       6.6      10.1       14.8       15.1   15.0   15.6   15.2   15.4   13.4   13.5
Prices
Consumer prices (percentage change)           9.3      53.2      29.1       28.3      18.3   36.9   24.1   21.8   26.1   34.5   36.8
Real exchange rate" (1980 = 100)             n.a.       n.a.    104.0       81.6     118.9  100.6   86.4   80.2   63.1   58.9   63.0
Terms of trade0 (1980 = 100)                139.7     114.7     101.8       89.6      89.4   92.7   92.6   90.4   86.8   86.9   88.1
n.a. Not available.
a. Preliminary estimates.
b. Resource balance is the difference between exports and imports of goods and nonfactor services.
c. Public and private, medium- and long-term debt outstanding and disbursed; data for 1965-72 refer to 1970-72 only.
d. Increase indicates real appreciation.
e. Ratio of export to import price indices of merchandise goods; increase indicates an improvement.
Source: World Bank and IMF data.



282  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
for the adjustment program on the part of key members of the govern-
ment and the public at large. The programs in Ghana and Turkey were
ambitious, but in both countries economic deterioration had induced a
consensus on the need for change. Although opposition was not totally
avoided, in both countries the programs were clearly seen as the only
way out of economic chaos, a position that is particularly evident in
Chad Leechor's study on Ghana (chapter 15). The same consensus never
developed in the case of Zambia, which eventually reversed its reforms
(chapter 16).
Second, the adequacy of financing also influences the effectivenss of
adjustment policies since it affects society's ability to sustain programs
that may have significant short-term costs. Adequate financing can pro-
vide the time and resources needed for an orderly transition to adjust-
ment, although care must be taken to ensure that external financing does
not lead to indefinite postponement of needed reforms. Inadequate fi-
nancing, by contrast, can force a country to make demanding adjustments
at a pace that is not politically and economically feasible and so lead to
a program's collapse. Mohsen Fardi's chapter on Zambia shows quite
clearly that inadequate funding contributed significantly to the collapse
of the program in 1987.
Countries that have been willing to undertake comprehensive pro-
grams, such as Indonesia, Korea, and Turkey, have been able to attract
substantial external financing, including commercial financing. Ghana
provides a good example of the extent of external noncommercial fi-
nancing available to a country for a rigorous program of adjustment.
The World Bank, the imF, and bilateral donors have assisted Ghana heav-
ily and continue to do so.
Ghana's needs for external financing were relatively easy to address
because it had not acquired a large stock of debt and interest payments
were therefore not a problem for the adjustment program. For many
other countries, however, debt servicing takes up a substantial portion
of the foreign exchange earnings. The debt overhang emerged as a key
factor in the financing package for Mexico, as John Nash discusses in
chapter 22. This is an important issue because it affects the flexibility
with which a country can handle its exchange rate policy. (Devaluation
leads to larger real income losses because the stock of foreign debt is
revalued.) With hindsight, and in light of recent debt initiatives, it is
evident that timely debt relief would have helped avoid some of the up-
down, stop-go features that adjustment programs have exhibited.
Third, the better performers-Indonesia, Korea, and Turkey-fol-
lowed relatively sound policies over the long term and promptly adjusted
to shocks. Unlike many other oil-exporting countries, Indonesia managed



Introduction to Part III 283
its oil resources well and largely succeeded in avoiding the "Dutch dis-
ease," as chapter 17 by Sadiq Ahmed shows. Korea followed a long-run
program of macroeconomic stability and, as Mansoor Dailami explains
in chapter 18, adjusted very quickly to stabilize the economy when eco-
nomic conditions appeared to be getting out of hand in the late 1970s.
Turkey's early successes are now threatened by growing macroeconomic
instability, but Turkey has nevertheless stuck to a policy of economic
liberalization for nearly a decade. In chapter 20, both Faezeh Foroutan
and the commentary by Dani Rodrik underline the dangers to the entire
program posed by the lack of fiscal discipline in recent years.
Fourth, the speed of the supply response is closely associated with
sustainability. A strong export response, based on a substantial depre-
ciation of the real exchange rate, few rigidities in labor markets and
wages, and the presence of unutilized capacity helped Korea, Turkey,
and, more recently, Indonesia continue their reforms. A powerful lobby
in favor of reform emerged quickly. By contrast, Zambia reversed its
policy in part because of lags in export growth. As exports continued to
stagnate, foreign exchange became scarce while the exchange auctions
caused the exchange rate to rise rapidly, and thus uncertainty and in-
stability increased. The institutionalization of reforms also affects the
supply response by giving the program credibility in the eyes of investors.
Reasonable macroeconomic stability is also necessary for a good supply
response, but it is not always easy to know just how much stability is
needed for structural reforms to be effective. Korea's experience, as
shown in chapter 18, illustrates the complementarity between macro-
economic stabilization and structural reform.
A fifth issue, which is related to the credibility of reform, is that of
political and social stability. Reform, of course, involves change, but as
Shahid Husain points out in his summary remarks, the Latin American
countries that have pursued structural adjustment-Chile and Mexico-
bave had both a relatively stable political structure and an institutional
framework that supported reform. Even in these two countries, however,
the process has been slow and painful. Indeed, the reform process is likely
to prove much more time consuming and arduous than anticipated, and
the World Bank needs to gear its adjustment lending operations to a
longer time frame.
Sixth, the country studies often show that insufficient attention has
been paid to the social costs of adjustment.2 For example, real wages
often decline during adjustment as resources shift toward tradables and
away from nontradables. The dilemma is how to design policies that
bring this resource shift about without a sharp drop in consumption or
in real wages for the poor. Is the targeting of subsidies a desirable ap-



284 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
proach, and if so, how can the targeting best be done? In the case of
Turkey, for example, the potential impact of adjustment on the poor was
barely addressed in the program design even though it was clear that real
wages would have to decline to effect adjustment. Real wages fell sharply
in Chile too, as Cristian Moran shows in chapter 21. Although attempts
were made to direct expenditures toward the poorer groups and children,
Moran argues that these measures were insufficient. In Ghana the issue
was how to protect the urban poor rather than the poor in general, since
farmers, poor and rich, on the whole benefited from the reform. Ghana's
response was to set up the short-term compensatory Program of Actions
to Mitigate the Social Costs of Adjustment. But except in these two cases,
efforts to address the social costs of adjustment have been limited.
The countries studied in the chapters that follow have been categorized
according to the relative influence of adjustment lending on performance
(table 13-3). (No effort has been made to separate the effects of World
Bank adjustment lending and of IMF programs.) Although this appraisal
is unavoidably subjective, it is based on the individual country studies
and the multicountry studies reported in this volume.
A low impact from adjustment lending can be found in both the suc-
cessful and the not so successful cases of adjustment. For example, Korea,
a successful adjuster, did not seem to be significantly affected by World
Bank adjustment lending; having decided on its own to adjust, Korea
would probably have done much the same thing without the adjustment
loans. In Zambia, an unsuccessful adjuster, adjustment lending also did
not affect performance. In contrast, adjustment lending seems to have
had a major positive impact on performance in five countries. Chile and
Indonesia benefited from the external resources provided. Adjustment
lending had a favorable impact on stabilization and exchange rate policies
in Indonesia (in the absence of a formal IMF standby) and on exchange
rate and structural reforms in Turkey and Ghana. In Mexico the positive
effect has been on recent trade policies. The positive impact was more
modest, although it varied considerably, in the remainder of the nine
countries. The strength of the positive impact is generally sensitive to the
period examined; thus, for example, positive results were greater during
an early period in Turkey and a later one in Mexico.
The country studies highlight the diversity of experience with adjust-
ment lending. They do not lead to any simple conclusions on the effect
of policy on performance, and they indicate both the strengths and
weaknesses of adjustment lending as a means of supporting reform.
Overall, the country experiences show that when a determined effort
is made, when the technical groundwork has been carefully prepared,
and when the political will and broad general consensus exist or can



Table 13-3. Terms of Trade Shocks, Performance, and Effects of Adjustment Lending in the Nine Case Study Countries
Influence
Indicator                                            Low                               Medium                                     High
Terms of trade'                               Ghana                               Chile, Korea,                        C6te d'lvoire, Indonesia,
Zambia                              Mexico, Pakistan, Turkey
Improvement in GDP growthb                    C6te d'lvoire,                      Chile                                Ghana, Indonesia, Korea,
Mexico, Zambia                                                          Pakistan, Turkey
Reduction in trade balance'                   Ghana, Pakistan                     Turkey, Zambia                       Chile, C6te d'lvoire,
(exports minus imports of                                                                                             Indonesia, Korea, Mexico
,      goods and nonfactor
,,     services as percentage of
GDP)
Incremental positive effect of                Korea, Zambia                       C6te d'lvoire,                       Chile, Ghana, Indonesia,
adjustment lending on                                                             Pakistan                             Mexico, Turkey
policy or financinge
a. High means relatively adverse terms of trade shocks.
b. Rating is based both on an improvement during three years after adjustment lending compared with three years before and on the level of GDP growth in 1982-86.
Low indicates a decline in the growth rate and growth of less than 3 percent. Medium indicates a decline in the growth rate but growth of more than 3 percent, or growth
less than 1 percent despite an increase in the rate. High indicates an increase in the growth rate and growth of more than 1 percent.
c. Rating is based both on the reduction in the trade balance (three years after adjustment lending compared with three years before) and on the level of the deficit in
1982-86. Low indicates an increase in the deficit or a deficit of more than 10 percent despite a reduction. Medium indicates a reduction in the deficit by up to 10 percentage
points or a deficit between 1 and 10 percent. High indicates a reduction in the deficit by more than 10 percentage points and a deficit of less than 1 percent.
d. Low indicates little effect of adjustment lending itself or an unsuccessful effort.
Source: World Bank (1988) and the studies presented in this volume.



286 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
be developed, adjustment programs can work-to the lasting benefit
of the country.
Notes
1. World Bank, Adjustment Lending: An Evaluation of Ten Years of Expe-
rience, Policy and Research Series 1 (Washington, D.C., 1988).
2. This finding confirms the results of a recent study by G. A. Comia, Richard
Jolly, and Frances Stewart, Adjustment with a Human Face (Oxford, U.K.: Clar-
endon Press for UNICEF, 1987).



14 C6te d'Ivoire: The Failure of
Structural Adjustment
Christophe Chamley
FROM 1965 TO 1975 economic performance in Cote d'Ivoire was im-
pressive, with real growth over the period averaging 8 percent a year and
never falling below 4 percent. This smooth pattern was disrupted by the
cocoa and coffee boom in 1976, when prices rose more than threefold
as a result of a frost in Brazil. The windfall revenue from the boom was
captured by the government. Producer prices remained stable while the
stabilization fund for coffee and cocoa accumulated a surplus equal to
16 percent of GDP in 1977 at the peak of the boom. These resources were
used by the government to launch a vast investment program.
Although there was no indication that a permanent change had oc-
curred in world coffee and cocoa markets that would prolong the price
increases, the government continued to pursue its expenditure programs
as though such a change had taken place, even after prices began to fall.
Between 1977 and 1980 prices fell by 30 percent. Current expenditure,
particularly civil service salaries, increased as a share of GDP. Investment
was accelerated at a rate that strained the absorptive capacity of the
economy, and the marginal profitability of new investments was low.
Two legacies of the commodity boom have been particularly detri-
mental to the Ivoirian economy. The first was the fiscal deficit, an out-
come, in part, of the belief that the "fair" level of coffee and cocoa prices
was above the preboom level. This belief seems to persist even today and
makes implementation of reforms more difficult. The second was the
increase in domestic prices that began during the boom and was accom-
panied by an equivalent expansion of the money supply. There are two
standard explanations for the inflation in C6te d'lvoire between 1975
and 1980. One emphasizes the impact of government spending on do-
mestic demand (similar to the "Dutch disease"), and the second stresses
the influence of monetary expansion. The first explanation may be closer
to the truth because C6te d'lvoire, as a member of the West African
Monetary Union, has a fixed exchange rate with free capital movements.
As a result C6te d'lvoire does not have independent control over its
monetary policy. Whatever the cause, the higher price level induced an
appreciation of the real exchange rate. The macroeconomic policy pur-
287



288 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
sued by Cote d'lvoire during the 1980s did not reverse the increase in
the price level.
For three years after the commodity boom was over, no significant
adjustment took place. Thus by 1981 the budget deficit equaled 12 per-
cent of GDP and the current account deficit equaled 17 percent of GDP,
while the ratio of foreign debt to GDP had reached 35 percent. It became
clear that the path being followed by the economy was unsustainable.
The Adjustment Program
The Ivoirian authorities launched a structural adjustment program at the
end of 1981, which was supported by the World Bank with three struc-
tural adjustment loans (SALS) (1981, 1983, and 1986) and by the Inter-
national Monetary Fund (IMF) through a multiyear arrangement under
the Extended Fund Facility (1981-83) and two one-year standby ar-
rangements (1984 and 1985).' The goals of the program were (1) to
restore internal equilibrium by reducing public expenditures and insti-
tuting administrative reforms for the control of the budgetary process
and (2) to restore external equilibrium in the short term by reducing the
fiscal gap and in the medium to long term by introducing economic
incentives for the diversification of the economy away from its main cash
crops. It is useful to divide the period into three phases.
1. The persistent fiscal gap (SAL 1, 1981-83). At this stage, the adjust-
ment program was not very well defined. Lending seems mainly to have
been a mechanism for policy dialogue. Since a "persistent oil boom" was
anticipated because of the recent discovery of oil, the seriousness of the
fiscal situation was greatly underestimated. Consequently, no significant
reform was implemented. Minor improvements in the budget process
had only a small positive impact on the fiscal balance, and these gains
were offset by a drought that reduced agricultural output and by the rise
in world interest rates. By 1983 the budget and external deficits were
about the same as in 1980, before the start of the adjustment program.
2. Partial restoration of equilibrium (SAL 11, 1984-86). During this sec-
ond phase, a reduction in government investment expenditure from 15
percent to 7.6 percent of GDP had an equal impact on the deficit. These
reductions were accompanied by a positive external shock-the end of
the drought and a temporary increase in coffee and cocoa prices. The
two effects contributed in roughly equal measure to restoration of the
short-run macroeconomic balances. The situation was transitory, how-
ever, because the prices of coffee and cocoa were well above their trend;
they fell again at the end of 1986.
3. The economic crisis (SAL III, 1986-88). The level of coffee and cocoa
prices hit a historic low. This led to macroeconomic imbalances similar
to those observed at the beginning of the adjustment program, but with



CMte d'Ivoire: The Failure of Structural Adjustment 289
a much larger stock of outstanding debt. Interest payments were sus-
pended in the spring of 1987, and since then the foreign debt to GDP ratio
has been about 90 percent.
The First Phase: 1981-83
PROGRAM AND CONDITIONS. The adjustment process began in the fall
of 1980 and led to the structural adjustment program supported by the
Extended Fund Facility of the IMF and the first World Bank SAL. From
the beginning the IMF took the lead in establishing and monitoring the
broad macroeconomic targets, and the World Bank, while supporting
these targets, concentrated on microeconomic issues such as the budget
process, restructuring of public enterprises, choice of investment projects,
and incentives for the diversification of the economy. (Table 14-1 presents
loan conditions for each SAL.)
The program established a set of policy actions that included ad-
ministrative reforms, a large number of studies, and specific economic
policy measures. Very few of these policy actions were specified as
conditions for disbursement. In general, the first SAL was a "soft" pro-
gram that set the basis for continuing policy dialogue. All parties
involved were conscious that the agreement was to be the first in a
series.
The structural adjustment program recognized the main imbalances
and identified the policies needed to address them. The seriousness of
the situation was grossly underestimated, however, because the postboom
prices of cocoa and coffee were considered to be temporary, and the
recent discovery of oil in C6te d'lvoire raised expectations of a new source
of revenue. Thus, for example, the structural adjustment program was
expected to introduce a pricing system (which remained unspecified) for
the agricultural sector that would adjust the producer prices of officially
supported crops sufficiently to offset the possible inflationary effect of a
prolonged period of high oil revenue while maintaining the country's
competitiveness in exports and import substitutes. What was expected
was a prolonged oil boom!
FISCAL POLICY. Government capital expenditure as a share of GDP was
reduced from 18 percent in 1980 to about 15 percent in 1981 and 1982,
but the level of current expenditure remained largely unchanged (table
14-2). (The reduction in current expenditure in 1981 was offset by an
equally significant increase in 1982; this variation may have been due to
payment delays.) On the revenue side, the receipts from cocoa and coffee
were relatively small because of low international prices and drought-
reduced output. The net surplus of the stabilization fund was between 2
and 3 percent of GDP (compared with 16 percent in 1977).



Table 14-1. Main Conditions of the Three World Bank Structural Adjustment Loans to C6te d'Ivoire
Category                               SAL I                                       SAL H                                        SAL III
Public sector
Administration         Improvement of budgetary accounting,        Continuation of improvements in ad-
the centralization of public finance,       ministrative efficiency and the unifi-
and the general efficiency of the pub-      cation of the budget process initiated
lic sector                                  in SAL I through (1) extension of the
Draft of a three-year plan (schema-di-        schema-directeur to the sectoral level,
recteur) to be translated into a three-     (2) a quarterly system of reporting
year rolling public investment pro-         for important investments, (3) stand-
gram (loi-programme)                        ardization and generalization of the
Integration of oil surpluses into the         economic appraisal process, (4) rein-
budget                                      forcement of programming and eval-
uation units in technical ministries,
and (5) integration of public enter-
prises (resource transfers and invest-
ment budget), resources of the mar-
keting board, and oil revenue in the
budget process
Expenditure            Ceiling of 5 percent in the growth of       As in SAL 1, overall macro-targets were
the level of nominal public expendi-        set and monitored in the framework
tures in 1981                               of the Extended Fund Facility with
Specific first-year performance condi-        the IMF. (Adoption of a financial pro-
tions for new foreign borrowing,            gram for 1984 acceptable to the
growth in domestic assets of the            World Bank as a condition for the
banking system, interest rates              second tranche.)
Satisfactory reviews of investment and
financial decisions, and the approval
of the 1982 investment budget with
its associated loi-programme



Revenue                No action on tax reform or on the rev-      No evaluation of the tax system be-
enue side of the budget                     cause it was thoroughly reviewed by
the IMF
Public enterprises                                                 Introduction of synoptic tables in the      Extended actions already taken under
transport and telecommunications            SAL It: (1) reduction in real transfers
sectors to determine performance tar-       to public enterprises, (2) disaggrega-
gets for basic financial data and key       tion of the system of synoptic table
indicators of management efficiency         for the thirty-one enterprises, (3) re-
and enterprise productivity                 habilitation of five enterprises, audits
Extension of this system to the thirty-       of three, improvement of supervisory
one most important public enter-            procedures of four
prises
Program of management audits (con-
ducted with some delays)
s     Agriculture            Restructuring of the Ministry of Agri-      Closing of the two least efficient (of       Diversification of export crops
culture                                     six) government sugar complexes           Phasing out of export taxes and appli-
Review of the budget of regional devel-     Some reduction in personnel of the ag-        cation of the value added tax for sig-
opment agencies                             ricultural agencies                         nificant agricultural products (palm
Financial restructuring of the regional     World Bank review of some of the               oil, sugar, cotton, rubber, coconut,
development agencies (with a 1982           agencies                                    coprah, and fresh pineapple)
plan as a condition for a second          Increase in producer prices of agricul-     Extension of the export subsidy scheme
tranche)                                    tural products, particularly nontradi-      to agricultural exports (begun in the
Increase in some producer prices for          tional exports. The price of some            middle of 1987 and only partially
nontraditional agricultural exports         products (rubber) was based on              implemented)
Removal of some subsidies and phasing         world prices.                             Further rehabilitation of the sugar in-
out of some inefficient projects          Support for a special pruning program         dustry through a comprehensive pro-
(sugar)                                     for coffee'                                 gram
Elimination of subsidies on food crop
seeds
Financial restructuring of national agri-
cultural bank
(Table continues on the following page.)



Table 14-1 (continued)
Category                                SAL I                                        SAL II                                       SAL III
Industry               Commission of studies; no specific ac-       Uniform tariff levels toward 40 percent      Removal of quantitative restrictions
tion                                         (on 60 percent of industrial value        Partial application of the export pre-
added for fertilizers and agroindus-         mium for non-cFA countries
tries, with the exception of coffee
and cocoa processing, chemicals,
plastics, and engineering industries)
Removal of quantitative restrictions
(replaced by an important surcharge
and five-year phaseout of a list repre-
senting 40 percent of local value
added)
Introduction of an export subsidy
scheme at the rate of 20 percent on
the domestic value added (which suf-
fered considerable delays)b
Acceleration of value added tax refunds
for exports
Proposal for a preferential interest rate
for export activities, which was re-
jected by West Africa Monetary
Union



Revision of the investment code, reform
of agencies in charge of promoting
small and medium-size enterprises,
creation of an investment promotion
agency'
Housing                                                            Sale of some subsidized units,d in-
creases in rents, introduction of man-
agement audits
a. As in SAL 1, the World Bank was aware that the prospects for coffee were better than for cocoa and that the Ivoirian policy of equal prices for the two crops should
be replaced by a price differential favoring coffee. The policy included in SAL ii, however, fell short of this action.
b. Although 1984 implementation was clearly intended, explicit conditionality required only submission to the national assembly. The review for the second tranche
states that the subsidy should be in place by the end of 1985. First payments were made only in August 1986, however, and the program was still not completely implemented
in 1988.
c. The effects of these reforms were either dubious or difficult to evaluate.
rS      d. The public housing program catered essentially to civil servants with middle to high income and was heavily subsidized (to the tune of more than 1 percent of GDP
t     by 1982). Following the actions under SAL II, urban public expenditure dropped by 45 percent between 1982 and 1985.
Source: World Bank.



294  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 14-2. ConsolidatedPublicSectorFinancein C6ted'Ivoire, 1980-87
(percentage share of GDP)
Budget item          1980  1981  1982   1983   1984  1985  1986  1987'
Total revenue         28.5  27.8   27.7   29.2  34.0  36.6   31.7  28.0
Tax revenue         20.4  22.1   21.3   20.8  20.5  19.8  20.3  23.1
Coffee and cocoa
stabilization fund  3.9   1.3    2.7    3.7   9.2  10.0   4.3  - 1.7
Public enterprise
surplus            0.7   0.2    0.5    0.7   0.3  2.6   2.6   2.9
Social security      1.4   1.2    1.3    1.3   1.0  1.1    1.3   1.2
Other revenue        2.2   2.9    1.8    2.7   3.0  3.1    3.2   2.5
Total expenditure     40.7  35.6   43.8   40.9  35.7  34.6   34.0  35.5
Current               22.6  20.6   28.6   29.6  28.1  27.4  27.7  29.2
Treasury            15.1  13.4   18.2   18.0  16.5  16.1   17.7  18.7
Wages              8.3   9.4    9.5   10.0   8.8  8.0   8.3   9.3
Materials          4.1   4.1    3.7    3.0   2.5  2.4   2.7   2.9
Subsidies          n.a.   n.a.    3.7    3.6   3.5  3.1   3.1   3.8
Unclassified       n.a.   n.a.    1.2    1.4   1.7  2.6   3.5   2.7
Interest on debt     3.2   4.8    7.2    8.2   8.5  8.4   7.2   7.4
Domestic           n.a.   n.a.    1.0    1.4   1.3  1.1    0.7   1.0
External           n.a.   n.a.    6.1    6.9   7.2  7.4   6.5   6.4
Other                4.3   2.4    3.2    3.3   3.1  2.9   2.8   3.1
Capital               18.1  14.9   15.2   11.3   7.6  7.2   6.3   6.3
Central government    n.a.   n.a.    9.9    6.9   4.1  4.6    3.0   3.7
Public enterprises   n.a.   n.a.    5.4    4.4   3.5  2.5    3.3   2.6
Surplus (pay basis)    -12.2  -7.8  -16.1  -11.7  -1.7   2.0  -2.4  -7.6
n.a. Not available.
a. Preliminary data.
Source: Ivoirian authorities and World Bank staff estimates.
The increase in international interest rates imposed an additional bur-
den on the budget. Interest rates on new commitments reached a peak
of 12 to 13 percent in 1981-82. As a consequence, interest payments on
the debt increased from 3.2 percent of GDP in 1980 to 7.2 percent in
1982.
The 3 percentage point reduction in capital expenditure from 1980 to
the levels in 1981 and 1982 was insufficient to reduce the deficit, given
the low revenues from cocoa and coffee and the increase in international
interest rates. A further reduction in capital expenditure (by 4 percent
of GDP) was implemented in 1983, but overoptimistic projections of rev-
enue delayed the implementation of sufficiently austere fiscal measures.
Thus by the end of 1983 the government deficit constituted about the
same share of GDP (12 percent) as it had before the adjustment program.
Ivoirian compliance with the conditions of SAL I was judged to be good,
mainly because the conditions applied to broad and soft overall targets



C6te d'Ivoire: The Failure of Structural Adjustment 295
for the budget of 1982 and the assessment was made in the middle of
1982. At this very early stage of the first adjustment loan period the
failure of the economy to adjust was not yet apparent. The pressure of
the fiscal gap became obvious later with the large expansion of credit by
the Central Bank. The expansion of credit overtook the targets set with
the IMF, and the last tranche of the Extended Fund Facility was. not re-
leased.
The Second Phase: 1984-86
During the period of SAL II three positive shocks affected the economy of
C6te d'lvoire: the drought ended, the value of the dollar increased by 24
percent in 1984, and the international prices of cocoa and coffee re-
covered. The CFA prices of cocoa and coffee increased by about 75 percent
between 1982 and 1984. Government expenditures were cut simultane-
ously. Internal and external balance was restored, but the recession that
had been initiated by the first expenditure cuts in 1983 deepened. Real
GDP fell by 2.5 percent in 1983 and by 2 percent in 1984. A brief economic
recovery in 1985 ended with the fall of cocoa and coffee prices at the
end of 1986.
PROGRAM AND CONDITIONS. The second SAL was intended to strengthen
the structural adjustment process that SAL I had begun. This follow-up
had been anticipated from the beginning, but by the time of SAL Ii the
situation in C6te d'Ivoire had worsened. Interest rates on foreign debt
were higher than in 1982-83, coffee and cocoa prices were low, and
cocoa output fell 30 percent from the 1981 level because of drought.
Furthermore, the expected oil boom did not materialize.
Although SAL II addressed the same issues as SAL 1, the urgency of un-
dertaking fundamental reforms was now very apparent, and the program
of reforms was considerably more comprehensive and detailed. In ad-
dition to the four main areas covered by SAL 1, SAL II provided for reforms
in public housing, particularly the subsidization of housing for public
sector employees. The restructuring and management of public enter-
prises also received special emphasis. Always important in the Ivoirian
economy, in 1977 public enterprises employed about one-third of the
formal labor force, produced 27 percent of value added, and were re-
sponsible for 68 percent of capital formation. But they lacked a clear
definition of their roles, their supervisory mechanisms were inadequate,
and their mediocre economic and financial performance strained the gov-
ernment budget. By 1985, however, the reorganization of the public en-
terprise sector had led to efficiency gains that were reflected in new sur-
pluses in operating funds.



296 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Tariff reform was another important element of the adjustment pro-
gram, particularly the substitution of tariffs for quantitative restrictions.
Under SAL 11 the tariff rate was brought to a uniform value of 40 percent.
Although the need for the tariff reform had been identified at the begin-
ning of the adjustment program and in the studies undertaken in SAL I,
no practical reform measures had been taken before SAL 11.
All of these policy measures contributed significantly to the restoration
of internal and external balance in the short term. The program had one
important shortcoming, however: it did not succeed in introducing a set
of incentives to diversify the economy.
FISCAL POLICY. The transition from the first to the second phase was
marked by further reductions in government capital expenditure in 1983
and 1984 to 7 percent of GDP, a level that has been maintained since that
time. Current expenditure was also reduced slightly by an amount equal
to about 1.5 percent of GDP. Thus the total reduction in expenditures was
about 5 to 6 percent of GDP (table 14-2).
The recovery of cocoa and coffee prices and the end of the drought
had a strong impact on the revenue of the stabilization fund. Its surplus
as a percentage of GDP was 5.5 percentage points higher in 1984 than in
1983. Standard taxes as a share of GDP remained unchanged at about 20
percent, their level since 1975, so taxes made no additional contribution
to revenue. The fiscal gap, which had been equal to 12 percent of GDP
at the end of the first phase in 1983, was thus met in roughly equal parts
by the reduction in expenditure and the increase in cocoa and coffee
revenues.2
By 1985 the numerous measures for reorganizing public enterprises
began to bear fruit, and the operating surplus of public enterprises as a
proportion of GDP increased by 2 percentage points. That same year the
budget showed a surplus for the first time (equivalent to 2 percent of
GDP). In 1986, however, the budget was again in deficit (2.4 percent of
GDP) because of the fall in world cocoa prices at the end of the year.
THE ECONOMY: RECESSION AND PARTIAL RECOVERY. In the first phase
of the structural adjustment program the level of real output stagnated
and output per capita fell. This recession was deepened by the drought
in 1983 and by the large reduction in government investment expenditure
in 1984. The change in real output per capita from 1980 to 1987 is
shown in figure 14-1. To exclude the contribution of the coffee and cocoa
sector, the value of output net of exports is also shown. (This is a rough
measure since there are induced effects between this sector and the others.
Exports other than coffee and cocoa were small and stagnant as a share
of GDP.)



C6te d'lvoire: The Failure of Structural Adjustment 297
Figure 14-1. Change in Real Output Per Capita in C6te d'Ivoire,
1980-87
Percent
25 -
20 -.*...~
10 - ~ ~   ~     ~     ~     ---         
15
10
Ou   ~~I    *          I    .    I
1980    1981   1982   1983   1984    1985   1986   1987a
- - - Real GDP per capita
.R..... Real GDP net of exports
-      Real GDP net of exports per capita
a. Estimated.
Source: World Bank data.
As real GDP increased by S percent in 1985 and then by 3 percent in
1986, a brief economic recovery occurred even though real government
expenditure remained stable.3 The recovery was very weak, however. An
increase in indirect taxes accounted for 2 percentage points of the growth
of measured real GDP. This contribution is only a statistical artifact and
does not represent any increase in the production of goods and services.
Thus the increase in production over the two years 1985 and 1986 was
only about 6 percent, which was barely sufficient to keep pace with the
population growth. In 1986 (for which there are detailed accounts), real
growth was led by expenditures of public entities while real output de-
creased by 7 percent in the textile and leather sector and by 9 percent
in construction.



298 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
The Third Phase: 1987-88
SAL III began in February 1985. The drought had finally ended, and the
economy was entering its 1985-86 recovery phase. Internal and external
economic balance had been restored, and the decline in output had been
arrested. Although the structural problems of the economy had not been
solved, the lack of pressure in the short term generated complacency and
contributed to the two-year lag between the initiation of SAL III and its
effect on the economy.
The fall in cocoa and coffee prices at the end of 1986 marked the
beginning of a new crisis, which led to the conclusion of the negotiating
process. Coffee and cocoa prices continued to fall and reached their low-
est level in recent history. In 1987, for the first time, the price stabilization
fund registered a deficit (of 1.7 percent of GDP).
PROGRAM AND CONDITIONS. Despite the progress in the first five years
of the adjustment program, the fundamental issues remained the same
under SAL III. The general purpose of SAL Ill was to extend and broaden
the changes begun in the previous years. The government's fiscal policy
program included detailed administrative reforms and changes in the
composition of expenditures, which resembled the fiscal program under
SAL II. Since reforms in the housing sector had been satisfactory, they were
dropped from SAL III and replaced by policy reforms for the energy sector.
The major reform under SAL III was the establishment of an export
premium program to offset the overvaluation of the real exchange rate,
since devaluation was not an option for C6te d'lvoire as a member of
the CFA community. The misalignment of the exchange rate had been
discussed since the beginning of the structural adjustment program in
1981, but no action had been taken to resolve the problem. The premium,
which was to be financed through an import tariff, was to be applied at
a rate of 20 percent to nontraditional exports directed to non-cFA coun-
tries. The policy was, in effect, a simulated exchange rate adjustment.
The premium was first extended to nontraditional agricultural exports
in the fall of 1986. Its implementation was limited, however, because it
excluded cotton and pineapple. Pineapple was excluded under the pretext
of a temporary lack of quality control in the industry, but since pineapple
represented almost half the agricultural exports that were eligible for the
premium, it is likely that the exclusion was based on the added cost.
Even for eligible goods, some payments were delayed because of the
budget crisis. The premium was extended in the fall of 1987 to non-
agricultural exports. Some payment delays occurred again in 1988.
No specific price policy for coffee and cocoa was set under the SAL
program. The World Bank and the Ivoirian authorities could not agree
on specific measures. There was only "a plan to arrest the projected



C6te d'Ivoire: The Failure of Structural Adjustment 299
decline in coffee output," which was included as a condition for the
second tranche. A shift in the incentive structure from cocoa to coffee
had been identified as a goal under SAL I, but even this vague condition
was abandoned as the prices of coffee and cocoa dropped dramatically
at the end of 1986.
FIsCAL POLICY. Since public sector capital expenditure had been re-
duced from 18 percent of GDP in 1980 to a little over 6 percent in 1986,
its level was no longer a problem for C6te d'lvoire, so no expenditure-
cutting targets were set. The program merely set standard goals for the
project selection process, without specifying particular policy measures.
Current expenditure was treated in the same way.
The fall in cocoa and coffee prices at the end of 1986 generated a
revenue shortfall of about 6 percent with respect to the "standard" ex-
pected surplus of 4 percent of GDP. Since current and capital expenditures
had not increased, the deficit for 1987, estimated at 7.6 percent of GDP,
was due entirely to this shortfall. The situation worsened as a result of
a further fall in commodity prices at the beginning of 1988, and in 1989
the prospects for a price recovery in the near future were still not good.
Interest payments on foreign debt were suspended in April 1987. Since
these payments represent a large fraction of GDP (see table 14-2), their
suspension provided needed breathing space for the Ivoirian authorities,
at least in the short term.
Macroeconomic Issues
Most of the conditions in the SALs addressed issues of microeconomic
management and did not emphasize an economywide approach.to struc-
tural adjustment. The World Bank was restricted in this respect by at
least three factors. First, the IMF has responsibility for determining and
monitoring the overall level of revenue, spending, and credit. Second, the
nominal exchange rate is a rigid institutional constraint; any change in
the nominal rate is ruled out because it would raise a set of complex
issues with the other countries of the CFA zone. Third, Ivoirian authorities
resisted any suggestions for changes in some key prices (coffee and cocoa,
civil service wages). Despite these limitations, the serious macroeconomic
problems facing the authorities in Cote d'Ivoire must be considered in
any discussion of structural adjustment.
Fiscal Policy
REVENUE. The structure of public sector revenue and expenditure is an
essential element of the economic crisis in C6te d'lvoire. The two main
sources of revenue are taxes on income, trade, and value added, and



300 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
taxes on coffee and cocoa (the stabilization fund). The share of revenue
from these two sources is presented in table 14-2.
Revenue from taxes on income, trade, and value added has remained
remarkably stable at between 20 and 23 percent of GDP since 1975. It is
unlikely that this ratio can be increased significantly, even with consid-
erable effort, because an increase in the tax burden would induce a shift
of activities toward the informal sector while an increase in tariffs would
lead to an increase in smuggling.
The second component of revenue is highly variable and arises from
the margin between the world prices of coffee and cocoa (net of pro-
cessing costs) and the producer prices. This component of revenue rep-
resented 16 percent of GDP at the peak of the boom in 1976, decreased
to less than 3 percent during 1981-82, increased again to more than 10
percent in 1985, began to fall after 1986, and became negative in 1987.
Since the beginning of the commodity boom in 1976, the prices of
coffee and cocoa have been adjusted to maintain parity between them
and to keep their real value with respect to the consumer price index at
a constant level. The stated policy (still in place) has been to provide a
"'stable price" to producers and to insulate them from the fluctuations
in international prices. The government has thus captured the windfall
gains of the commodity booms. The producer prices of coffee and cocoa
have been increased only gradually, at about the same pace as the do-
mestic price level. They have never been decreased, and any such move
would be considered a strong signal by the government. Until 1987,
producer prices were below the international prices (net of processing
costs), and the margin constituted a tax on producers. At present, how-
ever, international prices are so low that the margin has become negative
and producers are being subsidized.
This policy provided some protection to the agricultural sector from
the recession that was caused by the sudden decline in government ex-
penditures (this is partially illustrated in figure 14-1). Indeed, the rural
sector may have suffered less in Cote d'Ivoire than in other countries
from the adjustment measures.
Agricultural pricing and taxation policy has been an important element
of the structural adjustment debate in C6te d'Ivoire. Following trends in
the world market, the World Bank has consistently recommended an
increase in the price of coffee relative to cocoa. These recommendations
have led to only minor incentive programs for coffee, such as encour-
agement of pruning. At the then world prices, Ivoirian producers of coffee
and cocoa were being subsidized, and this situation was untenable from
a fiscal point of view. A lowering of producer prices seemed to be in
order, but the move was opposed by the Ivoirian authorities and by
producers, and for good reason. Because prices in the rest of the economy



C6te d'Ivoire: The Failure of Structural Adjustment 301
were rigid, a reduction in producer prices would have lowered the welfare
of the agricultural sector and reduced the incentive for exports.
A partial solution to this problem would be to lower the price of cocoa
and possibly coffee in order to finance an additional premium on non-
traditional agricultural exports. But the revenue generated from lower
coffee and cocoa prices would be insufficient to provide a meaningful
premium. Furthermore, this policy would not restore the "equilibrium"
values of the relative prices of agricultural products (including coffee and
cocoa) and nontradable goods in the economy. In other words, the fixed
nominal exchange rate and the apparent downward rigidity of the price
level impose severe constraints on agricultural tax policy.
EXPENDITURE. Government expenditure represents a very large fraction
of GDP (between 35 and 45 percent). Two changes in the pattern of ex-
penditure have characterized the adjustment period: a decrease in capital
expenditure and an increase in current expenditure. Capital expenditure
was kept at a high level (more than 15 percent of GDP) until 1982. It was
reduced to 7 percent of GDP from 1983 to 1984 and has been maintained
at that level since then (table 14-2). Changes in capital expenditure con-
stituted the dominant mode of policy adjustment in the 1980s.
Current expenditure actually increased during the adjustment period.
It is remarkable that the wage bill, one of the two major components of
current expenditure, increased as a share of GDP until 1983, decreased
slightly in 1984 and 1985, then increased again in 1986 and 1987 and
by 1987 was at a higher level than at the beginning of the adjustment
process in 1980. This increase represents an important failure of the
adjustment program: a sound principle of public finance is that the level
of current expenditure be geared to the long-run revenue capacity of the
government. This capacity, measured in terms of shares of GDP, was
smaller at the end of the 1980s than it was before the commodity boom
in 1975. Until the pattern of current government expenditure is altered,
a major component of structural adjustment will be missing.
The second major component of current expenditures is interest on the
debt. Most of the government debt is foreign debt because rules of the
West African Monetary Union prevent the government from borrowing
domestically. Interest payments as a share of GDP increased from 3 percent
to 9 percent between 1980 and 1984 because of the rise in international
interest rates. Although world rates have fallen since then, the interest
burden as a share of GDP has remained almost at the same level because
of the increase in the stock of the debt from 35 percent of GDP in 1980
to about 90 percent of GDP in 1989.
Because government expenditures represent such a large fraction of
GDP, the composition of these expenditures on goods and services affects
the allocation of resources between tradables and nontradables and so



302 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
may affect the orientation of the economy toward foreign markets. The
major links between fiscal policy and the allocation of resources are
through wage policy and the impact of fiscal policy on the composition
of aggregate demand.
Wage Policy and the Labor Market
The public sector is an important player in the labor market. Many an-
alysts believe that it contributes to the rigidity of the domestic price level
by ignoring macroeconomic conditions in the setting of wages. The gov-
ernment wage bill increased both as a share of GDP (see table 14-2) and
in real terms during the first period of adjustment between 1980 and
1983, when the recession was gradually deepening. Only in 1984, when
the fiscal crunch became acute, did the absolute level of the real wage
bill decrease. Estimates for 1987 indicate that these cuts may have been
only temporary.
The average wage in the public sector has an impact on the average
wage in the economy. Wages in the public sector are not determined by
conditions in the labor market but are a policy instrument, and current
policy rules out any decrease in nominal wages in the public sector. The
issue here is not whether wages are high in the public sector-although
casual evidence shows that they are much higher in CFA countries than
in non-CFA countries. The issue is whether the wage policy contributes
to the rigidity of nominal wages and prices throughout the economy and,
in addition, whether this rigidity prevents an adjustment of the real ex-
change rate. Given the fixed nominal exchange rate, this is an important
issue that needs to be examined. Although there have been some studies
of wage determination in Cote d'lvoire (Levy and Newman 1988; van
der Gaag and Vijverberg 1988), there have been no studies on the impact
of public sector wages on the labor market.
Resource Allocation and Impact on the Real Exchange Rate
Fiscal policy instruments such as the level of government expenditure or
the taxation of cash crops are important determinants of the real ex-
change rate, which should be taken into account in policy analysis. De-
varajan and de Melo (1987) emphasize the importance of this effect when
a large fraction of government expenditure goes to nontradables. An
increase in government expenditure then induces an increase in the supply
of nontradables and therefore a decrease in the supply of tradables. This
induces a movement along the economy's production frontier and a de-
crease in the real exchange rate, which is determined by the marginal
rate of transformation along the production frontier.



C6te d'Ivoire: The Failure of Structural Adjustment 303
This argument is correct, but it is not particularly useful for policy
analysis. An increase in expenditure on nontradables results in a shift of
resources because of the crowding out of domestic resources by the gov-
ernment. The impact on the production of tradables depends on the con-
cavity of the production frontier (that is, on the possibilities of substi-
tution in the economy). If the degree of substitution is large (small),
government expenditures have a large (small) impact on the production
of tradables. Therefore, it is useful to focus directly on the composition
of the crowding out of resources by the government.
The application of the Devarajan-de Melo analysis to the taxation of
cash crops is misleading because it does not focus on the actual link
between fiscal policy and the allocation of real resources. Thus govern-
ment taxation of cash crops to finance expenditures on foreign goods
would induce a shift of resources away from the tradables sector if a
fraction of the taxable revenue is spent on nontradable goods. The mea-
surement of the impact of fiscal policy on the allocation of resources
depends on the composition of expenditures and on the degree of sub-
stitutability in the economy. Thus the first task in an analysis of the
impact of government expenditure in Cote d'lvoire is to break down
expenditures by tradables and nontradables.
Savings, Investment, and External Equilibrium
Net savings by the government decreased after 1980 and became negative
by the end of the decade: capital expenditure was reduced from 18 percent
to 6 percent of GDP while the deficit was reduced only from 12 percent
to 8 percent. Without a drastic adjustment in expenditure or revenue, or
both, government net dissaving will prevent a resumption of growth. As
in other heavily indebted countries, rising interest on the foreign debt
has had a strong impact on the net savings rate. Most of the decrease in
the net savings rate has been due to the increase in interest payments on
the foreign debt. (This calculation assumes that interest payments of 7.4
percent of GDP are made; this has not been the case since April 1987,
when arrears began to accumulate, but the general analysis nonetheless
applies.)
The level of private savings has been remarkably stable, with the ex-
ception of the usual fluctuations due to recessions. The reason for this
stability, however, is that the government, through its fiscal policy, has
absorbed most of the external shocks to the economy. Devarajan and de
Melo (1987) show a correlation between the budget deficit and the ex-
ternal deficit.
Conclusion
Although the structural adjustment program resulted in some improve-
ments in the efficiency of the public sector, it did not achieve the main



304 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
objective of establishing the conditions for a resumption of growth. After
ten years, the price structure of the economy did not change to favor
diversification. Thus the economy remained heavily dependent on two
cash crops with bleak future prospects. Attempts to diversify the economy
toward nontraditional exports were a failure: nontraditional exports re-
mained at 5 percent of GDP since 1980. The policy measures that were
taken were too little and too late. The needed adjustments in current
expenditure and long-term revenue of the government and in the prices
of coffee and cocoa were not made, while the export premium was in-
troduced too late and at a time when deterioration in the government
budget prevented effective implementation.
The fixed exchange rate has constituted an important institutional con-
straint. Although it has provided price stability in CFA countries relative
to non-CFA countries in Africa, there have been costs associated with this
policy. The combination of rigidities in the price system and the policy
stance of the Ivoirian authorities has prevented a significant readjustment
of the real exchange rate. Current economic policy prevents adjustments
of relative prices within the agricultural sector, between the agricultural
sector and the rest of the economy, and between the tradable and non-
tradable sectors. Furthermore, the lack of price adjustment contributes
to the budget deficit. This worsening macroeconomic situation illustrates
that microeconomic policy actions alone are insufficient. C6te d'lvoire's
suspension of interest payments has delayed the day of reckoning, but
this is an unstable condition that will have to end soon.
Notes
1. This chapter describes the World Bank structural adjustment program in
C6te d'lvoire and discusses its contribution to the Ivoirian economy. It does not
discuss the 1MF program. Although additional analyses and counterfactual simu-
lations would no doubt provide interesting additional insights, they could not
be undertaken here because of space limitations and because a significant part
of the design of the program was left to the IMF. The purpose here was simply
to describe the macroeconomic adjustment of the Ivoirian economy and discuss
the main issues.
2. This attribution of shares to external and internal factors is particularly
simple and does not require a structural model for determining the endogenous
supply response because the output of coffee and cocoa is inelastic in the short
run. Thus an increase in world prices would have no effect on output. This
analysis also ignores the induced effect on revenue of the 1985 recession brought
on by the sharp reduction in government expenditure. These induced effects are
very much smaller than the initial expenditure shock and are not necessary in
this approximation of effects.
3. Treasury expenditures increased by 80 billion 1985 CFA francs, most of
them unclassified payments that may have been in arrears; the government wage
bill increased by only 6 billion CFA francs.



C6te d'Ivoire: The Failure of Structural Adjustment 305
References
Devarajan, Shantayanan, and Jaime de Melo. 1987. "Adjustment with a Fixed
Exchange Rate: Cameroon, C6te d'Ivoire, and Senegal." World Bank Eco-
nomic Review 1 (May):447-87.
Levy, Victor, and John Newman. 1988. "Labor Market Adjustment during a
Recession: The Micro and Macro Evidence." World Bank, Washington, D.C.
van der Gaag, Jacques, and Wim Vijverberg. 1988. Wage Determinants in C6te
d'Ivoire. LSMS Working Paper 33. Washington, D.C.: World Bank.
Comments
Shantayanan Devarajan
CHRISTOPHE CHAMLEY'S chapter documents one of the saddest stories of
the 1980s. After achieving an average annual GDP growth rate of 8 percent
in 1965-75, Cote d'Ivoire's economy slowed to a grinding halt in the
1980s. The growth rate at the end of the 1980s averaged a little over 1
percent. Moreover, this slowdown was caused not by an adverse external
shock but by a favorable shock: the coffee and cocoa boom of the mid-
1970s. The windfall revenue from this boom was spent on an investment
program. When the boom ended, the government continued spending,
which led to unsustainable current account deficits, a debt overhang, and
a secular decline in competitiveness. Worse still, this favorable terms of
trade shock was known to be temporary at the outset, since it was caused
by a frost in Brazil.
In his case study Chamley divides C6te d'lvoire's adjustment experi-
ence into three phases: (1) an initial phase, when a structural adjustment
loan was issued with only mild conditionality; (2) a second loan, with
stricter conditionality, which was able to be met in part because of a
favorable turn in coffee prices during the same period; and (3) a third
phase, up to the end of the 1980s, in which the economy continued to
perform poorly and many of the original structural adjustment issues
continued to be discussed. Chamley further divides the policy measures
into those dealing with internal and external balance, price incentives,
and administrative reform. His overall conclusion on C6te d'Ivoire's
structural adjustment experience can be summarized by the words "too



306  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
little, too late." I will focus on the first two policy areas and on his overall
conclusion.
On the subject of internal and external balance the most important
point, not sufficiently emphasized by Chamley, is that in C6te d'Ivoire
there is very little difference between internal and external balance. First,
when the government fixes the producer price of the two major export
crops (coffee and cocoa), changes in world prices will affect the fiscal
deficit and current account deficit symmetrically. Second, in C6te d'Ivoire
the fiscal deficit was financed by foreign borrowing (which is the mirror
image of the current account deficit). This was necessary because, on the
one hand, the country's membership in the CFA zone essentially prevents
financing its deficit through seigniorage revenue and, on the other hand,
the weakness of the domestic capital market makes government borrow-
ing from the private sector virtually impossible. Third, a quick look at
the facts shows that the current account and fiscal deficits were roughly
the same percentage of GDP throughout the adjustment period.
The link between the two deficits has implications for the setting of
price incentives in C6te d'lvoire. For example, it may be thought that
raising producer prices for coffee and cocoa would increase export in-
centives and thereby cut the current account deficit. But this would also
increase the fiscal deficit (if expenditure is held constant). Furthermore,
the fiscal deficit is financed by foreign borrowing, and the real appre-
ciation associated with the inflow of foreign capital may act as a dis-
incentive to export, with the end result being lower exports. By separating
price incentives from the discussion of internal and external balance,
Chamley fails to draw this connection between the two.
As for expenditure policy, Chamley points out clearly how this is both
the cause of and possibly the solution to C6te d'lvoire's problems. When
coffee and cocoa prices fell in the late 1970s, government expenditure
continued to rise, leading to an increased foreign debt and large current
account deficits. In adjusting to this crisis, too, expenditure policy may
be one of the few instruments available. As Chamley shows, the tax/GDP
ratio in C6te d'lvoire (leaving out the revenue from the price stabilization
fund) stayed constant throughout the 1980s. The produce marketing
board's revenue fluctuates with world prices and hence is procyclical.
Thus the only way to address the twin deficits is to cut expenditure. It
should be added that C6te d'lvoire did undertake major expenditure cuts
beginning in 1983. Capital expenditure fell from 18 percent of GDP in
1980 to 6 percent in 1986. Total government expenditure was reduced
from 41 to 34 percent of GDP in the three-year period 1983-86. Chamley
does not say why more could not have been achieved on the revenue
side. He states that the World Bank decided to defer to the IMF on this
issue. If this strategy has failed to yield the desired results, perhaps the



C6te d'Ivoire: The Failure of Structural Adjustment 307
World Bank should consider an alternative strategy. For example, the
value added tax may be used as a revenue-enhancing instrument.
Almost all the discussion about price incentives in C6te d'Ivoire is
related to the country's membership in the CFA zone. Although lowering
export taxes may provide price incentives, this may be counterproductive
in the end. The most important price incentive for adjustment, however,
is the real exchange rate, or the relative price of tradables to nontradables.
In the CFA zone, with the nominal exchange rate fixed and world prices
exogenous, the main means of affecting the real exchange rate is via the
price of nontradables. This is linked to wages, which are in turn linked
to the operation of the labor market in C6te d'lvoire. It would have been
helpful to learn more about this, especially about what steps are being
taken to depreciate the real exchange rate.
Under the second SAL an attempt was made to simulate a devaluation
by imposing a uniform tariff and subsidy scheme. First, quantitative re-
strictions were removed and a uniform import tariff of 40 percent was
introduced; then an export subsidy of about 30 percent was established.
The intent was to raise the price of tradables relative to that of non-
tradables and thereby achieve a real depreciation. Unfortunately, two
factors prevented this outcome. First, the import tariff gave rise to smug-
gling, which undermined the collection of revenue from the tariff. Second,
to cut costs the government stopped paying the export subsidies. Ex-
porters were then worse off than before because the harmonization of
the tariff and subsidy scheme led to higher tariffs on intermediate goods.
One final point about price incentives deserves mention. There is a
temptation during the structural adjustment process to try to get as many
prices as possible "right." There also appears to be a belief, implicit in
Chamley's chapter and more explicit in other documents, that prices in
C6te d'lvoire are far from "right." Yet, even with such distortions, how
did the economy manage to grow at 8 percent a year in the past? Was
it luck, foreign aid, or some dynamic internal forces in the economy? Put
another way, if only the internal and external balance of the country is
restored, will the economy resume rapid growth?
Chamley describes two kinds of institutional reform that were intro-
duced as part of the structural adjustment process: improvements in the
selection of public projects and new methods of budgeting. Although
these are useful reforms, why does nothing appear to have been done
about tax administration? After all, if the problem was the fiscal deficit,
it should be attacked from both the revenue and the expenditure sides.
Chamley's conclusion that adjustment in C6te d'lvoire was a case of
"too little, too late" should be qualified by asking what could have been
done differently, given the institutional constraints in the country. Plainly,
devalu-ing the exchange rate was not an option. It is true that some earlier
cutbacks in government expenditure might have helped, although the



308 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
expenditure/GDP ratio did fall significantly. Another option would have
been to increase the tax/GDP ratio, whose obdurateness remains a puzzle.
Of course, to answer this question properly would require, as Chamley
notes, some counterfactual simulations. He says that this would call for
more analytical work. Nevertheless, he attributes the success of the sec-
ond SAL tO good luck (high coffee and cocoa prices) and good policy in
equal proportions. This is a counterfactual simulation, although we are
not sure what analytical framework he used.
More generally, there are several analytical frameworks being used
both directly and indirectly as part of the SAL appraisal process. It is
worth asking what counterfactuals were performed as part of the ap-
praisal and what could be performed today with these models.
Some perspective on the constraints on C6te d'Ivoire's adjustment ex-
perience can be obtained by comparing its performance with that of
Ghana, an economy that is similar in many ways but with one crucial
difference: Ghana is not a member of the CFA zone and hence can use
the nominal exchange rate as an instrument of adjustment. During the
1980s C6te d'Ivoire had consistently lower inflation than Ghana. In fact,
the highest annual rate in Cote d'lvoire was lower than the lowest ob-
served in Ghana. But at the end of the period C6te d'Ivoire had a much
higher foreign debt: 125 percent of GDP, as opposed to Ghana's 40 percent
of GDP.
This contrast illustrates the tradeoff involved in belonging to the CFA
zone: by relinquishing the seigniorage tax as a means of raising revenue,
countries maintain lower inflation rates. However, they may have to rely
to a greater degree on foreign borrowing, which may leave them with
serious long-term problems.



15 Ghana: Ending Chaos
Chad Leechor
AT THE LOWEST POINT of the economic crisis in Ghana in 1982 and 1983,
the country was a scene of devastation. Power systems had ceased to
function, even in Accra, the capital city. Communication, postal, and
railroad services had broken down. Tax collection had fallen to about S
percent of GDP. Investment had fallen below the level needed to maintain
the capital stock. Foreign exchange reserves had been nearly depleted.
Real income per capita, which had been falling for almost a decade, was
by 1983 a third below the level reached in the early 1970s (figure 15-
1).
The main cause of the crisis was a failure of economic policy. Fiscal
discipline was weak as the government pursued extensive industrializa-
tion and public sector employment without commensurate efforts to in-
crease revenue. Budgetary deficits were financed mainly through mon-
etary expansion, which accelerated inflation. Because the official
exchange rate was virtually unchanged for a decade, it created substantial
overvaluation of the local currency. Domestic production was inhibited
by extensive controls on prices and distribution. Exports were depressed
by cumbersome regulations, explicit export taxes, and hidden taxes in
the form of low producer prices and overvaluation of the exchange rate.
Although these policies led to economic decline, the severity of the
crisis was also attributable to several shocks beyond the control of the
government. First, a protracted drought ravaged Sub-Saharan Africa, and
Ghana was not immune to its destructive effects. Agriculture, the largest
sector, was particularly hard hit, and the repercussions were felt through-
out the economy. Second, the world price of cocoa, Ghana's predominant
export crop, had taken a sharp turn downward, while the price of crude
oil, Ghana's major import, had gone up significantly. Third, about a
million Ghanaian workers returned home from Nigeria. Ghana was thus
deprived of a customary source of foreign exchange from remittances,
while its unemployment rose. External debt had little to do with the crisis,
however, as the stock of debt was only about US$1 billion in 1982 (4
percent of GDP at the official exchange rate or 20 percent at a more
realistic rate) and the debt service ratio was about 15 percent.
Before 1983 the Ghanaian authorities had progressively increased con-
trol over the economy. As inflation rose, there was a tendency to step
309



Figure 15-1. Index of Per Capita Income in Ghana, 1957-87
Index (1980   100)
140 -
130 -
120 -
o~j                             110                                                     A
100 a
90                                       I
80
1957   1960          1965         1970          1975         1980         1985 1987
Sources: IMF, International Financial Statistics, various issues; Roemer (1984).



Ghana: Ending Chaos 311
up enforcement of price controls. As trade deficits widened, imports were
increasingly restricted through licensing requirements and foreign ex-
change allocation. As foreign exchange reserves dwindled, efforts were
made to limit the number of exporters and to administer export prices,
both with the objective of stopping foreign exchange leakages. Much
attention was devoted to alleviating undesirable economic symptoms,
while the more fundamental causes were seldom addressed. The cedi (¢)
was devalued twice between 1974 and 1983, raising the exchange rate
from ¢1.15 to ¢2.75 to the U.S. dollar. But the adjustment was insig-
nificant in relation to the forty-six-fold increase in the domestic price
level that had occurred in the intervening years.
Adjustment policies had been deliberately avoided since 1971, when
an early experiment was prematurely aborted. At that time, in response
to a balance of payments crisis, the Busia administration devalued the
cedi sharply, by 82 percent against the U.S. dollar. The measure proved
divisive; although cocoa farmers and rural workers generally welcomed
it, urban consumers were vehemently opposed. A military coup was car-
ried out with little resistance, and the new administration quickly reversed
the exchange rate policy and set the cedi at its original level. Subsequent
administrations appeared to regard the episode as a warning against
adjustment policies in general.
The Adjustment Program
A far-reaching policy reform, the Economic Recovery Program (ERP), was
announced in April 1983. It called for increased fiscal discipline; more
realistic pricing of goods, factors, and foreign exchange; and a deter-
mined, if gradual, shift away from administrative control toward more
reliance on markets. In almost three decades of Ghana's independence,
this approach to policymaking had never taken hold, so at its onset it
seemed fraught with dangers. It was feared that those who would be
adversely affected might try to derail the reform process through eco-
nomic or political means. Structural rigidities threatened to slow and
diminish the supply response. In spite of these apparent risks, the ERP
was launched and sustained through a very difficult initial stage. As it
turned out, the program was successfully implemented over four years
and was followed by a more comprehensive plan designed to build on
the initial gains and to facilitate broader structural adjustments.
The ERP was conceived by the government of Ghana in collaboration
with the World Bank and the International Monetary Fund (IMF). Support
from the IMF was critical in the initial stage, which involved primarily
stabilization measures to restore internal and external balances. The role
of the World Bank was more prominent in a subsequent stage, when the



312 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
government, having demonstrated its commitment to stabilization, at-
tempted to rehabilitate the economy and restore growth.
Stabilization Measures
The ERP gave high priority to fighting inflation at home and securing a
sustainable external balance with an adequate flow of imports. To curb
inflation, which had been rising through the 1970s and had exceeded
100 percent a year in the early 1980s, efforts were made to improve fiscal
control and thereby to restrain monetary expansion. Initially, fiscal con-
trol was achieved mainly through a substantial cut in government ex-
penditure from 10.2 percent of GDP in 1982 to 8.6 percent in 1983. The
cut was implemented across the board and encompassed reductions in
real wages and operating expenditures as well as in development outlays.
Subsequently the emphasis shifted to resource mobilization, particularly
increased tax revenue, to support an expansion in public services and
infrastructure. Government revenue (inclusive of grants) rose steadily
from 5.6 percent of GDP in 1983 to 13 percent in 1986 (figure 15-2).
Budget deficits were reduced from about 7.5 percent of GDP during the
economic downturn in 1979-82 to 2.7 percent of GDP in 1983 and well
below 2 percent of GDP thereafter.
In working to achieve external balance, the Ghanaian authorities did
not want to rely on compression of imports, as had been done in the
past. But at the onset of the ERP, prospects for earning foreign exchange
were grim because the exchange rate had been misaligned for more than
a decade. The combination of high domestic inflation and little change
in the nominal exchange rate had made the cedi one of the most over-
valued currencies in the world. Since the required adjustment was ex-
tremely large, the government chose to undertake periodic devaluations
over several years. In 1983 the exchange rate was devalued from ¢2.75
to the U.S. dollar to ¢30, a magnitude seldom seen elsewhere. Further
devaluations brought the exchange rate to ¢50 to the U.S. dollar in 1984,
460 in 1985, and ¢90 in 1986 (figure 15-3). Since 1987 the exchange
rate has been determined by foreign exchange auctions.
Before adequate foreign exchange earnings were secured, however, es-
sential imports had to be procured to maintain domestic production and
to meet local demand. Although Ghana's external debt was moderate in
size, the fact that its payments were in arrears had blocked its access to
external resources. The IMF provided a sizable bridging fund, which al-
lowed Ghana to pay off a large part of its external arrears and thereby
improve its ability to borrow abroad to finance its imports.
Adjustment Measures
To assist recovery and encourage growth, adjustment measures under
the ERP attempted to shift away from administrative control of allocative



Ghana: Ending Chaos 313
Figure 15-2. Government Revenue and Expenditure as a Percentage of
GDP in Ghana, 1978-87
Percent
16-
14-
1 2 
\                       I   /
IExpenditure/GDP      I
\\           ~~~~~~I/
10                                  I
I__          /
I      I
8 - 
6                           R evenue/GDP
1978 79  80  81  82  83  84  85  86   87
Source: IMF, International Financial Statistics, various issues.
mechanisms and move toward greater reliance on markets. The shift was
meant to be selective and gradual, however, because the magnitude of
existing distortions was overwhelming. Moreover, the depleted infra-
structure severely constrained production and distribution, while the re-
sources available were extremely limited. Too abrupt a change might
have entailed excessive economic and social costs. Firms might not have



314 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Figure 1S-3. Nominal and Real Cedi Exchange Rate Relative to the
U.S. Dollar, Annual Average, 1978-87
400 -
350 -
Real average yearly rate
I
300-
250-
e  200 -
rrl                            ~~~~~~~/
ISO      /z                         /
100 A
so                               I
~*~%    I    Nominal average
yeryrate
0~                                            I    I 
1978  1979     1981      1983      1985      1987
Source: IMF, International Financial Statistics, various issues.
survived the transition period, and vulnerable groups might have expe-
rienced undue hardship. The World Bank played an active role in sup-
porting this transition, assisting in the design of the adjustment program
and providing resources for removing the most serious obstacles on the
path to recovery.
Given the environment in which the ERP was launched, extensive re-
forms were needed. A full description would be tedious and not partic-



Ghana: Ending Chaos 315
ularly instructive, so the following discussion concentrates on the major
reforms, grouped according to broad objectives.
THE EXTERNAL SECTOR. The first and probably most important mea-
sure under the ERP was to correct the massive misalignment of the ex-
change rate. The series of devaluations begun in April 1983 effected a
large real depreciation. By early 1987 the exchange rate had reached the
level of ¢160 to the U.S. dollar, compared with the rate of 42.75 pre-
vailing in early 1983. The process of adjustment was somewhat con-
voluted, consisting of multiple rates initially, followed by unification and
a few large devaluations. Finally, the official rate was left to be deter-
mined in a weekly auction.
On trade policy, the ERP attempted to increase exports and trade flows
overall, rather than relying on import restrictions to reduce external im-
balances. The exchange rate policy was a primary instrument for pro-
viding incentives to actual and potential exporters. In addition, foreign
exchange was allocated to ensure the rehabilitation and recovery of tra-
ditional exports. In the interim phase, a foreign exchange retention
scheme was introduced to reduce the remaining bias against exports. The
scheme was gradually phased out as the exchange rate moved toward a
sustainable level. Export procedures were simplified, and administrative
price setting was abolished. The World Bank assisted the export pro-
motion efforts through three adjustment loans. The first one was for the
rehabilitation of traditional exports; foreign exchange and local currency
funds were provided to revive the cocoa, timber, gold, and diamond
sectors. The other two, an industrial sector adjustment loan and a struc-
tural adjustment loan, attempted among other things to promote non-
traditional exports through incentives, access to foreign exchange, and
relief from indirect taxes.
Although import controls were extensive at the time the ERP was
launched, an immediate shift to a tariff-based import regime was con-
sidered too risky in the presence of remaining exchange rate overvalu-
ation and foreign exchange shortages. Instead, the licensing system was
streamlined, mainly by establishing priorities in the annual import pro-
gram and integrating import and foreign exchange licenses. In addition,
no restrictions were placed on imports unless foreign exchange was re-
quested from the government. Subsequently, a foreign exchange auction
was introduced as an allocative mechanism, and an open general license
was granted for the foreign exchange acquired. Import tariffs were re-
duced and simplified. The World Bank played a major role in this process
of gradual trade liberalization, supporting the adjustment measures
through a structural adjustment credit and two reconstruction import
credits.



316 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Apart from trade, the government significantly reduced its arrears in
external payments from more than US$200 million in 1983 to less than
US$20 million in 1985. Ghana's improved credit standing and its efforts
in seeking external assistance, both financial and technical, enabled it to
draw on external capital during the economic reconstruction.
PUBLIC FINANCE. Improved fiscal discipline was a major objective of
the ERP. The strategy adopted was to restrain expenditure sharply at the
outset and then to allow some increase in expenditure as more revenue
was raised. Much attention was devoted to the structures of the tax
system and the expenditure program to minimize the adverse impact of
individual measures.
Revenue gains came primarily from the devaluations, which increased
the duties collected on imports. A temporary import surcharge was also
applied to mop up scarcity rents resulting from quantitative restrictions.
Export taxes, with the exception of the cocoa tax, were phased out.
Personal income taxes had become highly punitive under protracted in-
flation since deductions lost their real value and tax brackets became
small in real terms. Gradual reforms were introduced. Another important
change was in tax administration, which was upgraded by layoffs of
incompetent staff, recruitment of a new managerial and professional
team, improved compensation, and ultimately reorganization of the tax
bureau. Administrative fees, which had declined in real value after many
years of high inflation, were raised across the board. The World Bank
and the IMF collaborated in providing technical assistance for the reform
of tax policy and administration.
Expenditures were severely restrained initially, despite obvious defi-
ciencies in infrastructure and civil service salaries. As revenue rose, how-
ever, remedial actions were quickly undertaken. In 1984 and 1985 civil
service wages and salaries were raised substantially. By 1985 the gains
in revenue were sufficient to permit a significant increase in development
expenditures as well, much of which was for rehabilitating transport
facilities. In 1986, however, the introduction of an overambitious pay
scale caused fiscal control to deteriorate. The overall fiscal balance for
the year was eventually restored, however, as the new pay scale was
revised downward and development expenditures were reduced.
The World Bank supported the process of rationalizing expenditures
initially with two import credits for measures to curtail subsidies, par-
ticularly for fertilizer, and to increase accountability for public invest-
ment. Subsequently, greater financial discipline was imposed on public
enterprises to limit their losses and the need for budgetary transfers. These
measures were supported in part by an industrial sector adjustment credit
and a structural adjustment credit.



Ghana: Ending Chaos 317
DEREGULATION. At the inception of the ERP, the domestic economy was
a fifth below the peak reached in the 1970s. Much of the capital stock
was in disrepair. Intermediate goods were unavailable because of foreign
exchange shortages. Moreover, many producers had terminated or scaled
down production as a result of pervasive controls on prices and distri-
bution.
A prominent feature of the ERP was its reliance on incentives and com-
petition as stimuli for production. This approach deviated sharply from
the previous reliance on price and distribution controls, which had de-
pressed profits and led to smuggling and shortages. The government
moved with caution, however, lifting controls gradually, and for some
products no deregulation was attempted.
Prices of consumer goods were decontrolled first. The number of goods
subject to price control was reduced from twenty-three to thirteen in
1984 and to eight in 1985. Price controls remained for sensitive inputs,
including petroleum products, cotton, tobacco, and cocoa. The objective
of price setting changed, however, from limiting consumer costs to stim-
ulating producer responses. Producer prices for cocoa, for example, were
raised well in excess of the devaluations; as a result, the ratio of producer
prices to border prices rose from about 20 percent in 1982 to about 50
percent in 1985. Similar increases were also granted for petroleum prod-
ucts, cotton, and tobacco.
Changes in distribution controls followed a similar pattern. State-
owned monopolies were progressively eliminated along with price con-
trols. In some cases, only prices remained subject to control while pro-
duction and distribution were open to competition. Cocoa, however, was
subject to both price and distribution controls because of its strategic
role in generating foreign exchange and government revenue. But some
improvements were made in efficiency by liquidating estate holdings,
cutting the work force, and streamlining operations.
INSTITUTIONAL REFORMS. The civil service and state-owned enterprises
were extensively restructured to contain operating costs and enhance
productivity. Staffs were widely cut in the public sector, and compen-
sation was upgraded for those remaining. (In 1986, for instance, the
cocoa board took 25,000 "ghost" workers off the payroll.) The man-
agement and operating procedures in many major state-owned enter-
prises, including the cocoa board and the gold mine, were reviewed with
the assistance of local and foreign experts.
Economic Performance
Following implementation of the ERP, aggregate production rose by more
than 6 percent annually in constant prices for three years, compared with



318 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
a 3 percent annual decline in the three years preceding the ERP. Domestic
savings and investment also rose significantly, their respective shares in
GDP nearly doubling. The largest gains came in 1984, the first full year
after the ERP was announced and implemented. In later years growth was
more moderate, but still significant.
Economic Responses and Recovery
The response was most dramatic in the industrial sector, where capacity
utilization had been very low. Imports of intermediate goods and spare
parts, made possible initially by the injection of foreign exchange from
the IMF and the World Bank, allowed manufacturing firms to expand
production rapidly. After the introduction of the ERP, real manufactured
output rose at an annual rate of 14 percent. The remaining industrial
subsectors showed some recovery, although not as much as manufac-
tures. For the industrial sector as a whole, real output rose by 11 percent
a year during the adjustment period, compared with a decline of about
the same magnitude before adjustment.
Agricultural response was also strong. In particular, food crop pro-
duction rose by about 80 percent in 1984 and by about 10 percent there-
after. This expansion, however, cannot be attributed to policy change
alone. The return of normal rains after a protracted period of drought
helped make the turnaround possible. The value of output in forestry
and logging grew by about 3 percent a year in constant prices. The ag-
ricultural sector as a whole grew by more than 5 percent a year. Exports
of cocoa rose in foreign exchange value by 43 percent and timber by
nearly 100 percent within two years of the introduction of the ERP.
The upsurge of activities in agriculture and industry was accompanied
by significant growth in the services sector. As expected, the transport
and trading subsectors, which had the most extensive linkages to agri-
culture and manufacturing, grew the fastest. The entire service sector
grew at an annual rate of 6 percent after the introduction of the ERP, in
contrast to its downward trend before 1983.
Along with production, savings and investment also recovered even
though they were not of immediate priority under the ERP. The rate of
fixed capital formation rose from a low 4.5 percent of GDP before the
ERP to 8.3 percent thereafter. Private saving was largely unchanged after
the introduction of the ERP, but dissaving by the public sector was cut
substantially so that net domestic savings rose from 4.2 percent of GDP
to about 7.2 percent. Despite this recovery, however, the level of savings
and investment remained low by international standards. This outcome
may have been anticipated by the authorities, who had emphasized re-
habilitation and increased utilization of existing productive capacity
under the ERP. The second phase of the ERP, which was launched in 1987,



Ghana: Ending Chaos 319
appears to have moved toward a greater emphasis on resource mobili-
zation and capital formation.
One casualty of the economic crisis was the financial sector. Inflation
and general economic deterioration substantially eroded public confi-
dence. Domestic financial assets lost their real value over time. Savers
turned to assets less susceptible to inflationary erosion, including foreign
currency (such as the CFA franc) and real assets. The considerable financial
depth that Ghana had attained by the mid-1970s was lost in the process,
as the ratio of bank deposits to GDP fell from 27 percent in 1977 to about
10 percent in 1983 and 1984. Some progress was made thereafter, as
nominal deposit rates were raised to achieve positive real returns, and
the deposits to GDP ratio rose slightly to about 14 percent in 1986. The
financial sector still has a long way to go to achieve full recovery, how-
ever.
Macroeconomic Balance
A major achievement of the ERP has been improved fiscal control. Budg-
etary deficits were substantially reduced over the adjustment period, from
an average annual rate of about 8.8 percent of GDP in the decade preceding
the ERP to an average of 1.6 percent from 1983 to 1986. This improve-
ment was not accompanied by comparable credit and monetary disci-
pline, however, and inflation remained a serious issue throughout the
adjustment period.
In each year after the introduction of the ERP, the central bank extended
far more new credit to the public sector than it needed for deficit financing
(see table 15-1).' This lending was the main cause of the rapid growth
in high-powered money (monetary base), which averaged about 50 per-
cent a year from 1983 to 1986. With the money multiplier remaining
unchanged over the period, the money supply grew by the same rate as
the monetary base. Such rapid monetary expansion put considerable pres-
sure on prices, even when scarcity rents fell as a result of increased pro-
duction and imports. The consumer price index rose by an annual rate
of 25 percent between 1983 and 1986, which was much lower than before
the crisis (about 60 percent), but quite high in view of the considerable
success in deficit reduction.
External Balance
Before the adjustment program, import and foreign exchange controls
had been the main instruments for balancing external payments. The
resulting shortages of imports had constrained production and con-
sumption and lowered the overall volume of international trade. Import



320 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 15-1. Selected Monetary Aggregates for Ghana, 1983-88
(billions of cedis)
Aggregate                   1983         1984         1985        1986
Central bank credit to
public sector           24.5         38.6         57.5          94.5
Government                24.3         3S.1         45.0         78.0
Parastatals                0.2          3.5         12.5         15.9
Increase in public sector
credit                   -           14.2         18.8         37.0
Monetary base               14.6         21.8         29.6         47.3
Increase in monetary
base                     -            7.2          7.8         17.7
Money supply (Ml)           16.7         26.8         38.3         55.2
Increase in Ml               -           10.1         11.5         16.9
Government deficits          4.9          4.8          7.2        -0.3
Source: Bank of Ghana and World Bank staff estimates.
restrictions had usually been retained until a favorable exogenous shift
occurred in the terms of trade.
A new approach was pursued under the ERP. Instead of relying on
import restraints, the authorities attempted to create three conditions
consistent with a sustainable external balance. They wanted to (1) in-
crease production, (2) reduce absorption, primarily by controlling gov-
ernment expenditures, and (3) increase external demand for domestic
goods (expenditure switching) by eliminating overvaluation of the cedi.
This new approach allowed imports to rise substantially, from about
US$539 million in 1983 to US$780 million in 1986. Exports, stimulated
by devaluation of the exchange rate, increased even more. Thus the trade
gaps were narrowed during the recovery without any sacrifice in domestic
consumption or production, and the volume of international trade was
expanded.
The new approach also inspired confidence in the government's eco-
nomic management. As a result, the prospects for foreign investment and
external assistance improved, and the problem of capital flight became
less serious. During the recovery period the net capital account balance
shifted significantly in the country's favor, complementing the improved
trade balance (see table 15-2). The strengthened external position allowed
the country to pay off a large part of its external arrears and to replenish
its foreign exchange reserves.
A Long-Term Perspective on the Recovery
Clearly, the economic recovery since the inception of the ERP has been
impressive. The standard of living for the average Ghanaian has improved



Table 15-2. Balance of Payments in Ghana, 1979-86
(millions of U.S. dollars)
Item                                 1979       1980        1981        1982        1983       1984        1985        1986
Current account
Merchandise                           184         132        -310          10       -100       -115         -95          -7
Exports (f.o.b.)                  1,066       1,104         711         641        439         566         632         773
Imports (c.i.f.)                  -882        - 972      -1021        -631       -539        - 681       -727        -780
Nonfactor services                    -95       -104         -113        - 85        -65        -91        -110        -130
Resource balance                       89          28       -423         -75        -165       -206        -204        -137
Net factor income                     -46        -80          -81        -82         - 82       - 81       - 111       -105
Private transfers (net)                -3         -3           -4         -1           17         73          33          49
Current account balance                40        -55        -508        -158        -230       -215        - 283       - 193
Capital account
Grants (official transfers)            82          83           87         84          72        141          93         115
Official long-term loans (net)        108          96          49          16          15        131         108         213
Official medium-term loans (net)      - 14       -13           42          97          13       -42         -70        - 102
Trust fund                              34          0            0          0           0         -1          - 7       -11
Direct foreign investment              -3          16           16         16           2           2           6          4
Private capital (net)                  -5           0         -13         -5           12       -14            5          18
IDA allocation'                         14         14           14          0           0           0          0           0
Capitalb                              -26          12            0         13          61       -85           15        -18
Errors and omissions                -105        -141            63       -36        -187        -38           16        -18
Net capital account                     85         67         258         185        -13          94         165         138
Overall balance                       125          12       -250           27       -243       -121        -117         -56
a. From the International Development Association of the World Bank.
b. Not elsewhere specified.
Source: Bank of Ghana and World Bank staff estimates.



322  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
considerably since 1983. Policymaking has improved as well. Its success
has made Ghana something of a model for other African countries and
earned praise from the world media and the international community.
A more cautious observer, however, would look at the recent recovery
not only in the light of the crisis that preceded it but also in the context
of longer-term development. At the depth of the crisis, per capita income
had fallen by more than one-sixth of its level at independence. By the
end of 1987, four full years after the adjustment program started, Ghana's
real per capita income was still one-third below the peak level reached
in the mid-1970s and one-tenth below the level prevailing at independ-
ence in 1956 (see figure 15-1). There has been no real gain in the average
standard of living after three decades. Although economic output more
than doubled in the postindependence period, population grew even
faster, leaving the average person with less income than at independence.
Thus Ghana has lost considerable ground in terms of development.
Once one of the most prosperous developing countries, Ghana is today
among the poorest. Since Ghana's independence, developing countries
as a whole have experienced an average annual growth rate in per capita
income of 3 percent, while African countries averaged 1.7 percent. In
this period per capita income therefore rose by 109 percent in developing
countries as a whole and by about 53 percent in African countries, while
it declined by about 10 percent in Ghana. Ghana's record since the ad-
justment program started has been much better, however, and compares
favorably with trends in the rest of the world. In the five-year period
1983-88 the economy maintained an average growth rate of 5 percent
a year.
Assessment of Conditionality
What was the role of the World Bank in Ghana's adjustment process,
and what was Ghana's record of compliance with adjustment loan con-
ditionality between 1983 and 1986? In answering these questions, only
the conditions made in the context of adjustment loans extended by the
World Bank between 1983 and 1986 are examined. The following anal-
ysis does not review the ERP as a whole, the conditionality set by the iMF,
or traditional World Bank project operations. Also excluded is the struc-
tural adjustment credit that became effective in fiscal 1986, since it is
too early to draw any firm conclusions about it. Through this process
of elimination, four adjustment loans are left for the present review: the
first reconstruction import credit, an export rehabilitation credit, the
second reconstruction import credit, and the industrial sector adjustment
credit.
Before presenting the results of this analysis, several caveats must be
stated. First, the assessment reflects judgments based on partial infor-



Ghana: Ending Chaos 323
mation. There are gaps in information about the deviation of actual
measures from required conditions, sometimes because no indicators for
measuring progress were specified. The conclusions reached are neces-
sarily tentative. Second, the conditions set for the loans do not necessarily
reflect the full role of the World Bank in the adjustment process. Many
actions considered desirable by the World Bank may not appear explicitly
as loan conditions. The government may not have been willing to endorse
them or, conversely, it may have been ready to implement them without
prompting from the World Bank. In other cases, a provision might not
be included if it were already covered in an agreement with the IMF. Thus
the conditions reviewed in this section cover only a small fraction of the
common ground between the World Bank and the government.
Scope of Conditionality
Tlhe scope of World Bank conditionality varied in the course of Ghana's
economic recovery. Initially, when the country was still in the depth of
the economic crisis, the World Bank set conditions mainly for emergency
relief. As the country recovered and the World Bank acquired experience
in policy-based lending, the focus shifted to rehabilitation and capacity
utilization. Finally, in 1986 and 1987 the questions of growth and the
quality of investment were tackled through industrial sector and struc-
tural adjustment credits. As the objectives grew more complex, the num-
ber of conditions expanded from ten per loan in 1983 to sixteen in 1985
and thirty-nine in 1987.
The conditions under these loans were generally clearly stated. As-
sessing compliance is not straightforward, however. In some cases no
indicators were specified for measuring progress and no targets or dates
were identified. For instance, one loan required reinstatement of the duty
drawback program before release of the second tranche, but no indicators
were specified. Reinstatement of the duty drawback program in itself
would have served no useful purpose, however; what needed to be dem-
onstrated was that the duty drawback program had been of benefit to
exporters. And it was virtually impossible to monitor compliance or to
determine whether any progress had been made in the case of conditions
that simply called for the establishment of a task force or study group
to resolve specific problems (such as divestiture of public enterprises or
dissolution of the cocoa marketing mechanisms). Clearly the mere ex-
istence of such teams served no relevant purpose, but no method was
specified for determining whether the intended work was being done.
Implementation
Ghana's record of policy implementation is nonetheless very good, per-
haps exceptional. The core conditions for all the adjustment loans were



324 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 15-3. Ghana's Implementation of Conditionalities under World
Bank Adjustment Loans
Satisfactory
implementation             Partial implementation    Little or no progress
Exchange rate policy     Civil service compensation   Cocoa marketing arrange-
Foreign exchange alloca-    Financial management re-  ments
tion                     form (cocoa board)     Monetary and credit disci-
Import liberalization    Divestiture of public enter-    pline
Resource mobilization      prises                 Prudential management
Decontrol of prices and  Petroleum products         and regulatory frame-
distribution           Domestic indirect taxation  work for the financial
Pricing of agricultural  Income tax policy          sector
products               Import taxation           Duty drawback scheme
Pricing of freight and utili-  Export taxation
ties
Institutional reforms in
civil service and mining
Commercialization of fer-
tilizer sector
Source: World Bank reports.
implemented to a satisfactory degree. This record is particularly re-
markable given the depth of the original crisis and therefore the difficulty
of meeting the adjustment conditions. Exchange rate adjustment provides
an illustration; the cedi was devalued from ¢2.75 to ¢90.0 to the U.S.
dollar in just three years, a change so drastic that it has seldom been
matched elsewhere. Resource mobilization is another example; the ratio
of revenue to GDP rose from 5.6 percent in 1983 to 13 percent in 1986,
another rare achievement. A summary of Ghana's record of compliance
with conditionalities is presented in table 15-3.
Implementation of loan conditions has generally been successful when
the government has endorsed the actions to be taken and when timetables
and indicators of progress have been clearly specified. Political difficulty
did not appear to impede progress when these two conditions were sat-
isfied. Devaluations and tax increases are clearly among the most polit-
ically sensitive matters, yet they were successfully implemented.
A lack of strong government endorsement has usually precluded sub-
sequent progress. In the case of the cocoa procurement system, for ex-
ample, the government consistently resisted the use of alternative mar-
keting mechanisms, although it allowed limited institutional reforms of
the existing monopoly. Proposals made by the World Bank under various
adjustment loans were modified along the way and appeared in much
weaker forms as legal conditions. The duty drawback scheme also lacked
strong government endorsement. Decisionmakers appeared not to fully



Ghana: Ending Chaos 325
appreciate its significance, while working-level officials were concerned
more with protecting revenue than with implementing the duty drawback
scheme. As a result, little was achieved in either of these areas.
A firm endorsement does not in itself guarantee successful implemen-
tation, however. Problems may arise because of technical difficulties (due
possibly to limited capacity) or the absence of any indicators of progress.
Inadequate control of monetary and credit aggregates is an important
example of a technical constraint to implementation. In the case of Ghana
the government was clearly committed to fighting inflation, but the thin-
ness of the bond market (for deficit financing) and the massive devalu-
ations that were undertaken impeded full implementation of monetary
and credit controls. Civil service compensation is another area in which
there was agreement on the merit of action but implementation was less
than satisfactory. The inadequacy of the data prevented the World Bank
from providing appropriate guidance on the size and timing of salary
increases. As a result, the pay scale implemented by the government in
1986, although consistent with the World Bank's condition, turned out
to be too costly and had to be canceled.
Sustainability
Since 1983 the government of Ghana has generally been receptive to the
World Bank's approach to economic management. Progress has been
substantial, and the economic gains have been widely shared. The gov-
ernment is likely to stay the course rather than revert to the policies that
led to the crisis. In this general sense the World Bank's approach, as
represented by its loan conditionality, appears to be sustainable. More-
over, through the adjustment loans, the World Bank has contributed to
institution building and an internalization of essential skills, which should
prevent some of the implementation problems and reversals observed in
the past.
The risks of reversal appear to be manageable. Unlike the case in many
countries where external debts represent a crushing burden, Ghana's
debts are of recent vintage and have been actively managed. Domestic
resistance to adjustment policies is limited to specific, uncoordinated in-
terest groups, with no threat of escalating to alarming proportions. The
economy's weak spots are mainly the continuing heavy dependence on
the cocoa sector and the slow emergence of new exports. Reforms are
also needed to prevent the financial sector from becoming a serious ob-
stacle to structural change and growth. In particular, it is necessary to
restructure the nonperforming assets of the banking sector and to build
up a system of prudential regulations.



326 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Tranching
The World Bank made extensive use of conditions for the release of
tranches in the four adjustment credits reviewed here and used the timing
of actions to achieve a variety of purposes. Under an early adjustment
loan prepared at the depth of the crisis, quick disbursement was deemed
essential, and many conditions were required before the appraisal. The
other early loans also had "up front" conditionality. When it would take
time to implement a particular condition (such as institutional reform,
training, and drafting of proposals) or when no government endorsement
was evident (as with retrenchment of the work force and increases in
utility rates), a tranche-release provision or dated covenant was used.
Conditions in this category were the most numerous. A large and in-
creasing number of actions were recommended but not mandatory. This
category covered measures that were desirable but not strictly essential
and measures that were essential but politically difficult to implement.
Soaal Implications of Adjustment
The adjustment program brought considerable gains to many tradition-
ally vulnerable groups. The largest group of beneficiaries was perhaps
the rural workers engaged in the production of traditional export items,
particularly cocoa and timber, who accounted for more than half the
work force. As producer prices were raised, real wages in rural areas
went up accordingly. The economic recovery also improved the circum-
stances of the rural population as the unemployed were rehired or found
new jobs. The working poor in urban areas benefited from large increases
in the minimum wage. In addition, improved resource mobilization by
the government enabled an increase in social services and improvement
of infrastructure.
The distribution of the economic gains of the adjustment program was
perhiaps uneven and was certainly not universal, however. People who
had had access to unrealistically cheap foreign exchange lost out under
the new regime. A large number of workers in the civil service and in
public enterprises were laid off, although the impact of the layoffs was
mitigated by severance payments and retraining. And the large propor-
tion of the population that remained outside the organized and monetized
economy was unable to share fully in the gains made under the adjustment
program. The government was aware of this problem and in 1987
launched a major poverty-alleviation effort with the support of the World
Bank (Program of Actions to Mitigate the Social Costs of Adjustment,
or PAMSCAD).



Ghana: Ending Chaos 327
Conclusions and Lessons
Several lessons may be derived from Ghana's experience with adjustment
and economic recovery.
First, for all the praise lavished on Ghana's performance, four years
of recovery have not brought the country back to where it was four years
before the crisis. Per capita income today is lower than it was at inde-
pendence three decades ago and far below the peak level reached a decade
earlier. But given the length and depth of the country's economic decline,
a long consolidation period may be needed. Confidence, once lost, is not
easily restored. Private sector responses to current policies may therefore
be slow although they may not yet have been exhausted.
Second, the experience of Ghana shows that fiscal restraint does not
necessarily rule out substantial recovery. Contrary to common appre-
hension, the economy rebounded at a time when government spending
was sharply reduced. Expenditure cuts often removed waste, as with the
termination of payments to "ghost" workers in civil service and state
enterprises.
Third, and also rather unsettling to conventional wisdom, fiscal control
in a developing country does not automatically produce monetary control
or price stability. In addition to fiscal control, other causes of credit and
monetary growth must be taken into account. In the case of Ghana, the
massive devaluations and the substantial primary price increases, which
increased the costs of state-owned marketing boards, appeared to have
contributed to the inflationary pressure.
Fourth, when external debt service obligations are small, devaluations
can provide incentives for the private sector and raise revenue for the
government at the same time. Of course, this outcome was made possible
in Ghana by its particular economic structure. The devaluations' stim-
ulatory effects on export production (in the primary sector) outweighed
their contractionary effects on imported inputs (in the industrial sector)
and consumer goods. Tariff revenue increased as the import bill grew in
terms of domestic currency. This result makes exchange rate adjustment
a more attractive instrument than tax increases.
Fifth, there are no clear indications that Ghana's experience can be
transferred to other countries. Some of the uncertainties arise from the
difficulties of making large exchange rate adjustments and sharp ex-
penditure cuts and mobilizing substantial revenue. The Ghanaians had
been through so much economic trouble for so long that they were willing
to put up with extraordinary hardship during the adjustment period.
Moreover, the Ghanaian leaders had an unusual strength of political will,
particularly during the first two years of adjustment when difficult mea-
sures had to be undertaken before any benefits of recovery were visible.
These two critical factors may not be present in other countries.



328 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Note
1. This excess borrowing from the central bank is a curious phenomenon.
Other instruments were also available for deficit financing (sale of government
bonds or borrowing from commercial banks, for example), and even if these
alternatives were not used the government could have limited its borrowing from
the central bank to an amount equal to the deficits. Why the government bor-
rowed so much more than it needed is an interesting question, but it lies beyond
the scope of this paper.
References
. Various years. International Financial Statistics. Washington, D.C.: Inter-
national Monetary Fund.
Roemer, Michael. 1984. "Ghana 1950-1980: Missed Opportunities." In A. C.
Harberger, World Economic Growth. San Francisco: Institute for Contem-
porary Studies.
Comments
Lyn Squire
CHAD LEECHOR has presented a very thorough survey of the adjustment
process in Ghana, and I am in broad agreement with his assessment of
the policy measures that have been taken. My comments go beyond his
analysis in two ways. First, I discuss some of the noneconomic factors
that lie behind the policy measures and that are essential for a successful
adjustment program. Second, I examine whether the adjustment process
in Ghana has yet reached the point of self-sustaining growth.
Two noneconomic factors are especially important: political leadership
and administrative capacity. Changing a country's basic approach to
economic development is a major task; maintaining commitment to a
new approach is an even more difficult one, while implementing the
program requires a certain level of administrative capacity.
Ghana's ability to chart a new course and stick to it has been re-
markable and is attributable mainly to the outstanding leadership of
Chairman Rawlings. Passive support for an adjustment program is not
sufficient-the aborted effort in Sierra Leone, for example, and the sub-
sequent failure to provide a new direction to economic policy are direct



Ghana: Ending Chaos 329
consequences of indecisive leadership. The Ghanaian government has
made its commitment to the program absolutely clear, and at least part
of its continued success must be attributed to the clear signals and un-
qualified endorsement of the reform effort provided by Ghana's top po-
litical leadership.
The second factor is the administrative capacity needed to implement
a program successfully. Managing an auction for foreign exchange, con-
ducting a review of public expenditures, and dismantling a system of
controls are just some of the tasks that place a heavy burden on the civil
service. In the case of Ghana, the civil service had undoubtedly suffered
as a result of the economic decline of the 1970s-salaries had fallen
drastically in real terms, morale was low, and retaining qualified staff
proved difficult. In this respect Ghana, like many other countries in Af-
rica, was probably not well placed to implement a reform program. And,
indeed, part of the program now includes reform of the civil service.
Nevertheless, Ghana was able to overcome these difficulties by drawing
on the skills and dedication of a few key members of government, whose
energy and enthusiasm kept a large program on track. Although stretched
to the limit, administrative capacity proved adequate in the end.
The importance of such a core team also emerges from the experience
of other countries. Guinea-Bissau's reform effort, for example, has ben-
efited from the presence of an economic team to implement its reform
program. In Nigeria, by contrast, despite its more developed adminis-
trative capacity, progress has been less smooth in the absence of a well-
knit team responsible for implementing the program.
In reviewing the Ghanaian experience, therefore, I see two important
lessons for other countries. First, passive support for adjustment is not
sufficient. The demonstrated commitment of the political leadership is
essential. Second, potential problems arising from a relatively weak ad-
ministrative capacity can be overcome provided there is a well-coordi-
nated economic core team with clear responsibility for implementation.
Next I would like to assess whether Ghana has yet achieved self-sus-
taining growth. For the purposes of this discussion, I assume that self-
sustaining growth requires a rate of investment of at least 20 percent of
GDP and a rate of national savings of at least 15 percent. These numbers
are based on the experience of fourteen countries in Sub-Saharan Africa
that achieved an average rate of growth in GDP of 4 percent or more a
year for at least four years. In 1988 Ghana was investing only 12 percent
of GDP and had a national savings rate of only 6 percent, while its average
annual GDP growth rate in 1985-88 was 5.4 percent. The implication of
these figures is that Ghana's excellent growth performance in recent years
can be attributed, at least in part, to better utilization of the existing
stock of capital and to a significant inflow of concessional assistance well
above the Sub-Saharan average.



330 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
As an illustration of the importance of these factors, consider the effect
of a smaller inflow of concessional assistance. Assuming that concessional
assistance to Ghana grew at the same rate as it did in the rest of Sub-
Saharan Africa and that consumption and the incremental capital-output
ratio (ICOR) remained unchanged, GDP growth in 1985-88 would have
averaged 4.4 percent a year compared with an observed rate of 5.4 per-
cent, that is, a reduction in the growth rate of one percentage point.
Clearly, this is a very rough calculation, but it indicates the role played
by concessional assistance. This calculation, together with the advantage
provided by a temporarily low iCOR, suggests that Ghana's recent high
growth rate is not an indication that self-sustaining growth has been
realized.
What are the prospects for achieving self-sustaining growth? In the
immediate future Ghana will have to increase its investment rate to offset
a rising ICOR. In addition, since foreign savings are unlikely to constitute
as large a share of total savings as they have in the recent past, financing
that higher rate of investment will require an increased rate of national
savings. According to the World Bank's most recent set of projections,
Ghana's investment and national savings rates will rise to my minimum
required levels of 20 and 15 percent, respectively, by 1994. Compared
with current levels, these are ambitious but not impossible targets, so
self-sustaining growth seems feasible.
What is critical to the achievement of these targets, however, is the
significant distributional change that occurs within aggregate investment
and savings. For both investment and savings the role of the private sector
is expected to increase relative to that of the public sector. Consider the
case of investment. Although the urgent need to restore the economic
and social infrastructure implied a prominent role for the public sector
in capital formation during the first phase of the reform program, the
engine of growth and hence the main source of investment in the long
run is expected to be the private sector. This, in fact, is an important
objective of the adjustment program. But private sector investment has
been very slow to respond to the new incentive structure, and at present
it accounts for only about one-third of total investment. By 1995 it will
have to account for nearly half of a much larger total.
The recent disappointing performance in Ghana notwithstanding,
there are signs of improvement: cocoa is being replanted, investment has
picked up in the gold and forestry sectors, and foreign investors are
beginning to explore possibilities in manufacturing. Nevertheless, the
basic lesson for other countries is that the response of private investors
is likely to take some time to emerge, especially if the previous policy
regime was clearly opposed to the private sector. For Ghana, the critical
issue is whether current indications of an upturn will indeed develop into



Ghana: Ending Chaos 331
a period of growth led by the private sector. The evidence of the next
two years will determine the final assessment of Ghana's reform program.
The issue is similar for savings. The rate of private savings is projected
to increase from about 3 percent of GDP in 1988 to almost 9 percent by
1995. Whether current efforts to restructure the banking system, liber-
alize interest rates, and mobilize deposits will be sufficient to achieve the
desired rate of savings remains to be seen.
These remarks lead to an important general conclusion. For countries
such as Ghana that have successfully completed the first stage of ad-
justment, the determinants of private savings and investment become
increasingly important. But our understanding of investment and savings
behavior in developing countries in general and in Africa in particular
is weak. Assuming that more and more countries will be reaching this
second stage, the World Bank should make sure that it is in a position
to offer the best possible policy advice on these matters. The important
lesson for the World Bank, therefore, is that the determinants of private
savings and investment should be seen as a priority area for economic
and sector work and for the research program.



16 Zambia: Reform and Reversal
Mohsen A. Fardi
THIS CHAPrER EXAMINES Zambia's attempts at stabilization and structural
adjustment during the 1980s, although the country's efforts in this di-
rection go back to at least 1972 when it obtained a program loan from
the World Bank. The first section summarizes the causes and origins of
the economic crisis in Zambia and early efforts at adjustment. The fol-
lowing sections discuss the policy content of recent stabilization and ad-
justment operations, the outcome of these programs, and the extent of
compliance with loan conditions. Conclusions and lessons derived from
Zambia's experience are presented at the end of the chapter.
Causes and Origins of the Crisis
Zambia began its structural adjustment program after years of low to
negative economic growth rates. Average annual real GDP growth rates
before 1980 (2.4 percent during 1965-72 and 0.3 percent during 1973-
80) were much below the average for developing countries. GDP in Zam-
bia declined an average of 0.1 percent a year from 1980 through the
years of the stabilization and adjustment programs.
The slight growth before 1978 resulted from copper production and
exports, which increased annually by a rate of about 2.6 percent, and
agriculture, which grew by less than 2 percent annually. The govern-
ment's agricultural policies, including price controls, food subsidies, and
inadequate export incentives, have prevented the country from achieving
its potential for agricultural development, which is considered quite high
since less than a quarter of arable land is now under cultivation Uansen
1988).
Zambia's export earnings come almost exclusively (95 percent) from
copper. Since 1976, export earnings have been declining continuously
primarily because of falling copper reserves and production and, until
1987, declining copper prices. Copper production and export were hin-
dered by inadequate maintenance, a shortage of foreign exchange, man-
agement inefficiency, and transport difficulties. Since Zambia's produc-
tion costs for copper are among the highest in the world, the decline in
copper prices has had an immediate impact on the profitability of the
industry and on Zambia's terms of trade.' The deterioration in the terms
332



Zambia: Reform and Reversal 333
of trade also contributed to a large decline in national income and to the
depressed state of the economy. The terms of trade index declined by 72
percent during 1969-80 (from 353 to 100).
Although the terms of trade were deteriorating, the exchange rate was
not allowed to adjust, which led to an effective real appreciation of the
domestic currency. The fixed exchange rate was maintained through an
import licensing system that was based on a priority list of goods and
services. Despite government efforts to control imports in line with avail-
able foreign exchange, commercial payments began to be in arrears in
1975. As proceeds from copper exports declined and foreign exchange
became more scarce, the effective exchange rate became more overvalued,
restrictions on import licenses became more severe, and arrears increased.
The difference between the official and the black market exchange rates
widened, inefficiencies in the distribution of scarce foreign exchange in-
creased, and graft and corruption emerged.
Consumer prices and nominal interest rates were also subject to gov-
ernment control until 1984. The average annual increase in consumer
prices was 14.5 percent during 1973-80. The interest rate on bank de-
posits ranged between 4 and 7 percent during 1965-83.
After independence the Zambian government took over a number of
key enterprises that had been established by foreign interests, including
the copper mining companies. The government also took an active role
in the economy by investing in import-substituting manufacturing en-
terprises during the copper boom years and by increasing recurrent ex-
penditure to expand programs for free public health and educational
services and to provide consumer subsidies. In 1982-83 the public sector
accounted for about 72 percent of formal sector employment (World
Bank 1987b). Public expenditure ranged between 10 and 25 percent of
GDP during 1975-82.
Until recently Zambia continued to import more than it exported by
borrowing externally and by accumulating arrears. On the strength of
its future earnings from copper exports, the country was able to attract
finance from commercial and official bilateral and multilateral sources.
With the exception of 1979, Zambia has had a significant current account
deficit every year since 1975. By 1980 the current account deficit was
14 percent of GDP.
To sum up, the roots of the economic crisis facing Zambia in the 1980s
lay in the need for a rigidly structured economy to adjust to a rapid
decline in per capita income because of a rapid decline in the terms of
trade. Maintaining a fixed rate of exchange, low agricultural prices, and
extensive price controls created a strong bias against noncopper exports
and increased the import intensity and capital intensity of domestic pro-
duction. The public sector was overextended because of the government's
unsustainable support of a large number of public enterprises and its



334 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
effort to provide free public services for a rapidly growing urban pop-
ulation. As copper exports and the national economy declined, the gov-
ernment relied on large-scale foreign borrowing, including International
Monetary Fund (IMF) resources, and the growth in arrears to stave off
the need for retrenchment and restructuring. Since structural adjustment
had to take place eventually, this delay simply made it even more difficult.
The Adjustment Program
The stabilization and adjustment program in Zambia had three parts:
an IMF-supported program, a World Bank-supported program, and debt
relief through reschedulings of debt by official and private creditors.
The IMF-Supported Program
Zambia has received the support of the IMF several times since 1976. A
review of these arrangements shows an unchanging pattern of objectives
and an unwavering optimism in expecting a resurgence in copper prices
and exports and a turnaround in government management of its fiscal
affairs.
The imF-supported programs endeavored to reduce the unsustainable
deficits in the fiscal budget and the current account and to create an
environment favorable to the development and diversification of exports
and a reduction in inflationary pressures. The policy measures that were
to bring about the desired changes were depreciation of the exchange
rate, unification of tariff rates, a reduction in restrictive economic and
trade regulations, an increase in interest rates, removal of price controls
and consumer subsidies, limitations on domestic financing of the fiscal
deficit, an increase in tax revenue, and limitations on any increase in the
public sector wage bill.
Despite some improvements in the policy environment, almost all
agreements with the IMF were suspended or canceled because of an in-
ability to meet the targets for the current account and fiscal budget def-
icits, the accumulation of external arrears, and insufficient control over
domestic credit. These repeated interruptions, suspensions, and cancel-
lations severely reduced the credibility of the stabilization program. The
government appears to have lacked sufficient control over the program
to maintain a consistent policy course.
The World Bank-Supported Program
Although the objectives supported by the IMF and the World Bank over-
lapped, the IMF-supported program focused on improving macroeco-
nomic stability, while the World Bank's support initially had a sectoral



Zambia: Reform and Reversal 33S
orientation. As management of the macroeconomic policies became more
critical, however, the World Bank-supported program focused more on
issues such as the foreign exchange auction market.
In January 1983 the government committed itself to three objectives
that would set the stage for renewed expansion: (1) adequate economic
incentives for the production of agricultural and manufactured exports,
(2) increased international competitiveness of the economy, and (3) the
allocation of resources in line with the first two objectives. Annual GDP
growth was projected at 2 to 2.5 percent for the next ten years, which
was an implicit acceptance of a continued decline in per capita income.
The World Bank supported Zambia's adjustment program by granting
a sectoral adjustment loan for rehabilitating the mining sector and two
sectoral adjustment credits for rehabilitating the agricultural and man-
ufacturing sectors. Growth in these sectors was expected to develop new
exports, reverse economic decline, and create the basis for sustained long-
term growth.2 The sectoral adjustment loans were to finance imports
required for the rehabilitation of the sectors concerned. None of these
operations was designed as quick-disbursing support for the balance of
payments. As economic conditions deteriorated and the need for policy
reform became more urgent, however, these operations acquired the
quick-disbursement feature. A fourth operation, the Economic Recovery
Credit, was designed to reinforce policy measures adopted under the
previous three sectoral operations. The adjustment strategy was based
on an expected increase in copper exports and in funds from external
sources.3
Structural Adjustment Policies
FIsCAL POLICY. The government's critical tasks were to reduce public
sector expenditures, to increase public revenue, and to limit financing of
the budget deficit to the amount of net external borrowing. Tight fiscal
and monetary policies, accompanied by positive interest rates and de-
preciation of the exchange rate, were expected to reduce the size of ag-
gregate demand (expenditure reduction) and change the orientation of
the economy toward tradable goods (expenditure switching). Expendi-
tures were to be reduced by cutting subsidies and imposing limits on
public sector employment and wage hikes. Revenue was to be increased
through the mineral export tax, rationalization of import tariff rates, and
new excise taxes.
Beginning with the 1984 budget, recurrent expenditures were tied to
each ministry's development priorities. Allocations for agriculture were
increased from an average of 6 percent of total public expenditure in the
years before 1984 to 9 percent in 1984. Priority was placed on finishing
existing projects in line with development objectives. Representatives of



336 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
external lending agencies were invited to review the government's plans
and the status of ongoing projects. The World Bank conducted a review
of public expenditure in 1986 and made recommendations that the Zam-
bian government accepted in principle. Preparations were made to im-
plement a three-year rolling fiscal plan, and a study of the tax structure
was undertaken with World Bank and IMF assistance. The recommen-
dations emanating from the public expenditure review and the tax study
were to be implemented in the 1987 and 1988 budgets.
MONETARY POLICY. A tight monetary policy to reduce inflation was
an important element of the reform program. One condition for the re-
lease of the second tranche of the Industrial Reorientation Project was
decontrol of interest rates. Accordingly, the government introduced a
daily auction market for treasury bills in September 1985, which led to
an immediate increase in the open market rate from 9.5 percent to 24
percent. Commercial bank lending and time deposit rates were also
raised.
During 1983-85 the growth of the money supply was controlled and
the money supply contracted relative to GDP. In 1986, however, the fi-
nancing of large budget deficits through domestic borrowing led to a 66
percent increase in the money supply, excess liquidity, and rapid depre-
ciation of the kwacha. Although the commercial bank rate was increased
to 30 percent in 1986, it was still much below the inflation rate (the
wholesale price index increased 64 percent in 1985 and 74 percent in
1986). In the absence of sufficient control over the money supply and
complementary fiscal adjustment, the interest rate remained negative in
real terms.
THE EXCHANGE RATE SYSTEM. Before 1985 Zambia followed a fixed
exchange rate system with occasional adjustments. From 1976 to July
1983 the kwacha was devalued three times: 20 percent in 1976, 10 per-
cent in 1978, and 20 percent in January 1983. In July 1983 Zambia
adopted an adjustable rate system pegged to the trade-weighted average
value of the currencies of five major trading partners. After considerable
examination of alternatives, the government settled on an auction market
to allocate foreign exchange and to establish a more realistic exchange
rate. With this mechanism the government also hoped to avoid the mis-
allocation of scarce foreign exchange and to eliminate the graft associated
with the existing administrative system.
Despite some reservations about the feasibility of an auction market
for Zambia, the prevailing view within the World Bank was that the
auction system would be the most efficient way to allocate scarce foreign
exchange, would provide guidance on the real value of the currency, and
would serve as an incentive to exporters. Zambia's aid consortium also



Zambia: Reform and Reversal 337
strongly endorsed the introduction of the auction system, although some
members were concerned about its effects on equity and domestic prices.
When the auction market opened on October 1, 1985, the value of
the kwacha fell as was expected, from 2.2 to S kwacha to the U.S. dollar.
During the next eight months the value depreciated gradually to 8 kwacha
to the U.S. dollar, but in the second half of 1986 it fell rapidly to 19
kwacha to the U.S. dollar, which was equal to a real effective depreciation
of 60 percent of its value in 1980. Government attempts to stabilize the
exchange rate by increasing restrictions or by increasing the amounts
available for auction failed to arrest the rapid depreciation. Accumulating
arrears and delays of up to ten weeks in delivering foreign exchange led
to a loss of confidence in the system. By the end of April 1987, when
the kwacha had depreciated to 22 kwacha to the U.S. dollar, the auction
system was abandoned and a fixed exchange rate regime was reinstated.
TRADE POLICY. With the introduction of the foreign exchange auction,
import licensing and import prohibitions that had been established for
protective purposes were eliminated. Instead, a registration system was
put into effect to assist with the management of external payments. As
part of a tariff reform, import duties on industrial capital equipment
were raised to 15 percent and a minimum tariff of 10 percent was imposed
on previously duty-free imports of intermediate goods. Foreign trade
regulations were streamlined by eliminating exemptions and rebates of
import duties on a large number of inputs. A uniform sales tax of 15
percent was imposed on domestically produced goods to reduce the ef-
fective protection of domestic production. Duty drawbacks were sim-
plified for exporters.
PRICES AND INCOMES POLICY. Until December 1982 the prices of es-
sential consumer goods were subject to government control at wholesale
and retail levels. This practice led to inefficiencies in production and
distribution, disincentives to expand output, and the emergence of black
markets and smuggling in the border areas. In late 1982 it was decided
to set the retail prices for most consumer goods in line with border prices.
As a result, the prices of most consumer goods were decontrolled, al-
though changes in the prices of some goods remained subject to govern-
ment approval.
The extent of the price liberalization became apparent after the opening
of the foreign exchange auction market. The depreciation of the exchange
rate had an immediate and significant impact on domestic prices. The
price of refined petroleum products jumped by 150 percent, public utility
rates by 80 percent, bread by 100 percent, sugar by 49 percent, maize
by 40 to 50 percent, and fertilizer by 77 percent. For some services, such



338 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
as telecommunications, that were denominated in U.S. dollars, prices
increased by the full extent of the exchange rate depreciation.
The government tried to control aggregate demand by imposing limits
on public employment and wages. Nominal wages were increased by less
than 10 percent annually, well below the inflation rate. The freeze on
employment was not very effective, however, because the hiring of tem-
porary workers and contract employees continued to expand overall pub-
lic employment.
DEBT RELIEF. Zambia's debt is one of the highest in the world. Debt
per capita in 1986 was US$840, or 3.6 times GNP and 7.2 times exports
of goods and services. Zambia accumulated its foreign debt when it was
considered a middle-income country, so the terms were too hard for it
to meet in the 1980s. A high proportion of Zambia's foreign debt is short
term, nonconcessional, and not subject to rescheduling; about 20 percent
of outstanding debt has been rescheduled at least once. Given the large
share of debt not subject to rescheduling and Zambia's high debt service
obligations, rescheduling has provided only short-term relief. Although
rescheduling lowered debt service in absolute amounts, external debt and
debt to the iMF increased from US$3.9 billion in 1984 to US$6.4 billion
at the end of 1987. By 1985 the scheduled debt service exceeded 80
percent of exports of goods and services. Zambia could not meet its
obligations under the rescheduling arrangements reached in 1983-84 and
1986, and arrears continued to accumulate, growing to twice the value
of exports at the end of 1987.
Economic Performance
GDP, which had grown by only 0.3 percent a year during the 1970s, did
not respond positively to the policy changes of the 1980s. Iri fact, as a
result of continued deterioration in the terms of trade, gross national
income declined by 2.4 percent annually during 1980-86, with most of
the decline occurring in 1986 when gross national income fell by 17
percent. (Table 16-1 shows changes in macroeconomic indicators for
Zambia from 1965 to 1987.)
Domestic supply responded much more slowly than expected to the
incentives provided by the price and exchange rate adjustments. The
volume of copper production continued to decline despite efforts to re-
habilitate and improve management of the industry. Refined copper pro-
duction declined from 610,000 tons in 1980 to 460,000 tons in 1986,
while the price fell by 48 percent in nominal terms (World Bank 1986
and 1988). Despite optimistic expectations of an imminent reversal in
the price trend, world copper prices continued to decline. Capacity uti-
lization in manufacturing remained low because shortages of foreign ex-



Zambia: Reform and Reversal 339
Table 16-1. Macroeconomic Indicators for Zambia, 1965-73 to 1987
(average annual growth rates based on constant prices)
Indicator                1965-73    1973-80    1980-86       1986      1987
Gross domestic product      2.4          0.3      -0.1         0.4    -0.2
Gross domestic income       2.2        -6.0        -0.3      -4.7        3.8
Gross national product      2.4          0.8      -2.0      -11.0        1.3
Gross national income       2.2        -6.0        -2.4     -17.1        6.5
Agriculture                  2.0         1.6        3.1        9.0     - 1.0
Industry                     2.7       -0.3        - 0.8     -1.6        0.7
Manufacturing              9.8         0.5        0.6        0.4       0.6
Services                    2.3          0.4       -0.6      -0.9      -0.6
Total consumption            1.1         1.9      -0.3       -1.7        4.3
Private consumption        -1.9          3.1        1.0      -4.6        6.1
General government         10.4        -0.6        -3.3        6.3       0.0
Gross domestic
investment                6.2       -16.5        - 8.2       6.0    -33.8
Fixed investment            n.a.      -15.4       -11.6      - 8.8      17.3
Total expenditures           3.0       -4.1        -1.8      -0.6      -1.5
Gross domestic savings      3.2       -18.8         1.3    -20.4         0.6
Exports'                    2.5        -0.1       -4.4         8.0      11.3
Imports'                    3.4        -8.5       -9.7      -13.2        9.1
Capacity to import          2.1       -11.2       -5.7      -26.3       33.7
n.a. Not available.
a. Goods and nonfactor services.
Source: World Bank IEC data base, November 1988.
change limited access to needed imports. Value added in the industrial
sector declined during 1980-86. The basic problem in the industrial sec-
tor remained the insufficient participation of the private sector. The only
relatively bright spot was agricultural production, which was helped by
higher producer prices and favorable weather.
Internal Balance
The government's efforts to establish internal balance were only partially
effective in the first phase of the adjustment program. Although con-
sumption in the public sector was reduced in real terms from 1980 to
1986 (annual real rate of change of -3.3 percent), general government
expenditure as a share of GDP remained at around 25 percent.4
With the rapid decline in per capita income and the strong resistance
of urban consumers to reducing their consumption, private consumption
increased by 1 percent annually in real terms during 1980-86 and its
share in GDP increased from 55 percent to 66 percent. Thus the burden
of adjustment in domestic absorption fell primarily on domestic invest-



340 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
ment, which declined by 8.2 percent a year (on a gross basis) during
1980-86 (after falling an average of 16.5 percent a year during 1973-
80). The fall was precipitated by the rapid economic decline, the existence
of excess capacity, the shortage of foreign exchange for imports of capital
goods, and the falling rate of private and public savings.
External Trade Balance
Exports continued to decline during the adjustment years, mainly because
of the decline in copper exports. Although exports of nontraditional
goods increased somewhat, the amounts remained insignificant. As a
result, merchandise exports declined by 53 percent (from US$1,457 mil-
lion in 1980 to US$681 million in 1986). Imports of goods and nonfactor
services declined from US$1,765 million in 1980 to US$809 million in
1986-an annual decline of 9.7 percent on average (26.3 percent in
1986), which came on top of an average decline of 8.5 percent during
the 1970s. This drastic cut in imports aggravated an already very difficult
economic situation.
Despite the shortfall in exports, the trade balance improved from a
deficit of US$157 million in 1980 to a surplus of US$20 million in 1986.
However, the current accounts deficit, which had improved from US$761
million (19 percent of GDP) in 1981 to US$153 million (5.5 percent of
GDP) in 1984, increased to US$345 million (21 percent of GDP) in 1986.5
Public Revenue and Expenditure
Between 1981 and 1985 total government expenditure declined in real
terms by 18 percent. All expenditure categories except interest payments
declined: capital expenditures by 30 percent, wages and salaries by 16
percent, and subsidies by 25 percent. The budget deficit also was cut by
18 percent during that period. Total revenue increased from 23.5 percent
of GDP in 1984 to 26.5 percent of GDP in 1986. The depreciation of the
kwacha had a positive effect on government revenue from foreign trade
(mineral export tax, import duties, and excise taxes). Rescheduling of
the major part of the interest due on external debt in 1983 and 1984
helped to limit interest payments on external debt to about 1.3 percent
of GDP and domestic finance of the deficit to about 3 percent of GDP.
Inflation, although relatively high at about 20 percent a year, was not
alarming and was considered a temporary result of the lifting of price
controls and reductions in consumer subsidies. Thus it appears that until
the end of 1985 the government succeeded in containing the budget deficit
despite some shortcomings in its approach.6 From about the fourth
quarter of 1985 until the adjustment program unraveled in 1987, the
government went through a severe fiscal crisis and ultimately lost the



Zambia: Reform and Reversal 341
Table 16-2. Central Government Budget as a Percentage of GDP in
Zambia, 1981-86
(current prices)
Item                  1981      1982    1983    1984    1985         1986
Revenue               24.45     24.14   25.57   23.54    22.41      26.50
Expenditure           39.84     40.90   34.49   32.87    36.86      63.83
Current             35.30     34.56   29.75   27.48    33.24      55.79
Interest           3.38      4.16    4.64     4.38      9.69    19.34
Foreign           1.35     2.07    1.11     1.46      6.11    13.78a
Domestic         2.03      2.09    3.53     2.93      3.58     5.56
Wages and salaries   8.02      8.13    7.25     6.96      7.61     5.46
Subsidies            3.17      3.49    1.97     1.80     2.67      4.71
Other               22.76     20.87   19.43    17.26    13.26     26.27b
Capital              4.54      6.34    4.74     5.39     3.62      8.05
Overall deficit     -15.39   - 16.76   -8.92   - 9.33   -14.44   -37.33
Source of finance
Foreign              7.98      3.12    2.42     1.20     6.03     21.81
Domestic             4.82     13.60    2.75     3.83     8.55     15.52
a. Interest on external debt in 1986 on accrual basis.
b. "Other" expenditure in 1986 includes the following:
Percentage of GDP
Transfers for 'MF charges                                         6.8
Transfers for foreign exchange losses                             3.8
Transfers and pensions                                            3.3
Recurrent departmental charges                                    4.7
Other constitutional and statutory expenses                       7.2
Source: World Bank data.
gains made in earlier years. During this period the government lost con-
trol of the budget. As shown in table 16-2, the overall deficit, on an
accrual basis, increased from about 9 percent of GDP in 1983 and 1984,
the two years preceding the reform program, to 37 percent in 1986.
Although a more elaborate analysis is necessary to establish causality, it
can be observed that the jump in interest rates and the depreciation of
the exchange rate were concurrent with unprecedented increases in
wholesale and retail prices, the level of public expenditure, the budget
deficit, and the money supply.
Whether the government had the will or the latitude to reduce its dis-
cretionary expenditures to offset the impact of increased interest and
exchange rates is debatable. Given the protracted decline in national
income and real public expenditures, and the large budget share of non-
discretionary expenditures, there was little that the government could do
in the short term to reduce the fiscal deficit. Since the transfer of external
resources was insufficient, the fiscal deficit was accommodated by ex-



342 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
panding the money supply and accumulating large external and domestic
arrears.
Net Flow of External Resources
Was the net inflow of external resources in support of Zambia's stabi-
lization and adjustment program sufficient, particularly in light of the
continuous decline in export earnings? Could a larger inflow of external
resources have reduced the fall in imports, domestic investment, and the
growth of per capita income?
Although definitive answers to these questions cannot be provided in
the context of this review, data do show a wide gap between planned
and actual transfers of external resources. The adjustment program was
based on expected improvements in copper prices and export volume
that never materialized. Yet even though export performance consistently
fell short of expectations, the program's assumptions and design were
never reexamined. Underlying this inaction was an overoptimistic belief
in Zambia's ability to export, based on expectations of high copper prices.
Four major sources of external finance were expected when the pro-
gram was designed: earnings from copper exports, bilateral assistance,
IMF standby arrangements, and adjustment credit from the World Bank.
All four fell severely short of expectations. Exports fell repeatedly below
even the most conservative estimates. Multilateral donors were unable
to provide sufficient resources to help Zambia, despite assurances that,
unlike in the past, the resources would be put to good use. Zambia could
not draw on the IMF'S facilities because it had not met its previous com-
mitments and was in arrears as well. In addition, because Zambia had
no standby agreement with the IMF, debt rescheduling could not take
place in 1985. As a result, accumulated arrears reached US$586 million
in 1985, including US$125 million to the IMF.
Net transfers from all long-term sources declined from a peak of
US$450 million in 1981 to US$1 million in 1987 (table 16-3). Net trans-
fers from the World Bank were negative until 1985; in that year and the
following, however, sectoral adjustment credits from the International
Development Association (IDA) increased net transfers from the World
Bank to an average of US$63 million.7 The World Bank also accelerated
disbursements from the existing sectoral adjustment operations in light
of Zambia's liquidity crisis. Bridge finance from commercial banks was
also sought in 1985 to pay Zambia's arrears to the IMF and so to pave
the way for a new standby agreement. All these efforts were insufficient,
however, and the external finance gap widened rapidly.
Assessment of the Program
Zambia's restructuring program began in 1983 and lasted for about four
years. Implementation began to suffer after April 1986 when the eco-



Zambia: Reform and Reversal  343
Table 16-3. Net Transfers to Zambia from All Sources, 1980-87
(millions of U.S. dollars)
Source           1980   1981    1982    1983   1984   1985   1986   1987
World Bank'     -19.3  -22.9    -15.9  -19.2  -12.8    66.6   62.1   22.0
IBRD          -20.8  -25.0   -23.3  -28.4  -28.0         4.0  -15.7  -19.0
IDA              1.5    2.1       7.4    9.2   15.2   62.6   77.8   41.0
IMF             -18.2   336.9  -111.4    10.7   21.9  -43.4  -98.0    n.a.
Purchases       65.1  423.7      37.5  185.7  151.2    0.0  121.8    n.a.
Repurchases     57.3   57.0      95.2  121.4   73.0   19.0  143.9    n.a.
Net purchases    7.8  366.7   -57.7    64.3   78.2  -19.0  -22.1        n.a.
Interest        26.0   29.8      53.7   53.6   56.3   24.4   75.9    n.a.
Other sources    360.9  136.1    188.1   81.2  161.8  140.4  104.2  -21.0
Net transfers    323.4  450.1      60.8   72.7  170.9  163.6   68.3    1.0
n.a. Not available.
a. The World Bank includes both the International Bank for Reconstruction and Development
(IBRD) and the International Development Association (IDA).
Source: World Bank (1987a) and the World Bank World Debt Tables data base.
nomic management team was changed and critics of the reform program
were put in charge of implementing it. They continued efforts to imple-
ment some elements until May 1, 1987, when the program was abruptly
abandoned.
During the adjustment period Zambia adopted policies for rational-
izing the trade and exchange rate regimes and for rehabilitating the min-
ing sector. Consumer prices were decontrolled and agricultural producer
prices were raised. A substantial real depreciation in the effective rate of
exchange took place. Interest rates were raised in nominal terms although
real rates remained negative.
Scope and Implementation of Conditionality
Zambia received three sectoral adjustment loans and a structural ad-
justment credit from the World Bank. An export rehabilitation project
in 1984 was designed to halt the decline in copper production and en-
hance the efficiency and profitability of the mining sector. Among the
policy conditions to be met prior to the release of the second tranche
were satisfactory progress in carrying out production and investment
plans for the government-owned mining company, allocation of the for-
eign exchange needed to rehabilitate the company, and establishment of
a uniform tariff system with a minimum and maximum rate. Although
foreign exchange allocations fell 30 percent short of the government's
commitments, progress in other areas was deemed sufficient to allow the
release of the second tranche.



344 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
An agricultural rehabilitation project in 1985 financed imports of farm
implements and consultant services. Among the conditions for release of
the second tranche were full cost recovery for state enterprises that were
marketing maize and fertilizer within a regionally differentiated price
structure, improved efficiency of the enterprises, adoption of a mecha-
nism for linking producer prices to border prices, and the opening of
maize and fertilizer marketing to private traders.
The second tranche was released although compliance was only partial:
a study of fertilizer usage was completed, a study of maize and fertilizer
marketing was begun, and maize and fertilizer subsidies were reduced in
accordance with a three-year plan for phased reduction. After the adop-
tion of a foreign exchange auction, disbursement procedures for this
project were modified so that the foreign exchange provided for the pur-
chase of eligible imports was sold in the auction market. Although no
data are available on individual sales, bids by farmers and suppliers of
agricultural machinery were generally not successful.
Another sectoral loan in 1985 supported the auction market and in-
dustrial and trade policy reforms. Conditions for disbursement included
adoption of the foreign exchange auction system, conversion from an
import licensing to an import registration system, elimination of import
prohibitions, and a reduction in maximum tariff rates from 150 to 100
percent. Conditions to be met before release of the second tranche in-
cluded decontrol of interest rates, enactment of a new investment code,
establishment of a minimum 10 percent tariff on remaining duty-free
items, simplification of the export duty drawback system, and imple-
mentation of an action program to improve forecasting and budgeting
for foreign exchange.
Again, the second tranche was released. The government attempted to
comply with loan conditions by introducing a daily auction for treasury
bills, which led to an increase in nominal interest rates although real rates
remained negative; passing a new investment act (with parliamentary
approval but not the presidential signature); rationalizing import tariffs;
and simplifying duty drawback procedures.
In contrast to the three sectoral loans, the Recovery Program Credit
of 1986 was a quick-disbursing policy-based credit. It was designed to
provide the minimum level of imports needed for Zambia's recovery
program. When the credit agreement was presented to the World Bank's
Executive Directors in June 1986, four major problems were observed.
First, the deficit in the current account was larger than agreed to because
of falling exports, shortfalls in aid flows, slow disbursement of existing
bilateral commitments, and larger than expected sales in the foreign ex-
change auction. Second, the government had failed to control the money
supply. Third, the budget deficit was much larger than expected because
of revenue shortfalls (resulting in part from unexpectedly low external



Zambia: Reform and Reversal 345
Table 16-4. Number of Major Policy Conditions in Zambia's
Adjustment Operations by Policy Area
Adjustment operation           All
Policy area          Export  Agricultural  Industrial  Economic  programs
Trade and foreign      0         0          8         4        12
exchange
Agriculture            1         5          0         5        11
Public enterprise reform  0      0          5         1         6
Mining                 3         0          0         2         5
Fiscal                 0         1          0         4         5
Monetary               0         0          1         0         1
Social impact          0         0          0         1         1
Total                  4         6         14        17        41
Source: World Bank data.
aid and growing losses among public enterprises arising from changes in
the exchange rate). Fourth, the government was unable to handle the
economic management of a tough stabilization and adjustment program.
This situation, which had already jeopardized Zambia's standby agree-
ment with the IMF, led the World Bank to postpone signing the credit
agreement until Zambia reestablished its eligibility for drawings under
the standby arrangement. As a result, the Recovery Program Credit did
not become effective until December 1986, and the second tranche was
never released because of the policy reversals after the first tranche. The
credit was canceled when the government abandoned the adjustment
program in May 1987.
Although some marginal gains were achieved, little if any significant
structural change took place in the Zambian economy during the rela-
tively brief period in which adjustment policies were in effect.8 Despite
this lack of real structural reform, however, Zambia's progress in car-
rying out its policy commitments was deemed sufficient to permit release
of the second tranche for the three sectoral adjustment operations. In
retrospect, it appears that the World Bank may have gone too far in
accommodating Zambia with the maximum amount of resources avail-
able, perhaps influenced by the Bank's strong desire to see Zambia suc-
ceed in its enormous adjustment efforts.9 Table 16-4 shows the number
of policy conditions in each major policy area for the four adjustment
operations.
Sustainability and Policy Reversals
Overall, the adjustment program in Zambia was not sustainable and was
reversed in late 1986 and early 1987. Although some controls were lifted



346  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
from the exchange rate, the interest rate, and major producer and many
consumer prices, high inflation during 1985 and 1986 nullified the eco-
nomic impact of the price changes. Furthermore, the price liberalization
measures were not institutionalized, and so prices remained subject to
overt or effective administrative controls. At the end of 1986 the foreign
exchange system was still subject to administrative controls and the auc-
tion market lacked stability, fuel prices did not reflect import costs and
remained subject to government control, the price of fertilizer was below
domestic cost, and interest rates remained negative in real terms.
Under some pressure from external lenders and with little planning,
the government in December 1986 removed the subsidy on high-quality
maize meal while maintaining it on the lower-quality maize consumed
mainly by low-income groups. Although the objective was to target the
subsidy to the poorer groups, the program did not work because the
subsidized maize immediately disappeared from the shops. Riots and civil
unrest followed in urban and mining areas. In response, the government
reinstated the maize subsidy and took over the milling plants. Similarly,
an attempt to raise fuel prices by 70 percent on April 20, 1987, collapsed
in response to riots and protests.
Since abandoning the reform program on May 1, 1987, the govern-
ment has tried to implement its own restructuring program, which has
the spirit of the previous reform program but little of its content. It
abandoned the foreign exchange auction market and reinstated the ad-
ministered allocation of foreign exchange and a fixed exchange rate. The
exchange rate was initially set at 8 kwacha to the U.S. dollar and adjusted
to 13 kwacha in October 1988 (the final auction rate had been 21 kwacha
to the U.S. dollar). A large degree of excess demand exists for imports
at the fixed rate of exchange, as indicated by the open market exchange
rate of up to 40 kwacha per U.S. dollar. Debt service payments were
limited to 10 percent of export receipts less payments for imports of
petroleum and fertilizer and imports for the mining sector.
Continuing high inflation since the fixing of the exchange rate has
given rise to a large effective appreciation of the kwacha. A general price
freeze on all commodities was replaced by price control of twenty-one
essential groups of commodities (Republic of Zambia 1987). The price
control system did not reduce the high inflation rate, which accelerated
as a general shortage of basic goods emerged.
Future Agenda
The agenda for Zambia remains the same as before: to stop the decline
in per capita income, to diversify the economy's production base, to
improve the management of public resources, and to resolve the external
debt problem. Diversification of domestic production and exports is par-



Zambia: Reform and Reversal 347
ticularly critical. Zambia has only about ten years of copper reserves
remaining, and as reserves are drawn down, average production cost is
expected to rise and profitability to fall. (Zambia is already a high-cost
copper producer.) Clearly, Zambia cannot rely in the long run on its
copper exports to generate foreign exchange, public revenue, employ-
ment, and growth.
Zambia's large external debt and debt service requirements are another
critical problem. Zambia has been in noncompliance with its major cred-
itors since May 1987. External debt, including short-term obligations
and debt to the IMF, was estimated at US$6.4 billion at the end of 1987-
3.4 times the size of GNP or 6.7 times the value of exports of goods and
services-while external debt arrears were twice the value of exports.
To resolve the present impasse Zambia must find innovative ways of
reducing the external debt to a manageable size and get back on an
economic reform track that is workable, politically feasible, and ac-
ceptable to international lenders.
Conclusions and Lessons
Zambia's performance under structural adjustment was at best mixed
during the three years it was pursued. Marginal reductions were made
in the current account deficit. The decline in GDP was arrested. Capacity
utilization (which was as low as 40 percent) in the industrial sector im-
proved slightly with the increase in imports of intermediate goods. There
was a substantial depreciation in the real effective exchange rate, which
would have helped to allocate scarce foreign exchange more efficiently
had it been sustained longer. Nominal interest rates, as well as the prices
of a large number of consumer and producer goods, were also raised.
On the negative side, management of the macroeconomy deteriorated
rapidly after 1985 because of the inflationary effects of the exchange rate
depreciation, a continuous decline in copper exports, and the accumu-
lation of external arrears. Although some public expenditures were cut,
increased government reliance on domestic banking to finance rapidly
rising current expenditures, particularly for interest payments, and the
depreciation of the kwacha contributed to a budget deficit exceeding 35
percent of GDP in 1986. Rapid declines in real wages and the standard
of living and increasing unemployment led to social unrest and the erosion
of political support for the adjustment program. The government's failure
to modify some elements of the reform program and still comply with
the terms of the IMF and the World Bank and its inability to meet rapidly
accumulating debt service obligations finally led it to abandon the reform
program.
Why did the reform program fail in Zambia? Some preliminary con-
clusions emerge from this brief review.



348 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Program Design
The program placed too much emphasis on short-term adjustment of the
balance of payments and fiscal budget deficits through the reduction of
aggregate demand, which had already been falling since 1975. Not
enough emphasis was placed on expanding domestic supply and arresting
the decline in per capita income. Although World Bank-supported sec-
toral restructuring programs were designed to increase supply, there was
a lag in the sectoral response, and stabilization policies could not buy
enough time for the supply response to manifest itself before conditions
had deteriorated.
No comprehensive assessment of the probable macroeconomic re-
sponse to the adjustment policies was made before the program was
designed. In particular, the requirement that Zambia adopt a foreign
exchange auction failed to take into account Zambia's ability to address
the fiscal deficit. Since Zambia had gone through years of significant
decline in per capita income, the public sector's ability to reduce the fiscal
deficit in the short to medium term was very limited. It was unlikely that
expenditures could be cut to any significant extent, particularly since
they had already been declining in real terms and nondiscretionary items,
such as interest payments and defense, constituted a large part of the
total.
The current account deficit was easier to control than the fiscal deficit
because the availability of external finance set the maximum limit on the
size of the current account deficit. On the domestic side, however, the
government has a freer hand in financing its deficit by borrowing and
expanding the money supply. When exports are falling and external fi-
nance is increasingly scarce, the burden of external adjustment is on the
domestic side. In Zambia this resulted in a high rate of inflation and an
unmanageable fiscal deficit.
External economic conditions also worked against Zambia. The an-
ticipated improvement in the price of copper never materialized. An in-
crease in the value of copper exports would have gone a long way toward
consolidating the government's achievements and attracting further po-
litical support for the program. But when this failed to happen, no con-
tingency plans were in place to allow for the necessary adjustment.
A more realistic assessment of Zambia's problems and capacities and
of the future prospects for copper exports might have significantly im-
proved the design and sustainability of the adjustment program. For ex-
ample, a more concerted effort could have been made to reduce nondis-
cretionary recurrent expenditures, the program could have been extended
over a longer time, a longer grace period could have been allowed for
rescheduled debt, and the World Bank could have monitored the program



Zambia: Reform and Reversal 349
more closely, to permit greater flexibility in responding to conditions in
Zambia.
ADEQUACY OF RESOURCES. The net flow of external resources proved
to be much less than expected. External assistance, although vital for the
adjustment program while it lasted, fell short of Zambia's requirements.
The net flow of long-term capital was also limited because of Zambia's
lack of creditworthiness and the general decline in international lending
to developing countries. Direct foreign investment in Zambia was neg-
ligible. The government's ability to mobilize external assistance was
linked to its success in carrying out its reform commitments, while per-
formance was to a great extent dependent on the timing and availability
of sufficient concessional external assistance. Keeping up the flow of
external resources was vital to the survival of the adjustment program,
but it did not happen.
In retrospect, it is apparent that the government and World Bank staff
were overoptimistic in assessing the availability of real resources, the
expansion of copper exports, the capability of public administrators to
execute the reform program, and the extent of the supply response to
the foreign exchange rate and other price adjustments. Revenue from
exports of copper continued to decline and export diversification failed
to materialize.
PROGRAM TIMING AND DURATION. The social and economic problems
of countries such as Zambia originated long before they became inde-
pendent nations. Imbalances that have been created over a long time
cannot be redressed overnight. The responses of productive sectors to
more favorable incentives take time to emerge. The development of non-
traditional exports and a reduction in import dependency also may take
more time than is allocated within the framework of structural adjust-
ment lending.
Zambia began its structural adjustment program at a time of severe
economic and financial crisis. Per capita income had been declining for
more than a decade. Not enough had been done when the economy had
more strength to withstand the secondary shocks of adjustment, and too
much was attempted after it had exhausted its strength. Perhaps Zambia
tried to correct in too short a time the damage inflicted on the economy
by years of inappropriate policies.
Political Will
Insufficient political support has been cited as a major reason for the
failure of the reform. An opposing viewpoint expressed by some members
of the World Bank and the Zambian government is that the reform pack-



350  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
age was enthusiastically "bought" by an influential minority in the gov-
ernment. Within the ruling party, however, support for the program was
weak from the beginning, since the party stood to lose political influence
and access to financial resources as a result of the reforms. In the political
struggle that ensued in April 1986, the critics of the reform program were
put in charge of its implementation, a factor that certainly contributed
to its abandonment soon after.
Zambia's experience indicates that it is not sufficient for some members
of the government to "buy" a reform package. The managers of the
adjustment program and those who support it must be able to stay in
power long enough to execute the reforms. For the program to be sus-
tainable, it must provide adequate incentives for the political base of the
government in power to continue to support it until restructuring begins
to have some positive impacts.
Social Impact
Conspicuously missing from Zambia's reform program was any mean-
ingful attention to the impact of the adjustment program on social wel-
fare. This might have been because of the sectoral orientation of the
reform program, and had the adjustment program continued more at-
tention might have been given to improving social conditions. In any
case, it was felt that Zambia had no alternative to adjustment and that
without adjustment the standard of living would continue to fall.
Although the expected gains from adjustment were substantial, the
hardship faced by many people during the transition was immense. The
consumer price index for low-income groups increased by 72 percent
from October 1985 to February 1987. Unemployment was very high,
with as many as 2 million of a working population of 3.5 million reported
to be unemployed. The removal of subsidies after a long period of decline
in real incomes and widespread unemployment caused far stronger pro-
tests than was expected.
Structural adjustment entails a redistribution of resources and thus
changes in the returns to factors of production. It further involves re-
ductions in unearned income, subsidies, and administrative interventions.
There are many tradeoffs between present costs and probable returns in
the future. Although ultimately the gains from adjustment will be wide-
spread, there must be some short-term rewards to induce those who bear
the cost to continue to participate in the process. In Zambia it appears
that the present cost of adjustment for a significant portion of society
became greater than its discounted future value. Although Zambia has
terminated its externally supported adjustment program, its need for ex-
ternal assistance remains as critical as ever.



Zambia: Reform and Reversal 351
Notes
1. For comparative international costs, see Takeuchi, Strongman, and Maeda
(1986).
2. Zambia had received two program loans from the Bank in 1972 and 1976.
The concern then, as in later years, was with declining copper production and
prices, deterioration in the fiscal budget, the balance of payments position, ef-
ficiency of public investment projects, and the slow growth of the agricultural
sector.
3. World Bank staff were aware of the critical role of external assistance for
the survival of the reform program. A 1984 World Bank document stated that
if at least US$350 million of new commitments annually and $200 million of
additional resources could not be mobilized (on top of debt rescheduling), the
adjustment program had virtually no chance of success.
4. The difference in the share of government consumption in GDP (about 25
percent) and the share of current government expenditure in GDP as shown in
the fiscal data (56 percent in table 16-2) is due to large interest and transfer
payments.
5. These data are from the World Bank International Economics Department
Standard Tables of March 1988. More recent data (December 1988 Country
Brief) show the current account deficit for 1986 to have been US$434 million,
or 24.5 percent of GNP, larger than previously indicated.
6. In real terms, total government expenditures declined after 1970. In 1979
they constituted only 58 percent of the 1970 total, and in 1984 only 46 percent
(see Jansen 1988).
7. Since 1956 the World Bank has granted twenty-eight loans and twenty-
three credits to Zambia, totaling about US$951.4 million (1988 Country Brief).
Because of budgetary problems, the implementation of World Bank-assisted
projects in Zambia deteriorated in the 1980s. Serious delays in project imple-
mentation, particularly in the agricultural sector, led to slow disbursements of
loans and credits.
8. In 1981 it was foreseen that Zambia's structural inflexibility would take
many years to overcome, while its economy would remain vulnerable to terms
of trade fluctuations (World Bank 1981).
9. An example of such action is the World Bank's agreement to release the
proceeds of the agricultural sector credit to support the auction market.
References
Jansen, Doris J. 1988. Trade, Exchange Rate, and Agricultural Pricing Policies
in Zambia. World Bank Comparative Study. Washington, D.C.
Republic of Zambia. 1987. "New Economic Recovery Programme: Progress Re-
port No. 1 on the Implications of the Interim National Development Plan."
Lusaka.
Takeuchi, Kenji, John E. Strongman, and Shunichi Maeda. 1986. World Copper
Industry: Its Changing Structure and Future Prospects. Commodity Working
Paper 15. Washington, D.C.: World Bank.
World Bank. 1981. World Development Report 1981. New York: Oxford Uni-
versity Press.



3s2  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
. 1986 and 1988. Commodity Trade and Price Trends. Baltimore, Md.:
Johns Hopkins University Press.
. 1987a. World Debt Tables. Washington, D.C.
. 1987b. "Zambia, Public Expenditure Review." Washington, D.C.
Comments
Ben King
MOHSEN FARDI'S chapter is a careful, comprehensive, and sober account
of what happened in Zambia. His account ends with five reflections, to
which I will return later. I have a few picky comments to make, but my
main criticism is that his account is understated. So in a reversal of roles
I will compensate by overstating.
At my station of life, one is usually permitted (on sufferance) the luxury
of reminiscence. So I will begin by indulging-not entirely irrelevantly,
but perhaps somewhat irreverently. In the summer of 1952 I went on
my first operational mission for the World Bank: to Zambia, Zimbabwe,
and Malawi, then Northern and Southern Rhodesia and Nyasaland. That
was not only preindependence; it was prefederation.
We spent the most time in Zambia, or Northern Rhodesia, preparing
a loan to that country to be on-lent to the railways. The original request
to the World Bank was for 10 million pounds (then US$28 million), which
we reduced to 5 million pounds. At this distance in time, it is legitimate
to speculate how much this reduction was due to common prudence and
how much to tit for tat. I shall not try to answer, but our justification
at the time alluded to the fact that the production of copper had gone
up over 50 percent in six years and that the price had quadrupled. Wages
and incomes had risen sharply, and government revenue, which was ap-
proximately one-third of mineral output the year before, had increased
by a factor of more than seven. Typically, it was being spent on the boom-
towns of the copper belt and countryside. The government could easily
have found the 5 million pound difference and come back to the World
Bank when times were tougher.
Other themes at the time were the need for a countercyclical reserve
fund, the predictable consequences of suppressing public utility prices,
the burden of maize and wheat subsidies, the growing capital intensity
of production, and the neglect of agriculture. Many of these criticisms



Zambia: Reform and Reversal 353
had been made in a Colonial Office report fourteen years earlier, and
most were repeated in an internal World Bank paper nineteen years later.
Not much has changed, except that there will be no more copper booms
in Zambia.
I am indulging in these autobiographical trivia to place the structural
adjustment in Zambia in a historical context: not only Zambia's but the
World Bank's too. In that vein I will quote from the appendix to a little
book based on the Clayton lectures at Tufts University given in 1960 by
Eugene Black: "The Bank has been at least equally concerned with the
economic environment in which its loans are to be put to work. In loan
discussions, it has as a matter of course consistently urged attempts . . .
to put economic and fiscal policies on a sound footing." In other words,
lending and what we turgidly call policy advice were mutually reinforc-
ing, not mutually exclusive. But as is obvious from the book's title, The
Diplomacy of Economic Development, policy discussion was supposed
to take place in a diplomatic environment; it was intended as a means
of "illuminating choices rather than imposing solutions."
Ten years later things had changed. In the World Bank's annual report
for 1970 is a map of Africa covered with red and green dots representing
agricultural projects: green for the entire period before 1968 and red for
the two fiscal years after 1968. There are twenty green dots and twenty-
five red ones. Their message is clear. We now know what happened to
the red dots and their successors; the story can be found in several reports
by the World Bank's Operations Evaluation Department (OED) and in-
terim papers from the research project on Managing Agricultural De-
velopment in Africa (MADIA). On that map, Zambia has three red dots
and no green ones. One of the three was a forestry project. The two
agricultural projects proper did not fare well. One never got off the
ground and was canceled. The second had a negative rate of return, as
did a third that followed soon after. Price policy was at least partly to
blame.
During fiscal 1966 to 1977, roughly the first twelve years after Zam-
bia's independence, the World Bank lent Zambia US$463 million (gross,
before cancellations). The three agricultural projects were only a small
fraction of this total. Nearly US$200 million was invested in hydroelectric
power. There were two program loans of US$30 million each, one of
which was partially canceled. This flow of funds occurred despite re-
peated strong criticism of economic and fiscal policy in Zambia-in par-
ticular, the neglect of agriculture and the growth of the government's
payroll. The new government managed to emulate most of its predeces-
sor's substantial policies and to add a few of its own. It is hard to excuse
the World Bank-and other lenders-from being party to these excesses.
During this time, the World Bank became hostage to its own lending
program. Diplomacy was hard to practice when it became divorced from



354 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
lending. Indeed, it became an article of faith that diplomacy could not
be practiced in the context of project loans. As a consequence, project
loans suffered from their environment. OED audits of Zambian projects
reveal the effect of the price structure on many projects, the high in-
vestment cost per job, the shortage of staff and recurrent funds, and,
when there was staff, the frequency of its turnover. For far too long on
one agricultural project, the World Bank accepted assurances that some-
thing would be done to turn it around. Program loans were an occasional,
but not very effective, vehicle for reform.
Expansion came to an end in 1975 when copper prices collapsed. They
never again returned in real terms to their previous high levels. Gross
domestic income per capita, which in 1969 had been 40 percent higher
than in 1965, fell to a level close to 20 percent lower than in 1965. It
was the beginning of a continuous decline. Government revenue dropped
by 30 percent in 1975. For the first time, there was a substantial deficit
on the recurrent budget. A deep cut was made in capital expenditures,
and operating expenditures were cut in real terms. At the same time,
personal emoluments rose. But the greatest causes of increased expend-
iture were subsidies, defense, and interest payments.
Staff resources devoted by the World Bank to the study of the Zambian
economy were greatly increased at that time. In the fall of 1974 a large
mission went to Zambia to prepare a report on the agricultural sector.
This was followed in 1975 by a mission that produced a Basic Economic
Report incorporating the findings of the agricultural mission and spelling
out its prescription for changes in policy and practice. The report made
a reasonably hopeful projection of Zambia's future, predicated on
changes in policy. But it pointed out that its projections were highly
sensitive to the price of copper.
In 1976 the World Bank made its second program loan to Zambia and
the IMF its second standby credit. The World Bank's objectives included
agreement on the broad outlines of an agricultural action program and
a letter of intent to the effect that the government would finally establish
a development equalization account. But the account came too late; there
has been nothing to put in it.
In the following six years the World Bank's lending activities were
rather low key. Zambia was now considered eligible for financing on
concessional terms from the International Development Association.
Loans and credits were smaller on average than in the previous twelve
years. During this period the IMF'S involvement was greater than the
World Bank's; use of its credits rose from SDR9S million at the end of
1976 to sDRS76 million at the end of 1982.' In 1978 the government
asked for a third program loan and the consultative group of multilateral
and bilateral lenders to Zambia held its first meeting. But little came of
it because the World Bank considered Zambia's Third National Devel-



Zambia: Reform and Reversal 355
opment Plan to be unrealistic. In its appraisal of the proposed loan, which
was never made, the World Bank probed more deeply into additional
aspects of the pricing system (uniform prices) and the inefficiencies of
the marketing system. There was discussion of a structural adjustment
loan in 1980, but it was felt that conditions that would be acceptable to
the World Bank would not be acceptable to Zambia.
By 1982 the groundwork was being laid for reform in Zambia. The
World Bank's representative in Lusaka prepared a paper entitled "Ac-
celerated Development in Zambia: An Agenda for Action," which again
stressed the now familiar strictures about the economy. It also made the
prophetic observation that measures to curb consumption "will be much
less painful if they are taken in an environment of economic growth."
The government, on its side, drafted a memorandum that represented a
break with past policies. In fact, it was the first of a series of documents
in the next three years or so in which the Zambian government became
its own most severe critic. They are so persuasive that it is hard not to
sympathize with those who were working on Zambia in the World Bank;
in a context in which the chosen instrument for reform was an adjustment
loan, the series of credits that followed must have seemed a reasonable
gamble. One must ask not only why things went wrong, but whether it
was reasonable to expect this in advance. One must also ask whether
there was any other feasible approach.
Fardi classifies the reasons for breakdown under five headings. I wish
to comment on only a few, the first being the supply response. It is quite
true that the supply response was insufficient, although there were a
number of promising responses in the early days of the foreign exchange
auction. But in agriculture, the sector that is the main hope for Zambia's
future, the seeds of unresponsiveness had been sown years earlier through
the almost complete neglect of the smallholder sector and the prevalence
of incentives to consume agricultural produce and disincentives to pro-
duce it. Although there had been some response in some crops to inter-
mittent improvements in prices resulting from devaluation or other mea-
sures, price policy is only a necessary, but not a sufficient, condition for
growth.
T-hus those responsible at the World Bank had been dealt a hand with
very few honors by their predecessors. If the World Bank had taken a
long view of the Zambian economy, it would have insisted on a better
allocation of resources to agriculture as a condition for lending-any
kind of lending-during the 1970s; the case had been very well docu-
mented by that time. But if it was assumed, when a US$100 million loan
for power was under discussion, that matters vital to the economy could
not also be discussed, then of course such insistence would have been
difficult. That assumption ought now to be rejected. A comment on aid
to Tanzania (from a book published in Sweden, where aid is not usually



356 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
regarded with disfavor) appears, in retrospect, to be equally true of Zam-
bia: "A lower level of aid allocated in a better manner would have been
more useful. ... The assistance was part of the making of the crisis."
Fardi suggests that there was not much the government could do by
this time. On one, perhaps theoretical, level, this can be questioned. Re-
form of the agricultural sector was still the most neglected. Uniform prices
were never removed, and marketing arrangements were not clarified.
Producer and consumer prices of maize were out of step, controlled by
different ministries. It was this that led to the December riots. The foreign
exchange auction was never really given a chance. It lasted only six
months under sympathetic leadership. From then on, manipulation
doomed it. According to the World Bank's report on public expenditure,
there was still room for reducing or switching expenditures.
Too much is made of the burden of interest payments on the budget.
The reason they increased so much as a percentage of GDP is arithmetic;
the kwacha depreciated, and interest denominated in foreign currency
necessarily increased as a percentage of GDP, but this did not in any real
sense constitute a greater burden on the economy than before. One could
also say the same thing about the incoming aid on the other side of the
ledger. Whether reduction of Zambia's huge debt would have been a
more appropriate form of aid is another and a pertinent question.
As for domestic interest rates, one reason, if not the main one, for their
increase was the freeing of interest rates. And here one may question
whether a free capital market has any useful allocative effect when the
government is the borrower of first resort. If the fiscal deficit cannot be
brought under control, there is a case for rationing the remaining credit
to priority customers. (It was alleged, for example, that high interest rates
discouraged farmers.) One should really look separately at the primary
deficit, exclusive of interest payments, and the cost of interest payments.
Nonetheless, on a realistic rather than theoretical level, the judgment
that there was little the government could do may well be right. Even if
the auction had continued under sympathetic management, would the
government machinery have had the reflexes needed to manage an econ-
omy going through such a transition? A long history of protection from
market forces and the conviction that profits were evidence of profit-
eering rather than of market response might have suggested otherwise.
Subsequent events bore this out. When one talks of political will, much
depends on expectations. Other countries have committed themselves to
progressive devaluation, with all its discomforts, because they expected
the benefits to outweigh the discomforts. This expectation was always
weak in Zambia, and since both a quick and a large supply response was
unlikely, whatever support there was evaporated. History cannot easily
be expunged. As we now know, you cannot exorcise the ghost of Peron
with a few sprinkles of holy water.



Zambia: Reform and Reversal 357
Fardi refers to a tradeoff between costs now and returns in the future.
If there were costs to the economy attributable purely to the adjustment
process rather than to events, they would be hard to identify. But no
doubt the phrase refers only to social costs. With adjustment there are
losers and winners. Even if one finds the figure of 2 million unemployed
hard to credit (it suggests that almost everyone not employed by the
government was unemployed), it would be reasonable to suppose that
some of the losers would have been losers anyway and some would not.
For the most part, they would have been urban; nearly half the population
is urban, an extraordinarily high figure in Sub-Saharan Africa. In the
rural areas people have been losers for years. It is doubtful that they had
anything more to lose from the adjustment process; they might even have
gained if consumer goods became available at prices lower than previous
black market prices-or at any price if they had not been available at
all before. This is not to deny that it would have been morally right and
politically wise to afford some protection for those hardest hit, but the
fact that they had no alternative occupation to turn to underlines the
point about supply response.
A final point should be made about underfunding. To be sure, in re-
lation to the expectations of the foreign exchange likely to be available,
there was underfunding. As the chapter points out, this was due partly
to disappointment among the external lenders, but much more to lower
copper prices. What should be done in a world where prices do not always
live up to expectations? Would the aid community have underwritten a
worst-case scenario for the price of copper? Exactly who should bear the
risk? These are not easy questions to answer. They do suggest that "un-
derfunding" is not a term to be used in an absolute sense. Underfunding
in relation to what?
The suggestion that "Zambia had exhausted its strength" may be true;
if so, it implies that it was too late for a structural adjustment loan. But
earlier the time was not ripe. When was the right time? One school of
thought holds that the kind of wholesale reform a structural adjustment
loan calls for is, in some countries at least, unlikely to be undertaken
except in desperation, when economic breakdown has reached such a
level that the costs of not adjusting have become unbearable. If this is
so, one must really ask whether there are alternatives to structural ad-
justment as presently conceived.
Structural adjustment lending was born in 1980 in the wake of the oil
shocks and the subsequent worldwide recession. It was originally con-
ceived as an instrument to help developing countries over a perfectly
genuine adjustment hump by supplying what amounted to working cap-
ital. However, to the World Bank's Executive Directors, who asked why
project loans could not be used for this purpose, it was also presented
as "a fundamental instrument for dialogue" between the World Bank



358 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
and the country. When the Baker Plan came along in 1985, adjustment
loans were seen as the obvious instrument for moving it off its pedestal.
They have become the most favored notion of the 1980s, much as poverty
lending was in the 1970s. In neither case can the objective be criticized;
bringing about economic order and growth on the one hand and tight-
ening up income distribution on the other are worthy goals. But rigidity
of format is another matter. There may be times when it is necessary to
concentrate on a few essentials for a long time, to accept second-best
solutions, and to be patient.
Note
1. IMP funds are denominated in SDR (special drawing rights), which are based
on a basket of currencies.
Comments on Africa
Edward V. K. Jaycox
THE AFRICAN CASES presented above were well chosen for their implica-
tions for the adjustment process. The cases of Ghana and Zambia show
a similar degree of distortions present at the beginning of the programs,
as well as a similar degree of dependency on single products. These cases
and the difficulties of C6te d'Ivoire present valuable lessons for our re-
flection.
As the person now responsible for these operations in the World Bank,
however, I find something missing from the discussion of Zambia. Zam-
bia's was a terribly underfunded program. We overestimated copper rev-
enue, overestimated aid flows, and did everything we could to paint a
picture of an internally consistent financing plan based on the resources
that we and others could bring to bear. If the case had been looked at
more closely or more skeptically, the plan's lack of realism would have
become apparent. Certainly this is clear with hindsight. A great number
of shocks took place as the adjustment process went along: copper prices
went down or stayed at the same level when they were expected to go
up, aid that was expected did not arrive, deals with the Paris Club that
were normative were made less liberal when the aid was increased. At



Zambia: Reform and Reversal 359
one point, the actual financing was US$200 million less than had been
expected in net capital inflows.
What happens in a very weak administrative situation like this is that
the government resorts to the budget-to the creation of money-to meet
vital needs in the economy, and to short-term borrowing at high interest
rates. I do believe, however, that the end of the unraveling string in the
Zambian case was the underfunding of that program. Even though the
string goes all the way back to the 1940s, my real interest now is in what
happened in the period of formal structural adjustment. In sum, the Zam-
bian program was administered in a very chaotic way, and the chaos
resulted in part from the inadequacy of financing and unrealistic fi-
nancing projections.
The case of C6te d'lvoire shows that attempting to undertake structural
adjustment with both bands tied behind the back-that is, not able to
use the exchange rate as a policy variable and not able politically to
adjust nominal wages and prices downward-is not workable. The best
solution for this economy would have been to lower the cost structure,
which had gotten out of line. But that was not feasible politically in 1981,
nor has it proved to be politically feasible since then. Nominal wages
have to go down and real wages have to go down even faster in a zero
or very low inflation situation-a very, very difficult task. Since the use
of the exchange rate was not a policy option, what has been called the
second-best solution was used: the simulated exchange rate adjustment.
As a matter of fact, that is probably the second-worst solution because,
as pointed out by Shantayanan Devarajan in his comments on the chapter
on C6te d'Ivoire, both revenue and discipline in the economy were lost
at the same time. That approach did not work and, in my view, it is not
going to work. Furthermore, the program in C6te d'lvoire was also un-
derfunded. Lack of access to concessional finance, problems of credit-
worthiness and limits on creditors' exposure, lack of agreed arrange-
ments for the reduction of debt or debt service, and the resulting very
limited new or net flows of capital to C6te d'lvoire, coupled with the
policy constraints I have mentioned, make this a most difficult case.
In the case of Ghana, there is the more happy circumstance of adequate
funding matched by resolute implementation of reforms. This type of
program should be sustainable throughout Sub-Saharan Africa supported
by the kind of efforts that have been made by the aid community as a
whole. We have managed to mobilize a large amount of quick-disbursing
grant money and money at very concessional terms-International De-
velopment Association (IDA) terms or better. For poor African countries
that are debt-distressed, we have mobilized some US$6.4 billion, match-
ing the funds of IDA, to support structural adjustment.
The process of structural adjustment will require a long time. In the
case of Zambia, we knew it would be a fifteen-year effort. We knew this



360 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
was not a little hump; it was a long haul. Indeed, if we had had at the
beginning the kind of money we have today and had applied it to Zambia,
I believe the results would have been far different. The prospects of ac-
tually making that long haul are good, but this will require considerable
discipline and considerable management capacity. I firmly believe it is a
feasible and a necessary thing to do.



17 Indonesia: Stabilization and
Structural Change
Sadiq Ahmed
UNLIKE MANY other oil-exporting countries, Indonesia has managed its
oil resources well and has largely succeeded in protecting the commodity-
producing sectors (agriculture and manufacturing) and preventing them
from lagging behind the rest of the economy (Gelb and associates 1988).
During the 1970s Indonesia experienced a rapid growth in income, con-
sumption, and investment. In addition, important social gains brought
about a substantial reduction in the incidence of poverty.' Nevertheless,
some structural problems remained at the end of the decade. These re-
sulted chiefly from the complexities of the government's regulatory
framework, which had intensified the adjustment challenges posed by a
series of external shocks beginning in 1981. In response to these shocks
the government implemented a broad range of adjustment measures and
structural policy reforms that significantly reduced the macroeconomic
imbalances while allowing economic growth to remain positive in per
capita terms. Although some short-term social costs have been incurred,
Indonesia's long-term growth prospects would undoubtedly have been
severely jeopardized had these necessary adjustments not been imple-
mented.
This chapter examines Indonesia's adjustment to a series of external
shocks since 1981. The first section discusses the nature and magnitude
of these external shocks. The following sections review Indonesia's policy
response, analyze the role of the various policy instruments in the process
of stabilization and structural change, and evaluate Indonesia's economic
performance. The last section examines the World Bank's role in the
adjustment process.
Nature and Magnitude of External Shocks
Indonesia's external terms of trade fell by about 38 percent between 1981
and 1989. Oil prices accounted for the major part of this decline. In-
donesia's average crude oil export price of US$15 a barrel in 1989 was
only 36 percent of the real price in 1982. The deep worldwide recession
of the early 1980s and the decline in other commodity prices also hurt
Indonesia's balance of payments, although lower import prices provided
361



362 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 198Os
Table 17-1. Effects of External Shocks in Indonesia, 1984-89, as a
Percentage of GNP
Effect               1984  1985  1986  1987  1988  19891  1984-88
Terms of trade (1981
prices)              2.5    2.0   5.7  15.6   13.3  14.1    8.9
Exchange rate (1981
exchange rate)     -0.2  -0.3  -0.4   0.3    0.9   1.5    0.3
Interest rate (1981
interest rate)       0.6   0.4   0.1   0.0   - 0.5   0.0    0.1
Total effect          2.9   2.1   5.4  15.9   13.7  15.6    9.3
a. Estimated.
Source: Ahmed (1989).
partial relief. The adverse effect of the worldwide recession on export
volumes was especially severe in 1983.
Another adverse external shock was the depreciation of the U.S. dollar
after mid-1985. Since a large proportion of Indonesia's foreign debt is
denominated in currencies that have appreciated in relation to the dollar,
Indonesia's total debt in dollar terms has surged. These currency fluc-
tuations, together with the loss of export revenue because of falling oil
prices, significantly increased Indonesia's debt service payments (interest
and principal). Unlike many other developing countries, however, es-
pecially in Latin America, Indonesia did not have its debt service burden
seriously increased by higher international interest rates, mainly because
of the predominantly concessional nature and fixed interest rates of its
debt. Another factor in Indonesia's favor was its low borrowing require-
ments (because of its large oil earnings) during 1978-81 when inter-
national borrowing rates surged.
The magnitude of the adverse effects of external shocks on the Indo-
nesian economy is illustrated in table 17-1. On average, Indonesia suf-
fered an income loss equivalent to about 9 percent of its annual gross
national product (GNP) due to external disturbances during 1983-88.
The loss of income was especially large in 1987 and 1988. As expected,
the terms of trade effect was dominant, while the interest rate effect was
minimal. The loss of income from international currency fluctuations
increased significantly in 1987-88. The burden of external shocks was
substantial and presented a major challenge to policymakers. Without a
forceful policy response, this large loss of income could easily have de-
stabilized the economy.
Indonesia's Adjustment Program
Indonesia's adjustment program was initiated in 1983 and intensified
after the collapse of oil prices in 1986. In broad terms, two types of



Indonesia: Stabilization and Structural Change 363
policy adjustments were made. First, the government adopted more aus-
tere macroeconomic policies to restore financial stability. Second, to sus-
tain the development momentum over the medium to longer term, it
embarked on a major program to restructure the economy by improving
economic efficiency and reducing the country's heavy dependence on oil
exports as a source of foreign exchange and budgetary revenue. This
structural adjustment program included a range of measures to
strengthen domestic resource mobilization, expand nonoil exports, and
promote a more competitive and dynamic nonoil economy. A summary
of the policy measures initiated through the end of 1988 is presented in
table 17-2. These policies can be grouped under four broad categories:
exchange rate management, fiscal policy, monetary and financial policies,
and trade and other regulatory reforms.
To restore balance of payments stability and sustain growth over the
medium term, the government devalued the rupiah by 28 percent in
March 1983.2 At the same time, the flexibility of the exchange rate was
increased through a more actively managed float. The rupiah was again
devalued by 31 percent in September 1986 in response to the rapid decline
in oil prices. Since then, the rupiah has depreciated against a falling U.S.
dollar. As a result of two large devaluations, a managed float policy, and
the government's ability to restrain inflation, Indonesia's real effective
exchange rate depreciated by about 55 percent between December 1981
and December 1988. The downward trend of the real effective exchange
rate was generally maintained throughout the adjustment phase. As dis-
cussed later, the substantial depreciation of the real exchange rate played
a crucial role in the adjustment process by stimulating nonoil exports
and restraining imports.
Fiscal Policy
The objectives of both demand restraint and structural change were
served by the implementation of strong fiscal measures designed to re-
strain public expenditure and mobilize public resources. Many large cap-
ital-intensive projects were rephased in 1983, with an estimated foreign
exchange savings of US$10 billion. In the late 1980s civil service salaries
were frozen and real capital spending by the central government was cut
by more than one-third. Some actions were also taken to restrain in-
vestment in public enterprises. Tight control was kept on the use of non-
concessional import-related credits after 1984, and equity participation
in public enterprises funded through the budget was reduced to minimal
levels. A sweeping tax reform implemented in 1984-86 boosted nonoil
tax revenue and improved the efficiency of the tax system. As a result



364 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 17-2. Adjustment Policies Initiated in Indonesia, 1983-88
Policy
instrument                             Policy description
Exchange rate    Rupiab devalued by 28 percent in March 1983.
Exchange rate made more flexible after March 1983.
Rupiah devalued by 31 percent in September 1986.
Exchange rate depredated against a falling U.S. dollar after September
1986.
Fiscal           Large capital- and import-intensive projects rephased in May 1983.
Major cutbacks in government real capital spending after 1983.
Tight control maintained after 1983 on the use of nonconcessional,
import-related credits.
Major tax reform initiated in January 1984.
Follow-up steps taken to strengthen tax administration.
Restraints established on civil service employment and salaries.
Monetary and     Major financial reform in June 1983 removed interest rate and credit
financial        ceilings for state bank operations and introduced new instruments
of monetary control.
New financial measures introduced in October and December 1988
to improve financial sector efficiency and develop capital markets.
Monetary management for controlling inflation improved.
Short-term monetary management for curbing exchange rate specu-
lation improved.
Trade and other   Across-the-board reduction in nominal tariffs implemented in March
regulatory       1985.
Measures to provide internationally priced inputs to exporters an-
nounced in May 1986.
Significant reduction in import licensing restrictions initiated in Oc-
tober 1986, January and December 1987, and November 1988.
Regulatory restrictions for exporters reduced in December 1987 to
limit the antiexport bias of trade policy.
Customs, ports, and shipping reorganized in April 1985 to reduce
freight costs and cut procedural time.
Steps taken in May 1986, October 1986, and January and December
1987 to reduce investment and capacity licensing requirements,
relax foreign investment regulations, and reduce the role of the local-
content program.
Substantial deregulation of maritime activities announced in Novem-
ber 1988 to reduce costs and encourage private participation.
of these fiscal measures, total nonoil taxes of the central and local gov-
ernments as a percentage of nonoil GDP grew from 8.3 percent in 1983
to 11.5 percent in 1989, while the overall public sector deficit declined
during that period from 4.9 percent to 3.4 percent of GDP.



Indonesia: Stabilization and Structural Change 365
Monetary and Financial Policies
Supportive monetary and financial policies were implemented to contain
inflationary pressures, prevent capital flight, mobilize financial resources,
and improve efficiency in the use of financial resources. A major reform
of the financial sector, initiated in June 1983, deregulated state bank
interest rates, simplified subsidized lending rates for priority sectors, and
replaced credit ceilings with a system of reserve money management (see
Balino and Sunderarajan 1986). As a follow-up to this reform, a new set
of financial measures was announced in two steps, on October 27 and
December 20, 1988. These initiatives were aimed at improving the ef-
ficiency of the financial sector by increasing competition, encouraging
the development of the capital market, and further improving the effec-
tiveness of monetary management.
Apart from these financial measures, the government's targets for mon-
etary expansion were generally based on maintaining low rates of infla-
tion. The deregulation of domestic interest rates helped restrain capital
flight, increase mobilization of private saving by the banking sector (see
Chamley and Hussain 1988), and reduce inefficiency in the use of fi-
nancial resources. The combination of cautious monetary management
and austere budgetary management has noticeably reduced the average
rate of inflation since the early 1980s.
Trade and Other Regulatory Reforms
The government initiated a series of trade and other regulatory reforms
to increase the effectiveness of demand management policies in reducing
macroeconomic imbalances and to facilitate economic recovery in the
medium term. Trade policy reforms sought to reduce the burden of im-
port licensing and to signal the government's intention to shift to tariffs
as the principal instrument of import policy. As a result of these policies,
the value of imports subject to controls declined from 43 percent in mid-
1986 to 21 percent in December 1988. More important, the proportion
of domestic production protected by import licensing restrictions was
reduced from 41 percent to 21 percent.
Notable progress was also made in simplifying and reducing other
regulatory impediments. Through a series of steps taken between May
1986 and December 1987, the investment approval process was stream-
lined, licensing requirements were eased, the bias against foreign invest-
ment was reduced, and the role of the local-content program was di-
minished. In May 1989 the government converted the Investment Priority
List to a short negative list, thereby simplifying the system and opening
up new sectors to domestic and foreign markets. Reforms in the area of
customs, ports, and maritime transport implemented in 1985 substan-



366  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
tially reduced freight costs and procedural time. Subsequently, the gov-
ernment removed virtually all regulatory constraints on the development
of an efficient and responsive maritime transport sector. Along with trade
measures, these other regulatory reforms boosted domestic and foreign
investment and nonoil exports by improving business confidence and
reducing costs.
Economic Performance
The economic effects of Indonesia's adjustment program are discussed
below with particular reference to the balance of payments, fiscal deficit,
inflation, interest rates, efficiency, growth, and expenditure. The final
section looks at the social costs of adjustment.
Balance of Payments
The effect of the adjustment program on the balance of payments is
summarized in table 17-3. The current account deficit, following a surge
Table 17-3. Balance of Payments in Indonesia, Selected Years, Fiscal
1982-89
(billions of U.S. dollars at current prices)
Item                          1982    1983    1986    1987   1988    1989
Merchandise exports (f.o.b.)   23.0    18.6    18.5    13.7    18.1    19.8
Oil and liquefied natural gas    18.8    14.7   12.3   7.0     8.6     7.7
Nonoil                        4.2     3.9      6.2     6.7     9.5    12.1
Merchandise imports (c.i.f.)  - 20.0   -20.7  -14.4   - 12.7  - 14.2   - 15.5
Oil and liquefied natural gas   -5.4    -4.8   -3.2   -2.3   -2.4   -2.1
Nonoil                     - 14.6  - 15.8  -11.2   - 10.4  -11.8   - 13.4
Current account balance       -2.7    - 7.3   -2.0    -4.0   - 1.6    -1.9
Oil and liquefied natural gas  9.8     7.2     5.9     2.4     3.8     3.0
Nonoil                     -12.5   - 14.5   -7.9    -6.4   -5.4    -4.9
Public medium- and long-term
loans (net)                   2.2     4.0      1.4     2.7     2.1     2.7
Other capital (net)             0.6    - 0.0     1.5    -1.7     0.7    -1.6
Use of net foreign assets     -0.1      3.3   -0.9       3.0   -1.2      0.8
Current account/GNP (percent)    -3.0    -7.9   -2.4   -5.9   -2.3    -2.5
Noninterest current account
balance/GNP (percent)       -1.3    - 6.0      0.7   - 1.9     2.3     2.1
Medium- and long-term debt
service/exports (percent)    10.4    16.8    24.4    37.3    33.9    36.3
a. Estimated.
Sources: Bank Indonesia and World Bank staff estimates.



Indonesia: Stabilization and Structural Change 367
to 7.9 percent of GNP in 1983, fell to 2.4 percent of GNP in 1986. The
collapse of oil prices in 1986 led to another sharp deterioration in the
balance of payments, and the current account deficit climbed to 5.9 per-
cent of GNP in 1987. Subsequently, the adjustment measures succeeded
in lowering the deficit to less than 3 percent of GNP. Given the large loss
of foreign exchange earnings caused by external shocks, the ability to
reduce the current account deficit below the preshock level of 1982 is
evidence of the success of the government's adjustment program in re-
ducing the external imbalance.
In the initial phase of the adjustment period (1983-85), the balance
of payments adjustment was effected largely through a strong reduction
in imports. Several factors contributed to this reduction. First, public
sector imports of capital goods were reduced sharply by a rephasing of
large projects in March 1983 and by reductions in real capital spending
in subsequent years. Second, by redefining priorities, the government
diverted its expenditures from sectors with a relatively high import-in-
tensiveness (mining and industry) to those that were less import-intensive
(agriculture, education, and transport). The government also restrained
the import intensity of capital expenditures of state enterprises by im-
posing strict limits on the use of nonconcessional trade credits. Third,
the large depreciation of the real exchange rate raised the price of imports
relative to domestic goods, thereby causing private sector demand to shift
from imports to domestic substitutes (see Ahmed and Chhibber 1989).
Fourth, a strong public investment effort during the oil boom increased
import-substitution activities in such important commodities as rice,
sugar, processed food, cement, and fertilizer. Fifth, the proliferation of
nontariff barriers in the form of import licenses after the early 1980s also
led to a compression of imports. (This process was reversed beginning
in late 1986 when the government recognized the loss of economic ef-
ficiency caused by nontariff barriers. It then initiated a series of measures
to deregulate trade and reduced the economy's reliance on import sub-
stitution.)
The relatively smaller impact of nonoil exports on balance of payments
adjustment during the initial years is partly explained by the small nonoil
export base at the start of the adjustment program. For example, the
share of nonoil exports in total exports of goods (in real terms) was only
24 percent in 1983. The performance of nonoil exports began to improve
immediately after the institution of the competitive exchange rate policy
in March 1983.3 The expansion of nonoil exports was especially en-
couraging during 1986-88, when they grew by an average of about 20
percent annually in real terms; by fiscal 1989 the share of nonoil exports
in total exports (in real terms) had increased to 41 percent.4 Deregulation
measures affecting trade and domestic production also contributed to the
expansion of nonoil exports.



368 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Despite Indonesia's good performance in external adjustment and its
cautious approach to external borrowing, the debt burden rose sharply
in the 1980s. By the end of 1988 the stock of medium- and long-term
public and private debt outstanding and disbursed had reached an esti-
mated US$47.5 billion. Reflecting the adverse external environment the
country has faced since the early 1980s, many of Indonesia's debt in-
dicators are above the average for all developing countries and in line
with those for the seventeen highly indebted countries.' Depreciation of
the U.S. dollar after 1985 added US$12.6 billion (31 percent) to Indo-
nesia's public debt at the end of 1988 and US$1.9 billion (25 percent)
to its debt servicing during 1988.' Over the same period, oil prices fell
by about one-half, severely reducing Indonesia's export earnings and
capacity to service its debt.
Despite the rise in its debt indicators, Indonesia is distinguished from
the highly indebted countries in several important respects: (1) it has
maintained sound economic policies and a prudent borrowing strategy;
(2) as a result, it receives strong financial support from official sources
on concessional terms; (3) it has substantial reserves available in the form
of foreign exchange and undrawn lines of credit; and (4) it has retained
access to new voluntary lending from commercial banks. For these rea-
sons, Indonesia has not faced a foreign exchange crisis or a cash-flow
constraint on payment of its external debt obligations. Net transfers of
external assistance have risen sharply from US$0.4 billion in 1986 to
US$2.0 billion in 1988. At the same time, the government has maintained
strict limits on import-related and commercial credits and has reduced
exposure to private banks. This voluntary restraint has limited the total
net transfer of resources to Indonesia but has improved the structure
(maturity and terms) of its external debt. As a result, Indonesia's debt
structure is now better than that of most developing countries, with a
relatively high share of concessional debt and a relatively low share of
variable-interest debt.
Indonesia's recent performance compares favorably with that of the
highly indebted countries. Even with recourse to debt reschedulings and
moratoriums, net transfers to the most highly indebted countries have
turned substantially negative in recent years, reflecting the loss of access
to new voluntary lending, especially from commercial banks. As a result,
economic growth has been curtailed and per capita consumption has
declined in eleven of the seventeen highly indebted countries. Although
Indonesia has also made substantial cuts in investment since the early
1980s, the government was able to sustain GDP growth at 3.3 percent a
year on average after 1983 through policies to improve efficiency and
promote the development of nonoil exports. With prudent external bor-
rowing and better economic performance several of Indonesia's debt in-
dicators have declined, including the ratios of debt outstanding and dis-



Indonesia: Stabilization and Structural Change 369
bursed to exports, debt outstanding and disbursed to GNP, and debt
outstanding and disbursed to resources.7
Fiscal Deficit
The effects of the fiscal measures on public resource mobilization and
the central government's budget deficit are summarized in table 17-4.
Responding to the tax drive, the central government's nonoil taxes grew
from 7.4 percent of nonoil GDP in 1983 to 10.8 percent in 1989. Measures
to control expenditure were also largely successful. Thus, despite the
growing burden of external interest payments, total expenditure declined
from 22.7 percent of GDP in 1983 to 20.1 percent of GDP in 1989. As a
result of revenue mobilization and expenditure control, the budget deficit
narrowed from 4.4 percent to 2.8 percent of GDP over the same period.
The restraining influence of the budget is further indicated by the sub-
Table 17-4. Indonesian Central Government Budget, Selected Years,
Fiscal 1983-89
(trillions of rupiah at current prices)
Item                                1983    1986    1987    1988    1989a
Revenue and grants                  11.9    18.8    16.7    21.8    23.5
Oil and liquefied natural gas taxes  7.6    10.7     6.3    10.4      9.5
Nonoil taxes                       3.8      6.4      7.9      9.0    11.9
Other                              0.5      1.7      2.5      2.4     2.1
Current expenditure                  8.2    12.7    13.4    15.5    16.8
External interest                  0.7      1.8      2.8      3.8     4.3
Subsidies                           1.4     1.1      0.9      1.4      1.0
Other                              6.1      9.5      9.7    10.3    11.5
Government savings                   3.7      6.1      3.3     6.3      6.7
Capital expenditure                  6.6      8.7      7.3     9.2    10.5
Overall balance                     -2.9    -2.7    -4.0    -2.9    -3.8
Financed by
Extemal loans (net)                2.0      1.8      3.7      2.8     4.2
Asset drawdown                     0.9      0.9      0.3      0.1    -0.4
Percentage of GDP
Nonoil taxes (percentage of nonoil
GDP)                             7.4      8.4      9.8      9.4    10.8
Overall balance                   -4.4    -2.8    -3.9    -2.4    -2.8
Total expenditure                 22.7    22.6    20.3    20.7    20.1
Net domestic expenditureb          11.4     8.5      3.8      4.1      2.2
a. Estimated.
b. Defined as the domestic content of expenditure less nonoil revenues.
Sources: Indonesia Ministry of Finance and World Bank staff estimates.



370 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 17-5. Overall Public Sector Balances in Selected Asian Countries,
Fiscal 1982-88
(percentage of GDP)
Average,
Country          1982   1983   1984   1985   1986  1987  1988   1982-88
Bangladesh'      - 11.7  -11.2   -9.2   - 7.5  - 7.3  - 8.1  - 7.8   -9.0
India             -6.5   -7.0   - 7.2   -8.6  - 8.6  -9.1   n.a.    -7.8
Indonesia         -1.9   -4.9   -2.6    0.6  -3.1  -4.4  -3.2        -2.8
Malaysia         -19.7  -18.0  -15.9  -12.3  -5.3  -9.8  -7.8    -12.7
Pakistan'         -5.5   -5.6   -7.4  -6.2  -8.1  -7.7  -8.8         -7.0
Philippinesb       n.a.    n.a.  - 9.0  -8.2  -5.9  -5.4  -3.2    -6.3
Tlhailand         -7.9   -5.7   -4.5   -6.3  -4.7  -2.6  -2.2        -4.8
Indonesia primary
balance'        - 0.9   -3.7   - 0.9   -2.6  -1.2  -1.6   0.1      - 0.8
n.a. Not available.
a. Central government deficits.
b. As a percentage of GNP.
c. The overall public sector balance net of interest payments.
Source: World Bank Country Economic Reports.
stantial decline in net domestic expenditure from 11.4 percent of GDP in
1983 to 2.2 percent of GDP in 1989.
TIhe central government's budget deficit, however, understates the true
level of the public sector's net claim on resources because it does not
include the operation of public enterprises and local governments. Table
17-5 compares overall public sector deficits in Indonesia with those in
other Asian countries. This comparison shows that Indonesia has gen-
erally maintained cautious fiscal management, despite the loss of oil rev-
enue. Also encouraging is the trend of the primary balance (the public
sector deficit net of interest payments), which registered a small surplus
during 1988.
Austere budgetary management has been an integral element of the
adjustment program and has served to reduce demand pressures on both
the balance of payments and prices. By maintaining low deficits and
restraining access to bank credit, budgetary management has helped to
contain inflation and avoided a crowding-out of private investment. By
restraining aggregate demand and lowering the import intensity of ex-
penditures, budgetary management has also assisted in reducing the cur-
rent account deficit.
Inflation
The effect of the adjustment program on inflation is summarized in table
17-6. All three measures of domestic inflation-consumer price index



Indonesia: Stabilization and Structural Change 371
Table 17-6. Domestic Inflation Trends in Indonesia, 1973-81 and
1981-88
(average annual percentage rate)
Indicator                                    1973-81                1981-88a
Domestic inflation
Consumer price indexb                        15.3                     8.4
Wholesale price index (nonoil)               16.6                    11.6
Nonoil GDP deflator                          15.9                     9.3
Selected price indices
Import goods'                                13.6                    11.4
Export goodsd                                19.0                    12.7
Rice'                                        15.7                     9.1
Nontraded goodsf                             14.0                     8.9
a. Price indices for 1988 were adjusted upward to reflect better the impact of international
inflation and rice prices.
b. Because data on the cpi for seventeen cities are available only from 1979, the CP[ for Jakarta
is used.
c. In domestic currency.
d. In domestic currency, excluding oil.
e. Medium-quality wholesale price.
f. Housing and other services.
Sources: Indonesia Central Bureau of Statistics and World Bank staff estimates.
(cPi), wholesale price index (wpi), and the nonoil GDP deflator-show a
significant deceleration of the average rate of inflation. The sharpest re-
duction is indicated by the cPi, while the wpi shows a more modest decline.
Despite large devaluations, cost-push pressures on domestic inflation dur-
ing 1981-88 were contained by lower levels of world inflation and the
government's ability to restrain rice prices. Wage pressures were very
limited because of restraints on public sector wages. (Indonesia has a
small formal wage employment sector in which the public sector is dom-
inant.) Even in the formal manufacturing sector, available data suggest
a slowdown in the growth of wages. In the informal sector, wage ad-
justments were generally linked to the price of rice, which grew in a stable
fashion until 1986, reflecting virtual self-sufficiency in rice and the sta-
bilization of rice prices.8 On the demand side, austere budgets and the
slowdown of domestic income growth assisted in reducing demand pres-
sure on the exchange rate and restraining growth in the prices of non-
tradables.
The low transmission of inflationary pressures through cost-push and
demand-pull forces was supported by an appropriate monetary policy.
The government curbed inflation partly by slowing the rate of growth
of reserve money but chiefly by inducing people to hold a larger volume
of money (M2) by deregulating domestic deposit rates (see table 17-7).
The surge in the growth of quasi-money (QM, time and savings deposits),



372  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 17-7. Growth of Money and Credit in Indonesia, 1973-81 and
1981-88
(average annual percentage rates)
Item                                     1973-81             1981-88
Reserve money                              28.0                15.0
Currency and demand deposits (Ml)          32.8                12.3
Time and saving deposits (QM)              33.5                39.9
Money supply (M2 = Ml + QM)                33.0                25.1
Domestic credit                            21.5                25.7
Private credit                             26.6                28.0
Quasi-moneylGDP                             5.3                13.8
M2/GDP                                     15.I                24.5
Sources: Bank Indonesia and World Bank staff estimates.
despite much slower growth of nominal income, contributed to a sharp
increase in the financial deepening of the economy (reflected in the rise
in the QM/GDP and M2/GDP ratios). The increase in demand for money
in turn relieved demand pressure in the goods market and brought down
inflation.
The financial deepening and the government's austere budgetary stance
allowed the expansion of credit to the private sector, thereby avoiding
a supply crunch.9 The ability to bring down inflation even with two major
devaluations is a strong indicator of improved overall monetary man-
agement. Reserve money growth has been substantially reduced in recent
ye-ars. Moreover, despite some initial learning costs, short-term monetary
management has also improved, as shown by the government's success
in curbing speculative pressures on the exchange rate in March-June
1987.
Interest Rates
A major element of the June 1983 financial deregulation package was
decontrol of domestic interest rates. The sharp rise in state bank deposit
rates in conjunction with lower inflation caused real deposit rates to
become strongly positive-increasing from - 10.7 percent in 1981 to 9.8
percent in 1985. Similarly, the removal of the ceiling on lending rates
led to an increase in the average lending rate. High real lending rates
since 1983 have generated some concern about their adverse effect on
private investment, however.10 One factor underlying the high domestic
lending rates is the cost of financial intermediation. The current large
spread between deposit and lending rates (about 4 percentage points) is
a consequence of significant inefficiencies that have persisted in the fi-



Indonesia: Stabilization and Structural Change 373
nancial sector despite the 1983 financial reforms. The October-Decem-
ber 1988 financial measures addressed some of these outstanding prob-
lems. The reforms are expected to improve efficiency in the financial
sector significantly by increasing competition. They are also expected to
increase the availability of long-term finance by stimulating the growth
of capital markets.
One issue that has evoked intense debate is the role of monetary policy
in the determination of domestic interest rates in Indonesia. Because In-
donesia does not regulate private international capital flows and the ex-
change rate is not fully flexible, the domestic interest rate over the long
term will be determined by the international interest rate and the expected
rate of depreciation of the rupiah."'
In the past, three variables affected exchange rate expectations: the
swap premium rate,"2 the differential between domestic and international
inflation, and the price of oil. Exchange rate expectations appear to have
been especially sensitive to oil price movements. From 1983 to 1986 the
average difference between the domestic deposit rate and the interna-
tional interest rate was 6.3 percent, higher than the swap premium (5.1
percent) and inflation differential (cpi) between Indonesia and the United
States (4.6 percent). This happened because of uncertainty about the price
of oil, which prompted fears of an exchange rate depreciation that would
exceed the level implied by the inflation differential or the swap premium.
The September 12, 1986, devaluation and the recovery of oil prices in
1987 tended to dampen these fears in 1987-88, but two offsetting in-
fluences were the increases in the inflation differential (5.8 percent) and
the swap premium (9.1 percent). These increases widened the difference
between domestic and international interest rates even further and fueled
additional fears of a depreciation of the exchange rate. The recent move
to allow the swap premium to become a market-based value reflecting
international interest differentials rather than a policy variable is intended
to provide flexibility to the swap rate. In addition, releasing the premium
from its role as a policy instrument is intended to persuade people that
the swap premium does not reflect the government's view about future
exchange rate adjustments.
Since domestic deposit rates will need to be linked to international
interest rates to avoid capital flight, Bank Indonesia has rightly main-
tained that the main way in which monetary policy can affect long-term
interest rates is by sustaining low inflation rates and thereby dampening
fears of an exchange rate depreciation. In the short term, monetary policy
has also been used to protect domestic rates from the destabilizing in-
fluences of speculative capital flight.13 But an expansionary monetary
policy designed to achieve low interest rates has generally been avoided
because it would be self-defeating over the longer term by leading to high
inflation, capital flight, and balance of payments instability.14



374 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Economic Efficiency
With the substantial decline in oil revenue and the new limitations on
the public sector's capacity to support growth through its expenditure
programs, the need to improve economic efficiency and stimulate the
private sector became evident. The government recognized that a thor-
ough revamping of the regulatory framework was essential for achieving
this goal and, as mentioned previously, began a series of comprehensive
and far-reaching deregulatory measures in 1985. Because of the time lag
before many of these policy initiatives can affect behavior, it is too early
to evaluate their full quantitative impact. Even so, there is some evidence
of an improvement in economic efficiency in recent years.
A rough index of productivity is the aggregate return on investment.
Large investments financed through the oil boom fueled the growth of
the nonoil economy in 1973-81. Given the low initial capital base in
these sectors, these investments yielded a fairly high average rate of return
(see table 17-8). The return per unit of investment declined sharply in
1982-85 but has picked up noticeably in recent years. A closely related
indicator of capital productivity, the incremental capital-output ratio
(ICOR), reveals a similar pattern. A more comprehensive measure of the
gain in macroeconomic efficiency is the change in total factor productivity
(summarized in table 17-8 with the use of value added as a measure of
Table 17-8. Indicators of Aggregate Efficiency in Indonesia,
1973-81 to 1986-88
Indicator                        1973-81          1982-85          1986-88
Rate of return on investment
(percent a year)a              31.4             13.1            21.8
Incremental capital-output
ratio                           2.8              7.8              5.2
Total factor productivity
change (percent a year)'
Value added'                      8.0              4.0              5.1
Factor inputs
Labor                           3.0              2.8             2.8
Capitald                       10.7              9.8             5.2
Total factor productivity           0.9             -2.5              1.0
a. Rate of growth of nonoil GDP as a percentage of average investment rate during the period.
b. Total factor productiviry change is calculated as the difference between rates of growth of
value added and factor inputs (labor and capital); the inputs are weighted by their income shares.
c. Using nonoil GDP.
d. Capital stock was derived by using the perpetual inventory method.
Source: World Bank sraff estimates.



Indonesia: Stabilization and Structural Change 375
Table 17-9. Indonesia's Economic Structure, Fiscal 1982, 1986,
and 1989
(percent)
Item                                   1982           1986           1989
Oil and liquefied natural gas/total
exports'                             80.8           66.6           35.8
Nonoil exports/nonoil importsa         28.8           55.4           90.3
Oil and liquefied natural gas
revenues/total revenue               70.7           57.5           41.3
Government expenditure/GDP             22.1           22.6           20.1
Private fixed investment/total fixed
investmentb                          52.1           47.9           56.9
Manufacturing/nonoil GDPb               8.4            8.8            9.5
a. Goods only; in current dollars.
b. Calendar year; in 1983 prices.
Source: World Bank staff estimates.
production). A striking finding is that the main determinant of economic
growth in the oil boom period was the rapid buildup of capital stock.
Improvement in factor productivity played a marginal role. External
shocks reduced factor productivity in 1982-85, but the trend was re-
versed after 1985. Nevertheless, the low contribution of factor produc-
tivity to economic growth illustrates the scope for future gains in effi-
ciency.
The structural reforms that allowed the recent gains in economic ef-
ficiency also supported a noticeable improvement in the structure of the
economy (table 17-9). Indonesia's dependence on oil was substantially
reduced, and the reliance on more promising and sustainable sources of
growth increased. Especially encouraging is the strong positive response
of nonoil exports and private investment. The process of structural trans-
formation is not yet complete, however. With sustained progress in im-
plementing ongoing reforms and initiating new policy changes in required
areas, Indonesia can be expected to reach a higher real growth path on
a self-sustained basis.
Economic Growth and Expenditure
The government's balanced adjustment program also helped sustain a
rate of economic growth that was better than expected, despite the loss
of oil revenue. Although economic growth decelerated significantly over
the adjustment period, it remained positive in per capita terms (table 17-
10). Given the severity of the external shocks, sustaining a 3 percent
annual expansion of the overall economy and 4 percent annual growth



376  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 17-10. Overall and Sectoral GDP Performance in Indonesia,
1973-81 to 1988
(annual percentage real growth)
Item                   1973-81          1981-88          1987          1988
GDP                       7.5             3.3             3.9          4.7
Nonoil GDP                8.0             4.3             4.6          5.6
Agriculture               3.4             2.9             1.8          3.8
Manufacturing'           14.1             5.2             7.1          9.0
Services                 10.0             5.0             5.6          5.6
GDP per capita            5.2             1.3             1.9          2.7
Note: The Central Bureau of Statistics is currently reviewing its data and methodology for
estimnating GDP. This is expected to result in higher estimates of real growth rates for some sectors,
especially in 1988.
a. Excludes liquefied natural gas and oil refineries.
Source: Central Bureau of Statistics.
in the nonoil economy while also substantially reducing financial im-
balances is a notable achievement. Although the factors underlying the
rapid growth of the 1970s-oil earnings, rice production, and import-
substituting manufacturing-weakened during the adjustment phase,
nonoil exports, especially manufactured exports, emerged as a new
source of growth. Already there are encouraging signs of economic re-
covery, such as the upward trend in the rate of growth of the nonoil
economy since 1987, which reached an estimated 5.6 percent in 1988.
The lower rate of growth of GDP and the deterioration in the terms of
trade had an adverse effect on domestic income during 1981-86 (see
table 17-11). The shortfall in resources brought about substantial cut-
backs in public investment. Private investment fell noticeably until 1985
but has been recovering since then. The growth of private consumption
also declined significantly, although it still grew faster than national in-
come. Overall, the main adjustment happened in the public sector. As a
result of expenditure restraints, especially the large reduction in budget-
financed capital spending, public expenditure (consumption and invest-
ment) as a share of GDP fell from 22.4 percent in 1981 to 18 percent in
1988. In the private sector the decline in investment in the first few years
of the adjustment period was caused by higher real interest rates and
lower aggregate demand (see Ahmed and Chhibber 1989). Since 1986
private investment has been recovering in response to the improved in-
centives resulting from the regulatory reforms and a more buoyant level
of economic activity. Much of this new investment has been directed
toward export activities.



Indonesia: Stabilization and Structural Change 377
Table 17-11. Impact of Adjustment Policies on National Expenditure
in Indonesia, 1973-88
(percent)
Annual growth rate               Share of GDP'
Item               1973-81      1981-88     1987     1988      1981     1988
National income      11.4         1.8        4.7      4.8      97.9     88.6
Consumption           8.2         3.3        2.2      4.0      69.5     69.3
Public             10.1          2.2      -3.7      1.5      10.6      9.8
Private             7.8          3.4        3.3     4.4      58.9     59.5
Fixed investment     11.7        -0.5         1.9     7.0      24.7     18.9
Public             11.0        -2.0       -1.7      6.3      11.8      8.2
Private            12.3          0.7        4.9     7.6      12.9     10.7
Per capita national
income              9.1        -0.2         2.7     2.8       -        -
Per capita private
consumption         5.5          1.4        1.3     2.4       -
a. In 1983 prices.
Sources: Indonesia Central Bureau of Statistics and World Bank staff estimates.
The Social Costs of Adjustment
The adjustment process inevitably involved short-term social costs. De-
spite the recent recovery in economic growth, per capita income was
adversely affected by the terms of trade losses. A full assessment of the
social costs of adjustment is precluded by the lack of sufficient data on
how the adjustment process affected poverty and income inequality. Frag-
mentary evidence suggests, however, that the urban sector was more
adversely affected than the rural sector.
Part of the labor market adjustment took the form of an increased rate
of open unemployment in urban areas, especially among young school
leavers.15 Because the flexibility of educated urban job seekers is rather
limited, the slowdown in demand resulting from lower economic growth
and budgetary constraints on resources reduced the rate of their absorp-
tion in employment (especially in the civil service). For the bulk of the
urban work force, however, the main burden of adjustment fell on earn-
ings (especially in the informal sector); labor earnings fell in trade, and
real wages of construction workers and civil servants stagnated. In man-
ufacturing, real wages continued to grow, although at a slower pace than
before the external shocks. The slowdown in demand during 1982-85
hurt manufacturing sector profits, but the recent surge in nonoil man-
ufacturing exports has supported the recovery of profits in export-based
manufacturing enterprises, including small-scale enterprises.



378 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
The adjustment burden appears to have been generally less severe in
rural areas because farm incomes were relatively protected by several
factors: the increase in real rice prices; the real exchange rate deprecia-
tion, which largely offset lower agricultural export prices; and some prof-
itable diversification away from rice toward nonrice crops, especially
fruits and vegetables. Overall agricultural output was pulled down, how-
ever, by the slower growth of rice production, which in turn reduced the
demand for farm workers. Similarly, a slowdown in real public expend-
iture reduced the demand for rural construction workers. Because of the
flexible nature of rural labor markets, however, the result was not a
reduction in employment but rather a stagnation or decline in real ag-
ricultural wages (see Papanek 1988; Manning and others 1988)."6 Real
wages in rural construction activities apparently also remained flat.
The declining trends in labor earnings are attributable to the major
external shocks facing Indonesia and not to the adjustment response
itself. Indeed, the government moved to mitigate the social costs of ad-
justment by reallocating public spending to important social programs.
For example, the shares of development expenditure allocated to social
services and agriculture were significantly increased from 1983 to 1988.
An effort was also made to protect regional transfers, which normally
finance small-scale and labor-intensive infrastructure projects at the local
level. Given the large magnitude of the income loss, however, there was
some scaling back of social programs and some reduction in the quality
of service. But without this effort to redefine expenditure priorities in
support of lower income groups, the social costs of adjustment would
have been much higher. More important, by reducing financial imbal-
ances and promoting restructuring in the economy, the government's
reform program prepared the ground for sustaining a higher growth path
over the medium term. This is essential for absorption of the rapidly
growing labor force at higher levels of productivity and income.
The World Bank's Contribution to the Adjustment Effort
By any standard, the government's efforts to restore macroeconomic sta-
bility and to accelerate the process of structural change were remarkably
successful. The agenda for policy reform was carefully developed, and
the government, by making difficult but correct decisions, has clearly
established its credibility since 1983. The World Bank responded to the
government's adjustment efforts by focusing policy analysis and advice
on the key issues of macroeconomic management and structural reform,
while its lending program has expanded significantly to support contin-
ued policy reform by the government. At the same time, two adjustment
loans and annual economic reports have enabled the World Bank to be
a catalyst in mobilizing substantial additional support for the adjustment



Indonesia: Stabilization and Structural Change 379
effort from other external lenders who constitute the Intergovernmental
Group on Indonesia.
The World Bank's analysis and policy dialogue have supported the
government's efforts to define appropriate policy responses to emerging
issues of macroeconomic management. The World Bank prepared major
studies of trade and industrial policies in 1981 and 1985 that helped
identify the costs of the complex regulatory environment in Indonesia
and to show the broad directions for future policy change. With the
decline in oil prices, the push for reforms within Indonesia gathered mo-
mentum. To support this effort, the World Bank provided analytical
inputs on reforms in the trade regime and the regulatory environment.
With the sharp drop in oil prices in 1986 and the rapid increase in the
external debt burden, Indonesia needed to rely primarily on concessional
aid from official sources, rather than on commercial borrowing, to meet
its financing gaps. Indonesia's well-developed agenda for policy reform
and the government's well-established credibility for making sound policy
decisions enabled the World Bank to support structural adjustment
through sectoral loans for housing, urban development, and irrigation
and to support macroeconomic reforms through two trade policy ad-
justment operations.
In general, the World Bank's overall lending program has contributed
to important achievements at the sectoral level. Lending for irrigation
has supported self-sufficiency in rice production; lending for transmi-
gration and the development of tree crops in the Outer Islands has helped
reduce the incidence of rural poverty and increase nonoil export earnings;
lending for energy development, transport infrastructure, and industrial
finance has helped expand nonagricultural employment and diversify the
sources of growth; and lending for human resource development has
contributed to a major expansion in the supply of trained staff, the in-
creased availability of health services, and the success of one of the best-
managed and -administered population programs in the world.
The World Bank's annual economic reports and two adjustment loans
have provided the basis for the support of the Intergovernmental Group
on Indonesia, whose members substantially increased their aid flows and
made available fast-disbursing, untied balance of payments loans. This
assistance provides temporary support to the government while it im-
plements policies to adjust the balance of payments and the budget to
external shocks. The response from Intergovernmental Group members
has been very encouraging, with disbursements totaling US$3.7 billion
over the past three years. This valuable assistance has helped the gov-
ernment push ahead with deregulation and facilitated the recovery of
private investment and economic activity. It has also enabled Indonesia
to restructure its external debt, by improving the average maturity and



380 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
terms, while maintaining the confidence of the financial market and inves-
tors in the viability of the adjustment program.
Notes
I would like to thank Mark Baird, Amar Bhattacharya, Russell Cheetham, Ajay
Chhibber, Javad Khalilzadeh-Shirazi, and Attila Sonmez for helpful comments.
Errors are my sole responsibility.
1. The percentage of people falling below the designated poverty line dropped
from 57.1 in 1970 to 39.8 in 1980.
2. The magnitude of devaluation is calculated according to the International
Monetary Fund (IMF) method, which measures devaluation in terms of the rate
of change in the value of foreign currency per unit of the domestic currency (for
example, the value of the U.S. dollar per rupiah).
3. Empirical verification of the strong positive response of Indonesia's nonoil
exports to depreciation of the real exchange rate is contained in Ahmed and
Chhibber (1989) and Kincaid (1983).
4. In current prices, the share of nonoil exports grew from 21 percent in fiscal
1983 to 61 percent in fiscal 1989. This growth reflects the combined effects of
higher volumes and prices of nonoil exports and lower oil prices.
5. This group comprises Argentina, Bolivia, Brazil, Chile, Colombia, Costa
Rica, C6te d'lvoire, Ecuador, Jamaica, Mexico, Morocco, Nigeria, Peru, Phil-
ippines, Uruguay, Venezuela, and Yugoslavia (see World Bank 1989).
6. Excluding credits for the expansion of liquefied natural gas and liquefied
petroleum gas and for paraxyelene projects.
7. The ratios of medium- and long-term debt outstanding and disbursed to
exports, and debt outstanding and disbursed to GNP, declined from 236.8 and
71.0 in 1987 to 220.8 and 64.7 in 1988, respectively. The ratio of debt out-
standing and disbursed to resources fell from 137.3 in 1987 to 127.1 in 1988.
This ratio is obtained as a weighted average of debt outstanding and disbursed
to exports and of debt outstanding and disbursed to GNP, and it neutralizes the
bias of an exchange rate policy in the measurement of the debt problem (see
Cohen 1988).
8. Because the 1987 drought had adversely affected rice output, rice prices
surged starting in late 1987, and the upward trend continued until mid-1988.
Subsequently, rice prices stabilized in response to improved supply.
9. Indonesia's experience supports the McKinnon-Shaw argument for finan-
cial deregulation in developing countries. McKinnon's theory that the rate of
inflation can be brought down without reducing real money supply appears to
be of strong relevance to Indonesia's experience (see McKinnon 1973; Fry 1988).
10. Ahmed and Chhibber (1989) show that there is a statistically significant
negative relationship between the real interest rate and private investment in
Indonesia.
11. The domestic rate may adjust to the foreign rate in a lagged fashion, and
there may be some transaction costs.
12. The swap facility was set up in 1979 by Bank Indonesia to allow investors
borrowing abroad to hedge against the risk of a major devaluation by purchasing
a swap rediscountable at a fixed premium. The swap facility provided only a
partial arbitrage support, since it was constrained by limited access and a rigidly
set premium (until October 1988).
13. This was done very successfully during May-June 1987.



Indonesia: Stabilization and Structural Change 381
14. A framework linking the interest rate, exchange rate, and inflation in
Indonesia is provided in Ahmed and Kapur (1989).
15. Given the structure of Indonesia's labor markets, a slowdown in economic
activity is manifested largely in a reduction in labor earnings rather than in em-
ployment levels, except in the urban formal sector. Recent data show that more
than half of all employment is in agriculture and more than 70 percent in rural
areas. Employment outside agriculture is concentrated in services, with manu-
facturing accounting for only 20 percent of nonagricultural employment. Outside
of government service, low and unstable earnings in informal sector activities
predominate in urban areas. The considerable flexibility of rural and urban in-
formal labor markets allows a rapid expansion of employment during periods
of low economic growth, but at reduced real earnings.
16. There are some exceptions to these trends, especially in areas where a
more diversified rural economy and rural-urban linkages have continued to sup-
port increases in real agricultural wages (see, for example, Collier and others
1988).
References
Abmed, Sadiq. 1989. "Indonesia: External Shocks, Policy Response, and Ad-
justment Performance." Asia Regional Series, Report IDP 39. World Bank,
Washington, D.C.
Ahmed, Sadiq, and Ajay Chhibber. 1989. "Reconciling Non-Inflationary Growth
and Debt in Indonesia." Asia Region, World Bank, Washington, D.C.
Ahmed, Sadiq, and Basant Kapur. 1989. "A Short-Run Econometric Model of
the Indonesian Monetary System." World Bank, Jakarta, Indonesia.
Balino, Thomas, and V. Sunderarajan. 1986. "Financial Reform in Indonesia:
Causes, Consequences and Prospects." In Hang-Sheng Cheng, ed.,
Financial Policy and Reform in the Pacific Basin. Lexington, Mass.: Lexington
Books.
Chamley, Christophe, and Quaiser Hussain. 1988. "The Effects of Financial Lib-
eralization on Thailand, Indonesia, and the Philippines: A Quantitative Eval-
uation." PPR Working Paper. Policy, Planning, and Research, World Bank,
Washington, D.C.
Cohen, Daniel. 1988. "The Management of the Developing Countries' Debt:
Guidelines and Applications to Brazil." World Bank Economic Review 2
(1):77-103.
Collier, William, and others. 1988. "Employment Trends in Lowland Javanese
Villages." U.S. Agency for International Development, Jakarta, Indonesia.
Fry, Maxwell J. 1988. Money, Interest, and Banking in Economic Development.
Baltimore, Md.: Johns Hopkins University Press.
Gelb, Alan, and associates. 1988. Oil Windfalls: Blessing or Curse? New York:
Oxford University Press.
Kincaid, Russell. 1983. "The Impact of the 1978 Exchange Rate Adjustments
in Indonesia on the Non-Oil Trade Account." DM/83/60. International Mon-
etary Fund, Washington, D.C.
Manning, Christopher, and others. 1988. "Indonesia: Employment Problems,
Prospects, and Policies." World Bank, Jakarta, Indonesia.



382 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
McKinnon, Ronald. 1973. Money and Capital in Economic Development. Wash-
ington, D.C.: Brookings Institution.
Papanek, Gustav. 1988. Changes in Real Wages: Employment, Poverty, Com-
petitiveness. DSP 24. Boston, Mass.: Boston Institute for Developing Economies.
World Bank. 1989. World Debt Tables 1989-90. Washington, D.C.
Comments
William H. Branson
SADIQ AHMED presents a careful, balanced, and accurate account of In-
donesia's adjustment to the fall in oil prices in the mid-1980s and the
subsequent liberalization of its trade and industrial policies. In general,
I agree with Ahmed's analysis and with the judgment that Indonesia is
a successful case in terms of both adjustment and the interaction with
the World Bank in the adjustment process. My discussion will focus on
Indonesia's macroeconomic adjustment policy as an example for the ap-
plication of the classic analysis of policy for external and internal balance
associated with Salter (1959) and Swan (1963). I will also discuss briefly
the desirability of an investment pause during a structural adjustment
program, as was observed in Indonesia. Before turning to my analysis,
however, I will dispose of the two major points of disagreement I have
with the chapter.
The first point concerns the effect of dollar depreciation on debt service,
which, Ahmed argues, "significantly increased Indonesia's debt service
payments." This effect is included in the exchange rate effect in table 17-
1. This interpretation seems to reflect a kind of accounting illusion. If
the diversification of Indonesia's debt across currencies reflects the pat-
tern of its export earnings, movements of the dollar against the European
Currency Unit or the yen should have no effect on the burden of the debt
service. If for historical reasons the debt is not so diversified, the service
can be via interest rate swaps. If Indonesia does not diversify its debt
service, it is speculating on movements in exchange rates.
The second point concerns the assertion that an external shock is man-
ifested initially in a balance of payments disequilibrium. The recent lit-
erature on adjustment shows that this is the case only if the disturbance
is perceived to be temporary. But if, for example, the disturbance comes
from a variable that is perceived to follow a random walk over time,



Indonesia: Stabilization and Structural Change 383
such as the exchange rate, then each change is permanent. In this case,
absorption would adjust immediately to the income effect of the dis-
turbance, and no change would be observed in the balance of payments.
This is a standard result from intertemporal maximizing models such as
that of McKibbin and Sachs (1988).
Now I turn to a brief examination of Indonesia's macroeconomic pol-
icy, which provides a good example for the application of the classic
Salter-Swan analysis of policy for internal and external balance in a small
open economy with a pegged exchange rate. Indonesia has internationally
open capital markets and an exchange rate pegged at any point in time
to a currency basket. If we take low inflation as the target of internal
balance and stable foreign exchange reserves as the target of external
balance, we can apply the Salter-Swan analysis, in which fiscal policy is
aimed primarily at internal balance and monetary policy at external bal-
ance. With a mild increase in inflation following the 1986 devaluation,
and some apparent loss in reserves in 1988, this approach would call for
a tightening of both fiscal and monetary policy. This has been the di-
rection of policy in Indonesia since 1987.
In the Salter-Swan analysis, fiscal policy is aimed primarily at internal
balance, represented here by a low rate of inflation. Indicators of domestic
inflation are shown in table 17-6. The relatively high inflation rates of
the early 1980s were reduced to 5 to 7 percent on the general indicators
in 1985 and 1986 as the primary budget deficit was reduced. The discrete
nominal devaluation in September 1986 raised the prices of traded goods
in late 1986 and 1987, but the inflation rate on price indexes for export
and import goods slowed in 1988 to about 12 percent.
The single best indicator for internal balance is probably the inflation
rate for nontraded goods. With the prices of imported inputs and rice
rising, and some shift of demand toward nontraded goods with the de-
valuation, prices of nontraded goods rose by about 8 percent in 1987.
But even with some pressure from the rice price, inflation for nontraded
goods was down to about 5 percent in 1988. This shows that the de-
valuation of 1986 changed the price ratio of traded to nontraded goods
and was not passed through as a general increase in prices. In terms of
maintenance of internal balance, the tightening of fiscal policy (see table
17-4) was the complement to the devaluation that prevented its spread
to the general price level.
The proximate target for monetary policy in the Salter-Swan analysis
for Indonesia is external balance, represented by the maintenance of the
desired foreign exchange reserve position. This policy objective is re-
quired by the combination of open capital markets and a pegged exchange
rate. Although the exchange rate is adjusted against a currency basket
daily, it is not floating. At any point in time, the authorities are committed
to the existing nominal rate against the dollar. In this situation, monetary



384 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
policy must be tight enough to keep interest rates high enough that the
private capital inflow finances the current account deficit less official net
borrowing-that is, tight enough to maintain the desired level of reserves.
With a stable demand for money that is a function of real income, the
price level, expected inflation, and the interest rate, the given values for
the first three variables and a proximate target for the interest rate set
the demand for money. This, in turn, sets the growth in the domestic
component of the monetary base that is consistent with stable foreign
exchange reserves. Indonesian monetary policy is a good example of the
monetary component of the classic analysis. This is discussed clearly in
chapter 17.
The rise in the ratio of quasi-money to GDP since 1984 (implicit in table
17-7) presumably reflects the financial reform that began then and the
reduction of inflation. The continued growth in the ratio of quasi-money
to GDP since 1986 is also consistent with the public's expectation of con-
tinued low inflation despite the devaluation. Thus the movement of the
monetary aggregates is consistent with a gradual tightening of policy,
combined with a growing public demand for deposits in the banking
system. The only modification of this policy stance that might be sug-
gested is the maintenance of a sufficient real international interest dif-
ferential to maintain the central bank's net foreign asset position.
Finally, I will briefly discuss the investment response to structural ad-
justment. Adjustment programs in general are attempts to increase the
efficiency of the economy, and in general the programs seem to have
succeeded. In chapter 11 Riccardo Faini and his associates, using a cross-
section regression analysis, find significantly better growth performance
in countries with adjustment programs than in those without, and they
also find that the amount of World Bank-IMF lending contributes sig-
nificantly to the result. Indonesia is an example of a country that has
improved its growth performance with adjustment, however incomplete.
Faini and his associates also find, however, that the adjusting countries
have lower investment/GDP ratios than nonadjusters and that the invest-
ment ratio is significantly and negatively related to the amount of lending.
In Indonesia private as well as public investment fell during the adjust-
ment period, as suggested by table 17-11. How should we interpret this
investment response?
First, consider the extreme case of costless adjustment in a neoclassical
growth model. If the adjustment simply rotates the per capita production
function up, the golden-rule savings rate will fall as some of the windfall
is absorbed by a reduction in investment. The example is greatly strength-
ened if there are adjustment costs for the current population. The ad-
justment presumably puts the economy on a higher growth path, which
will provide benefits to future generations, but the costs are being borne
by the present generation. The costs can be spread equitably into the



Indonesia: Stabilization and Structural Change 385
future if expenditure cuts come out of investment instead of current con-
sumption. This argues for the intertemporal optimality of an investment
pause during the adjustment program, if not a permanent reduction in
the investment ratio.
This intertemporal efficiency argument may be implicitly recognized
by the political attractiveness, if not necessity, of maintaining consump-
tion during the adjustment program. The political justification for the
program is that it will provide future benefits. It seems reasonable for
the political process to require that the present population be compen-
sated for their efforts in behalf of the future.
The pause in private investment is likely to be motivated by the tem-
porary increase in uncertainty that is caused by the program. In no case
is a complete program announced and credibly implemented immedi-
ately. Decisions are taken over time and frequently against some political
opposition. As a result, some degree of secrecy exists in the political
decisionmaking process, which adds to the genuine uncertainty about the
best course of the program over time. If this process is understood by
private investors, their best course is to wait until the uncertainty is re-
duced to its normal level. So the investment pause is rational from the
private investor's perspective, and is socially rational as well.
An example of this investment response is provided by the Indonesian
adjustment in the 1980s, as described by Ahmed. Growth in real private
consumption slowed during the stabilization period until 1987, but much
less than growth in national income. The growth rate of private real
consumption per capita bottomed at 1.3 percent in 1987. Total fixed
investment actually fell, while private fixed investment was flat from
1981 to 1986. The private investment slump ended in 1987, and private
fixed investment in real terms grew 7.6 percent in 1988, as the shape of
the adjustment program became clear. The Indonesian government was
clear on the need not to squeeze consumption excessively in the adjust-
ment period, and the policy seems to have succeeded.
References
McKibbin, Warren, and Jeffrey Sachs. 1988. "Coordination of Monetary and
Fiscal Policies in the OECD." In J. Frenkel, ed., International Aspects of Fiscal
Policies. Chicago, Ill.: University of Chicago Press.
Salter, W. 1959. "Internal and External Balance: The Role of Price and Ex-
penditure Effects." Economic Record 35 (August):226-38.
Swan, T. W. 1963. "Longer-Run Problems of the Balance of Payments." In H.
W. Arndt and W. M. Corden, eds., The Australian Economy: A Book of Read-
ings. Melbourne: Cheshire Press. (Reprinted in R. E. Caves and H. G. Johnson,
eds., Readings in International Economics. Homewood, Ill.: Irwin.)



18 Korea: Successful Adjustment
Mansoor Dailami
IN MANY RESPECTS, the process of structural adjustment, which was pur-
sued with varying degrees of intensity by many indebted developing coun-
tries in the 1980s, finds its clearest expression in the experience of the
Republic of Korea. The economy adjusted to the global shocks of the
1980s with remarkable promptness and success. Economic growth re-
sumed, and the economy is again advancing at a double-digit rate. In-
flation was sharply reduced from an average of 13.7 percent in the 1970s
and a peak of 29 percent in 1980 to about 7 percent for 1988. The balance
of current account underwent a significant turnaround from a deficit of
US$5.3 billion (9.4 percent of GNP) in 1980 to a hefty surplus of US$10.0
billion (8.3 percent of GNP) in 1987, and the outstanding level of external
debt was reduced by about US$16 billion from its peak of US$48 billion
in 1980 (see table 18-1).
These achievements relative to those of other debtor countries place
Korea in a unique position that can be compared only with its own
"economic miracles" of the 1960s and 1970s. Yet neither the severity of
the economic problems faced by Korea at the beginning of the 1980s nor
the sagacity of the policy measures undertaken in subsequent years should
be underestimated. This chapter describes the crisis of the early 1980s,
reviews the policy measures adopted, evaluates the role of the World
Bank adjustment lending programs formulated in the context of two
structural adjustment loans (SALS) and two industrial finance projects and
attempts to draw policy implications for other indebted countries.
Background: Causes and Origins of Crisis
When the global shocks of the early 1980s hit the Korean economy, it
was already suffering from several structural and macroeconomic im-
balances. These imbalances were reflected most noticeably in high and
rising inflation, tight labor market conditions, overvalued exchange rates,
deteriorating external competitiveness, high excess capacity in the man-
ufacturing sector (particularly in heavy and chemical industries), low
business profitability, high corporate indebtedness, and segmented and
overregulated financial markets.
386



Table 18-1. Macroeconomic and Balance of Payments Developments in Korea, 1970-74 to 1988
Indicator                                 1970-74           1975-79          1980          1985          1986           1987          1988
Real GNP growth (percent)                     8.9              9.8           -5.2            5.4          12.3          12.04         12.2b
Rate of inflation (cpi, percent)             13.7             16.7            28.7           2.5           2.8           3.0           7.1
Current account balance
>     Billions of U.S. dollars                   -0.8             -1.5            -5.3          -0.9           4.6           9.9           13.0 '
NI    Percentage of GNP                          -6.6             -3.8            -9.4          -1.0           4.7           8.3a          n.a.
External debt (billions of U.S. dollars)      n.a.             n.a.           29.8          48.0          45.1          35.6          32.0'
n.a. Not available.
a. Estimate.
b. Based on data for three quarters only.
Sources: Bank of Korea (various years); Far Eastern Economic Review, February 9, 1989; and World Bank (1987b).



388 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Domestic Factors
The sources and causes of these imbalances were varied, but there is
growing agreement that most could be traced to the rapid growth strategy
of the 1960s and 1970s and, most immediately, to the nature of the
adjustment to the oil price increase of the mid-1970s. The response to
the oil shock was to expand aggregate demand, both internal and ex-
ternal, and thereby to maintain the growth momentum of the 1960s.
Between 1973 and 1978 real GDP grew by an average of 10 percent each
year, exports by 25 percent, and investment by 21 percent. This strategy
avoided interruption in the growth of per capita income, which tripled
in real terms between 1960 and 1978.
This vigorous economic growth was accompanied, however, by a sub-
stantial shift in industrial structure and resource allocation in the second
half of the 1970s. In anticipation of continuing domestic growth and
rapid expansion of international trade, the authorities changed course in
the second half of the 1970s and promoted large-scale investment in basic
chemicals and heavy industries such as steel, ship building, and the man-
ufacture of heavy equipment. The goal of these investments was to re-
structure the composition of exports and move toward more sophisti-
cated and skill-intensive products. But as demand for these industries'
products fell below planned targets, substantial excess capacity emerged.1
At the same time, this overinvestment in heavy and chemical industries
deprived traditional and light industry exports of necessary investment
funds, which in turn weakened their external competitiveness.
The industrial restructuring and the associated investment boom ex-
erted unsustainable strains on the country's resources, which were re-
flected in tight labor market conditions and accelerating inflation. Also
contributing to the tightness of labor markets was the departure of a
large number of skilled construction workers to the Middle East oil-
exporting countries. In response to this development, increases in real
wages exceeded gains in productivity by a wide margin, which further
aggravated inflationary pressures. And as inflation gained momentum,
it threatened external competitiveness and economic growth. Further-
more, with the exchange rate fixed after the devaluation of 1974, the
won became progressively overvalued and export growth weakened. In
fact, export volume, which had grown at an average annual rate of 27
percent over the previous two decades (1960-78), slowed in 1978 and
actually fell by 1 percent in 1979.
The widening structural imbalances in the Korean economy and the
ensuing macroeconomic problems originated with the government's basic
strategy of deploying finance as a mechanism for channeling resources
to priority investment sectors. This strategy, which involved both direct
lending to industry through the National Investment Fund and subsidi-



Korea: Successful Adjustment 389
zation of the cost of capital via interest rate controls and fiscal incentives,
was a hallmark of the Korean approach to industrialization in the 1960s
and the 1970s. By exercising control over the funds available to the
private business sector, particularly the corporate sector, the government
became, in essence, an active partner in the finance process.
Two aspects of this partnership were particularly significant: the sys-
tem provided funds to the priority sectors at preferential rates and, prob-
ably more important, transferred much of the risk of long-term invest-
ment to the government. By investing in a project that had the
government's blessing, a firm could anticipate a stable and subsidized
flow of funds, irrespective of its economic and financial performance.
This assurance of funds reduced the cost of financial distress by reducing
or even eliminating the risk of bankruptcy.
This interventionist approach to financial allocation seems to have
served its objective of increasing industrial investment in the 1960s and
1970s. Over time, however, it led to structural imbalances in the financial
markets. First, it made the financial institutions, particularly those in the
banking sector, dependent on the government's directives. Second, it led
to wide differentials in the cost of capital for various sectors and firms
of different sizes. Third, it retarded the development of capital markets
for the supply of long-term and risk capital. In a sense, the risk part-
nership between the government and the private sector functioned as a
substitute for organized securities markets.
External Shocks
Because of these widening domestic structural and macroeconomic im-
balances, the series of adverse global events of the early 1980s had a
particularly profound effect on the Korean economy. The country was
so dependent on exports and on imports of essential raw materials and
energy products that its external payments position was dealt a major
blow by the oil price increases of 1979 and 1980 and the subsequent
recession in countries of the Organisation for Economic Co-operation
and Development (OECD). Summary indicators of this deterioration in
the external environment for 1976-82 (table 18-2) reveal the evolution
of three adverse factors: (1) a substantial decline in the terms of trade
of about 28 percent from 1979 to 1981; (2) a sharp increase in inter-
national interest rates, from an annual average of 6.8 percent for the
three-month Eurodollar rate during 1976-78 to 11.6 percent during
1979-82; and (3) a significant slowdown in economic growth among
OECD countries, Korea's major trading partners, from an average annual
real GDP growth rate of 4.1 percent during 1976-79 to an average of 0.9
percent in 1980-82.



390 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 18-2. Summary Indicators of the External Environment Facing
Korea, 1976-82
Real GDP
International          growth
Terms of              interest rate            in
trade index           (three-month             OECD
Year           (1980 = 100)           Eurodollar rate)        countries
1976              105.1                   5.57                   5.0
1977              112.4                   6.0S                   3.9
1978              117.8                   8.85                   4.0
1979              115.3                   2.09                   3.3
1980              100.0                  14.19                   1.3
1981               97.9                   16.87                  1.6
1982              102.2                  13.29                 -0.2
Sources: Bank of Korea (various years); IMF (1988); and OECD (various years).
The most immediate and direct impact of these external shocks on the
Korean economy was a sharp deterioration in the balance of current
account. The current account deficit, which had averaged less than
US$0.5 billion during the 1976-78 period, rose sharply to US$4.15 bil-
lion (6.9 percent of GNP) in 1979 and to a peak of US$5.3 billion (9.4
percent of GNP) in 1980. Not all of this deterioration can be attributed
to external shocks, however. Domestic factors, notably an estimated 23
percent appreciation of the won in real terms during the 1974-79 period
also played an important part. A determination of the relative factors at
work is not easy, but an analysis of the counterfactual, based on the usual
methodology, yields some useful insight by providing a quantitative ac-
counting of the extent of external shocks and the nature of adjustment
responses (see table 18-3). The analysis verifies the substantial impact of
external shocks on the balance of current account. In 1980 in particular,
when various shocks converged to hit the economy with full force, almost
all the deterioration in the current account could be attributed to' the
impact of external shocks.
The analysis also provides some insights into the nature of adjustment
responses and their effect on the current account balance. In 1980, for
instance, adjustment in the current account took place primarily through
a combination of import compression (14 percent decline) and export
expansion (11 percent increase in real terms). The nature of adjustment
changed fundamentally in 1981, however, when export expansion was
large enough to balance the current account and also to permit some
increase in imports. Thus, unlike other debtor developing countries which
continued to use import compression to achieve external balance, Korea
succeeded in weathering the external shocks of the early 1980s through



Korea: Successful Adjustment 391
Table 18-3. External Shocks and Current Account Adjustment in
Korea, 1979-82
(percentage of GNP)
Contributing factor          1979         1980          1981         1982
Exogenous factors            -1.3         -9.4          -3.2         -2.2
Oil price                  -1.9         -4.4          -1.8         +0.3
Foreign interest rates     -0.7         -1.1          -1.1         +0.8
Nonoil terms of
trade and other'         +1.3         -3.9          -0.3         + 1.1
Adjustment efforts           - S.6        + 7.3         + 4.3        + 1.0
Import volume              -9.2         +4.2          -1.3         -1.4
Export volume              +2.4         +3.0          +4.9         + 1.8
Otherb                     +1.2         +0.1          +0.7         +0.6
Total change in current
account deficit          -6.9         -2.1          +1.1         +3.2
Note: A negative sign indicates an increase in the current account deficit; a positive sign indicates
a decrease.
a. Includes the effect of extraordinary rice imports.
b. Mainly the change in net receipts from overseas construction.
Source: World Bank estimates.
strong expansion of exports. The ability to sustain the growth of imports
after 1980 facilitated the launching of broader adjustment measures to
restructure the industrial sector and liberalize the trade regime.
The Adjustment Program
The government began the process of adjustment in Korea with measures
to stabilize the economy by decreasing inflation. It then moved on to
policy reforms and structural adjustment.
Stabilization Measures: The Disinflation Process
In the design of stabilization measures in 1979-80 the government gave
high priority to restoring price stability and improving external com-
petitiveness by reducing inflation. Inflation was not a new phenomenon
for the Korean economy. It had averaged 13.5 percent in the 1960s and
16 percent in the 1970s and has shown a high degree of cyclical sensitivity
to external factors, particularly international primary commodity prices
(see figure 18-1). What distinguished the inflation of 1977-79 from ear-
lier episodes was its dynamics and the belief that it was mostly home-
made. There was growing evidence that the expansionary aggregate de-
mand policies of those years, financed in part by accumulating foreign
debt, had exerted excessive pressure on the labor market and pushed real



Figure 18-1. Trends in Inflation and the Import Price Index in Korea:
Annual Percentage Rate of Change
Percent
60 -
Import price index, in
/  \domestic currency
50 -                                                     A
40-
30 - 
20         I
10              h   olesale       o
price index   I-
0                                   -
1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987
Source: Author's calculations based on data from Bank of Korea (various years).



Korea: Successful Adjustment 393
wages far ahead of gains in productivity. Growth in real wages in the
manufacturing sector, for instance, averaged 16.25 percent between 1976
and 1979, compared with a gain in productivity of 12.5 percent over the
same period. In addition, the exchange rate had been fixed after the
devaluation of 1974, and in this environment of rising inflation it led to
considerable overvaluation of the currency, amounting to 23.6 percent
between 1973 and 1979 (Corbo and Nam forthcoming).
The inflation of the 1977-79 period prompted an important shift in
the management of macroeconomic policy. The goal was no longer sim-
ply to promote growth but to promote growth with price stability. The
additional requirement of price stability posed a serious challenge to
policymakers. Given the overheated state of the economy, some degree
of demand contraction seemed to be necessary, which meant curbing the
growth of real wages. But such a policy had to be both temporary and
perceived as only a brief aberration in the government's traditional ap-
proach to macroeconomic management and economic growth, which had
been so successful. In the past, Korea's strategy had been accompanied
by rising real wages, made possible by rapid increases in labor produc-
tivity. To restore such trends in the late 1970s required a resumption in
the growth of productivity, which in turn depended on keeping invest-
ment high. Encouraging investment, however, conflicted-at least at first
glance-with the objective of lowering inflation.
Another aspect of the policy dilemma was how to address the dete-
rioration in trade competitiveness resulting from the rigid exchange rate
policy and high inflation at home. Currency devaluation was one option
for reversing the downward trend in export volumes, but it would be at
the cost of intensifying inflationary pressures, at least in the short run.
Trying to improve competitiveness by relying on demand contraction to
reduce inflation would also be costly in terms of forgone output and
employment.
The policy response adopted by the authorities consisted of a mix of
currency devaluation and contractionary demand measures, with a rela-
tively heavy reliance on monetary policy. The won was devalued by 17
percent in January 1980, and monetary policy was tightened. The initial
impact of the devaluation was a rapid surge in the rate of inflation, which
peaked at 28.7 percent in terms of the consumer price index (cpi) and
38.9 percent in terms of the wholesale price index (wpi) by the end of
1980. At the same time, the tight monetary policy dampened domestic
demand so much that in 1980 aggregate private investment declined by
11.5 percent while investment in machinery and equipment registered an
even steeper decline of about 25.5 percent (see table 18-4).
Although these measures contributed to the sharp decline in real ag-
gregate output in 1980, they also brought about several improvements;
in particular they reversed inflationary trends and restored external com-



Table 18-4. Responses of Macroeconomic Policy Indicators in Korea, 1978-88
(percentage change)
Indicator                      1978      1979      1980      1981     1982      1983      1984      1985      1986      1987      1988
Growth in money supply (M2)     35.0     24.6       26.9      25.0    27.0       15.2       7.7      15.6      18.4     19.1     21.3a
Real growthb                  11.0      3.9        1.3       8.3     19.3      10.9       3.7      11.1      15.3     14.8      n.a.
Growth in credit'               47.1     39.2       39.6      26.3    25.1       17.6      14.3      19.0       7.0     12.0      n.a.
Real growthb                  21.4     16.0       11.5       9.4     17.5      13.2      10.0      14.4       4.2      8.1      n.a.
Real effective exchange rate
index (1980 = 100)         104.5     97.1      100.0      96.0    95.1    100.9    103.6        110.1    128.5       n.a.     n.a.
4     Rate of inflation
Consumer price index          14.4     18.3       28.7      21.3      7.2       3.4       2.3       2.5       2.8      3.0       7.1
Wholesale price index         11.6     18.8       38.9      20.4      4.7       0.2       0.7       0.9    -1.5        0.5       2.7
Growth in real investment       29.8     15.9     -20.7      -6.3       0.1      17.5      18.6       1.6      10.6     15.1      n.a.
Private                       34.0      7.7    -11.5       -4.8      12.4      17.2      10.4       4.3      18.6      n.a.     n.a.
Machinery and equipment       59.9     22.0     -25.5      -2.8       4.3       8.8      15.6       5.2      30.3     16.6      n.a.
n.a. Not available.
a. Preliminary estimate.
b. Deflated by GDP deflator.
c. Total credit to private sector.
Sources: Bank of Korea (various years); Nam (1988).



Korea: Successful Adjustment 395
petitiveness. By 1982 inflation had been lowered to 7.2 percent (in terms
of the cPI), and a recovery in output and investment was well under way
(see table 18-4). Inflation declined further in the next two years, reaching
a low of 2.3 percent in 1984, and remained remarkably stable until 1987
despite vigorous expansion in real aggregate investment and output. Since
1988, however, inflationary pressures have gained momentum and are
likely to become a matter of major policy concern again. Nevertheless,
the decade of the 1980s was a period of sharp disinflation in Korea,
which is comparable, in many respects, to the experience of major in-
dustrial countries.
What factors contributed to Korea's achievement of low inflation in
the 1980s? Presumably, both internal and external factors played im-
portant roles; the evidence, however, is more firmly on the side of external
factors, as illustrated by the results of equations 1 through 9 reported in
table 18-5. The equations are based on a simple model of the inflationary
process in Korea: the rate of inflation in terms of the wpi is explained as
a function of lagged inflation, the current and lagged import price index
in domestic currency, the rate of growth of the money supply (M2), and
the rate of growth of real GNP. The model was estimated by ordinary
least squares, using annual data from 1971-87 with several lag speci-
fications. The results are familiar: inflation in Korea was strongly influ-
enced by external inflation and by changes in the exchange rate. Statis-
tically, the estimated coefficient of the import price index (in terms of
domestic currency) remains significant and robust throughout all esti-
mated equations, with a short-run elasticity of around 0.5.
Two domestic factors seem of particular importance in Korea's dis-
inflation of the 1980s: (1) the strong performance of business investment,
particularly in machinery and equipment, which helped to keep produc-
tive capacity in pace with the growth in demand, and (2) the appreciation
of the nominal exchange rate. Investment in machinery and equipment
averaged 17 percent a year in 1984-88 in real terms, while the exchange
rate appreciated in nominal terms by 21 percent during the same period.
From Stabilization to Adjustment
The prompt upturn in aggregate investment and output in 1981, coupled
with lower inflation and a flexible exchange rate policy, provided the
macroeconomic framework for Korea's broad adjustment efforts encom-
passing industry, trade, energy, and finance. With hindsight, the roughly
two-phased sequencing of these adjustment efforts seems to have been a
desirable ordering of reform measures.
The first phase (1981-85) focused on the real side of the economy.
The goal of policy was to foster industrial reform, to reduce Korea's high
energy consumption and its heavy dependence on petroleum imports, to



Table 18-5. Regression Results for Determinants of Inflation in Korea, 1971-87
Explanatory                                                           Equation
variable            1            2            3             4            5             6            7            8              9
Constant           -0.07       - 0.07        -0.05        -0.07         -0.09         -0.07         0.00        -0.08         -0.09
(-1.50)      (1.52)      (-0.84)       (-1.04)       (-1.56)       (-1.48)       (0.14)      (-1.46)       (-1.62)
P_,                 0.21        0.40          0.44          0.40          0.33          0.45        0.31          0.23          0.36
(0.70)      (2.42)       (2.42)        (2.31)        (1.63)        (3.21)      (2.47)        (1.84)        (1.90)
P-1                 0.12        0.09          0.12          0.10          0.07 
(0.79)      (0.61)       (0.79)        (0.58)        (0.44)       -            -            -              -
PM                  0.51        0.56          0.57          0.56          0.52          0.54        0.57          0.49          0.50
(4.64)     (6.28)        (6.12)        (5.33)        (4.91)        (6.66)      (6.62)       (5.35)         (5.39)
O  PM1                    0.16        -            -            -             -             -            -            --
(0.72)
0.74        0.70         0.79          0.72          0.50          0.74        -            -              0.50
(1.64)      (1.60)       (1.68)        (1.52)        (0.88)        (1.78)      -            -              (0.91)
M2                  -           -            -0.19         -             _             -            _            _
(-0.65)
M2_                -            -            -            -0.05          -                                       -             -
(-0.15)
M2_2               -            -            -             -              0.22         -           -              0.46          0.26
-            -           -             -             (0.59)        -            -            (1.67)        (0.74)
R2                  0.88        0.87          0.88          0.88          0.88          0.87       0.83           0.87          0.88
- Variable not included in the equation.
P = wholesale price index, PM = import price in local currency, Y = real GNP, M2  money supply, and = percentage rate of change; t statistics are in parentheses.



Korea: Successful Adjustment 397
reduce import barriers, and to promote efficiency in the public sector.
These initiatives were supported by two structural adjustment loans from
the World Bank (SAL I in December 1981 and SAL 11 in November 1983)
and a one-year standby arrangement with the International Monetary
Fund (IMF) in February 1982. In the second phase (1983 to the present),
attention turned to the financial sector, where the government has fol-
lowed a cautious liberalization strategy directed toward deregulation of
interest rates, enhanced competition, and a gradual shift from direct in-
tervention in credit allocation and monetary policy to indirect interven-
tion through reserve requirements, open market operations, and market-
determined interest rates. These reforms, which were initially supported
by two industrial finance project loans from the World Bank June 1983
and May 1985), have accelerated since the emergence of the current
account surplus in 1986.
Policy Reform and Structural Adjustment
Concerns about the sustainability of economic growth and the manage-
ment of the balance of payments in Korea have traditionally focused on
the performance of industry-particularly the manufacturing sector be-
cause of its role as the engine of growth and its rapid structural change
from light to heavy industry during the 1970s. Thus a key objective of
both SALS was to support the government's promotion of industrial ef-
ficiency and productivity. To this end, a wide range of medium-term
financial, trade, and industrial reforms were initiated to reduce the direct
role of the government in the finance process, to provide access to foreign
capital and technology, and to expose industry to greater external com-
petition. These reforms were intended to stimulate policy actions that
would reduce government direction and control over credit allocation at
the enterprise level. Other reforms included relaxation of controls gov-
erning interest rate ceilings, liberalization of the import regime by 1986
to the level prevailing in industrial countries, reduction in the spread of
tariff rates, and relaxation of foreign investment provisions.
The financial components of policy reform were strengthened and ex-
panded at a later stage under the two industrial finance projects. These
reforms included liberalization of commercial banking regulations and
supervision, control of corporate indebtedness, and development of an
equity market. Collectively, financial reforms were intended to address
the main structural weaknesses in the financial sector.
DIRECT AND SUBSIDIZED CREDIT. Although much of the financial reform
took place at a later stage of the adjustment process, two measures were
initiated under the two SALS: government direction of credit allocation
was reduced and interest rates on controlled funds were raised to the



398  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
level of the minimum rates on commercial bank lending. Directed lending
programs through the National Investment Fund and via the rediscount-
ing facilities at the Bank of Korea were the main instruments of direct
government intervention in the allocation of credit to industry. Loans
arranged through the National Investment Fund were used to finance
long-term investment in machinery and equipment and were targeted to
high-priority sectors, such as chemicals, power, and heavy industry. For
example, of the 4,445 billion won (W) in total loans made over the 1974-
84 period, about W2,714 billion, or 61,percent, was earmarked for the
heavy and chemical industries (World Bank 1987a).
The other major instrument of government-directed credit allocation
was the use of rediscount facilities at the Bank of Korea. Gains from the
use of these facilities contributed to commercial bank revenues and pro-
vided important incentives to extend credit to priority sectors. Depen-
dence on these facilities also rendered the commercial banks' balance
sheets dependent on government finance- and hence dampened the com-
mercial banks' initiative and reduced their managerial autonomy.
When viewed from the perspective of the SAL period only, financial
sector reform appears to have been slow. When viewed from a longer
perspective that includes the two industrial finance projects, however,
progress is considerable. Among the important achievements are the fol-
lowing:
* Since 1982, credit allocations through the National Investment Fund
have been made on a sectoral basis rather than by targeting specific
firms. Banks handling these loans therefore now have greater auton-
omy in credit evaluation and decisionmaking.
* The volume of loans arranged through the National Investment Fund
was reduced from W733 billion in 1983 to W506 billion in 1984.
. Even more important, net contributions to the National Investment
Fund were reduced from W444 billion in 1983 to W73 billion in 1984.
During 1984 and 1985 the government gradually allowed interest rate
ceilings to rise on both lending and deposit rates.
- The government promoted domestic savings by encouraging the in-
troduction of new financial instruments such as negotiable certificates
of deposit and cash management accounts.
IMPORT POLICY REFORM. The reform of import policy has been a key
aspect of the adjustment program since 1981 with the support of the two
SALS. The process accelerated after the 1986 turnaround in Korea's cur-
rent account position. Major reforms include liberalization of import
policy, tariff reduction, and the opening up of financial services, including
banking, insurance, and securities.



Korea: Successful Adjustment 399
Government efforts to expose industry to greater external competition
have involved a reduction in nontariff barriers and in the rates and dis-
persion of tariffs. Automatically approved imports as a percentage of
total imports (the import liberalization ratio) progressively increased
from 69 percent in 1980, to 80 percent 1983, 93.6 percent in 1987, and
94.8 percent in 1988 (Korea Ministry of Trade and Industry 1988). Im-
port liberalization also included phased general reductions in tariff rates
and greater uniformity of rates. In compliance with the Tariff Act of
1984, the average (arithmetic) tariff rate was reduced from 31.7 percent
in 1982 to 21 percent in 1985 and 18.1 percent in 1988, with a further
reduction to 7.5 percent planned by 1993. At the same time, the gov-
ernment continued sliding the tariff schedule downward to narrow the
dispersion in the rates. The targets for 1988 were uniform rates of 5 to
10 percent for raw materials, 20 percent for intermediate goods, and 20
to 30 percent for capital goods.
INDUSTRIAL RESTRUCTURING: THE CASE OF THE AUTOMOTIVE INDUSTRY.
In the wake of the serious economic difficulties facing the Korean econ-
omy in the early 1980s, including low capacity utilization in heavy in-
dustry and high oil prices, the World Bank took a pessimistic view of
the prospects for expansion of the automobile industry. Several prelim-
inary studies in 1980-82 raised serious questions about the rationale
behind this expansion. There were grave doubts about the industry's
ability to penetrate foreign markets, particularly in developed countries,
because of competitive price structures, sophisticated marketing net-
works, and the need for advanced product design. For these reasons and
in view of the general conditions of excess investment and declining do-
mestic and export demand, the World Bank questioned the wisdom of
the government's directed investments in the automobile industry.
Accordingly, under SAL I the government agreed to defer its support
for the expansion of automobile plants pending completion of a study
reviewing export prospects. Because a relatively higher comparative ad-
vantage was thought to exist in the spare auto parts industry, the World
Bank recommended that a task force prepare a plan for promoting this
industry as a substitute for or complement to the automobile industry.
As events unfolded, however, the outlook for the automotive industry
in Korea became far more positive than earlier conditions had led the
World Bank to assume.2 The firm determination of business leaders and
a recovery in economic activity in the years following the negotiation of
SAL I helped to alleviate the World Bank's concern. Consequently, this
issue was not followed up under SAL ii eighteen months later. Discussion
of the related issue of restructuring the heavy machinery sector was also



400  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
dropped under SAL II, and attention shifted to other issues, including trade
liberalization and energy efficiency.
PROMOTION OF ENERGY EFFICIENCY. A combination of a relatively high
energy to output ratio and heavy dependence on petroleum imports in
the late 1970s left the Korean economy particularly vulnerable to the
vagaries of the world oil market. In 1978 energy consumption relative
to GDP was about 60 percent higher in Korea than in other middle-income
developing countries. Among the factors accounting for this high energy
use was Korea's high stage of industrialization, its concentration on en-
ergy-intensive heavy industries, and the relatively heavy use of energy by
the residential sector.3
The main energy-related objectives of structural adjustment measures
under SAL I were to promote overall energy efficiency and to diversify the
sources of energy. More specifically, the goals were to lower the elasticity
of energy consumption in relation to GDP from the 1.3 average during
1975-80 to 0.9 by the late 1980s and to shift from the use of petroleum
toward coal, nuclear energy, and natural gas. To achieve these goals,
Korea planned to bring domestic energy prices in line with the world oil
price increases of 1979 and 1980, to offer financial and fiscal incentives
to promote energy conservation, to cancel or postpone energy-intensive
investment projects, to substitute liquefied natural gas for petroleum
products, and to strengthen legal and institutional structures to promote
energy efficiency. Some of these objectives and proposed measures were
initiated under an IMF standby arrangement and then reinforced under
SAL I. They were further strengthened and broadened under SAL 11 to in-
clude electric power planning and pricing, refinery pricing and trade de-
regulation, and formulation of a long-term strategy for the development
and deregulation of the coal mining industry.
Judged by several standards, progress in conserving energy and diver-
sifying supply has been remarkable. Even before the SALS the government
had demonstrated a bold and strong commitment to the long-term pro-
motion of energy efficiency by raising domestic energy prices in lockstep
with increases in world oil prices. Thus between March 1979 and No-
vember 1980 electricity rates were raised by 36 percent, coal prices by
14 percent, and petroleum prices by a range from 136 percent (diesel
fuel) to as much as 172 percent (gasoline) (see table 18-6). These price
adjustments, reinforced by additional policy measures under the two sAL
programs, achieved impressive reductions in energy intensity and greater
diversity in energy sources. Growth of total energy consumption (mea-
sured in tons of oil equivalent) decelerated from an average of 9.8 percent
in 1976-80 to about 6.2 percent in 1981-87, and the energy/GDP ratio
declined from 1.3 in the late 1970s to 0.7 in the mid-1980s (table 18-
7). With regard to the composition of energy the share of petroleum



Korea: Successful Adjustment 401
Table 18-6. Principal Energy Prices in Korea, 1979-81
(U.S. dollars)
March             November              April
Product                                   1979               1980                1981
Petroleum products (per
gallon)
Gasoline (regular)                      0.53                1.44                1.44
(172)                (0)
Diesel                                  0.39                0.92                1.07
(136)               (16)
Bunker                                  0.36                0.86                0.94
(139)                (9)
Electricity (per kilowatt hour)          0.066               0.090               0.095
(36)                (6)
Coal (per ton)                           34.09               38.92               41.59
(14)                (7)
Note: Numbers in parentheses indicate growth rates.
Source: World Bank data.
declined from 61.3 percent in 1980 to 49.1 percent in 1985, while the
shares of coal and nuclear power rose to 39.2 percent and 6.4 percent,
respectively. These achievements surpassed the targets established under
the SAL-supported adjustment programs.
Table 18-7. Changes in the Composition of Energy Consumption in Korea,
1980-87
Composition                1980    1981    1982    1983    1984    1985    1986    1987
Total energy consumption
Millions of tons of oil
equivalent            44.1    46.1    46.0    49.7    53.9    56.0    61.0           67.1
Growth rate (percent)     -       4.5      0.0      8.04     8.4      3.9      8.9     10.0
Energy use by fuel
(percentage of total
use)
Coal                    29.9    33.1    33.6    33.1    36.8    39.2    37.6           34.2
Petroleum               61.3     58.4    58.0    56.2    52.3    49.1    46.7          47.4
Gas                      1.0      1.1      1.6      2.1      2.3      3.8      4.4      3.1
Hydroelectric             1.1     1.5      1.1      1.4      1.1      1.3      1.6      2.0
Nuclear                  2.0      1.6      2.0      4.5      5.5      6.4    11.6      14.4
Fuelwood                 5.7      5.4      5.3      4.8      4.3      4.0      2.4      2.0
Energy/GDP ratioa          0.77     0.75     0.71-    0.70     0.70     0.68     0.68     n.a.
n.a. Not available.
a. Energy/GDP ratio = tons of oil equivalent/GDP in billions of U.S. dollars.
Source: Korea Economic Planning Board (various years).



402 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
The Bank's Role in the Adjustment Process
The two structural adjustment loans to Korea in 1981 and 1983 were
important vehicles of policy dialogue, financial support, and cooperation.
Both SALS included conditions for the disbursement of tranches, but all
tranches were disbursed on schedule. Given the government's strong com-
mitment to policy reform and its success in macroeconomic management,
however, the loan agreements were free of specific conditionality. As a
result, the government had a considerable degree of policy flexibility and
maneuverability, which proved to be important in the volatile world
economic conditions of the early 1980s. For example, the government's
unwavering support for the expansion of the automotive industry, despite
the World Bank's initial reservations, led to handsome returns as exports
of vehicles grew more than fourfold between 1980 and 1985.
The World Bank policy dialogue with the Korean authorities continued
in the context of the two industrial finance projects in June 1983 and
May 1985. These projects laid the groundwork for later reforms in the
financial sector. The reforms were motivated by Korea's awareness of
its new role in the international financial system and its understanding
that financial sector reforms are as crucial to its transition to the 1990s
as export expansion policies and strategies were to the economic growth
of the 1960s and 1970s.
It is difficult to evaluate the precise influence of the World Bank's
adjustment lending operations and the programs they supported, but
there is little doubt that they were important to Korea's successful ad-
justment experience in the 1980s. Although the two SALS (totaling US$550
million) met 13 percent of Korea's external financing needs during the
period of disbursement (1982-83), other aspects of the programs were
more important than this resource transfer. Of particular importance
were (1) the programs' catalytic role in mobilizing resources in inter-
national financial markets by helping Korea maintain its creditworthiness
and (2) the policy analysis, discussion, and dialogue which underlay loan
operations and agreements. Given their comprehensive and wide-ranging
character, the SAL programs provided a broad framework for the design
and integration of the various components of policy reforms in industry,
trade, energy, and finance. This is an aspect of SAL programs that merits
more attention.
Conclusions and Recommendations
According to the major economic indicators, Korea's stabilization and
adjustment program can be considered a remarkable success. The Korean
economy underwent a fundamental transformation in the 1980s. The
most visible and truly impressive aspects of this transformation relate to



Korea: Successful Adjustment 403
improvements in internal macroeconomic conditions and external bal-
ance, as this chapter has revealed. The extent of this transformation is
evident from a comparison of the present state of robust economic
growth, contained inflation, and a hefty surplus on current account with
the crisis situation in 1980, when the economy was in a deep recession,
aggregate output and investment were declining at unprecedented rates
of 5 percent and 24 percent, respectively, and inflation was soaring.
Although these macroeconomic successes are impressive, Korea's struc-
tural adjustment experience really stands out with respect to its broad
and far-reaching policy reforms in trade, industry, energy, and finance.
Successive efforts to liberalize imports for industrial production achieved
the target import liberalization ratio of 94.8 percent by 1988. The tariff
system was simplified, and the average tariff rate was reduced from 31.7
percent in 1982 to 18.1 percent in 1988, with a further reduction to 7.S
percent planned by 1993, a level that corresponds to that in OECD coun-
tries. Increased exposure to foreign competition and a reduction in gov-
ernment-directed credit, particularly to specific firms, had an important
influence on the industrial sector. The government's perseverance in its
plans to expand the automotive industry despite World Bank misgivings
reaped large benefits for the industry and the economy. In addition, it
made significant advances in reducing the country's growth rate of energy
consumption and in diversifying its sources of energy.
In the financial sector, the government has so far followed a cautious
strategy of liberalization. The strategy is directed toward deregulating
interest rates, enhancing competition, and gradually shifting from direct
intervention in credit allocation and the conduct of monetary policy to
indirect intervention through reserve requirements, open market oper-
ations, and market-determined interest rates.4 Two important consid-
erations influenced the pace and scope of these reforms: (1) the high
degree of corporate indebtedness and (2) the emergence in 1986 of a
substantial balance of payments surplus (amounting to US$13.0 billion
in 1988). Should this trend in the external payments position continue
in the 1990s, it would have significant implications for Korea's rela-
tionship with the international financial community, including the World
Bank and the IMF. Indeed, Korea's recent acceptance of the obligations
of Article VIII of the IMF's Articles of Agreement already marks the be-
ginning of a new era in Korea's international economic position.
Overall, four conclusions emerge from the review presented here. First,
success in the design and implementation of structural adjustment efforts
affecting the functioning of product and factor markets requires more
than the elimination of price distortions and the promotion of competitive
forces. The process also requires continuing efforts to strengthen the
capability of the economy to respond promptly and favorably to market
incentives and signals. Thus changes in both the incentive system and



404 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
the response capability are integral to the adjustment process. Korea's
success in the adjustment process was as much a function of the econ-
omy's capability to respond positively and promptly, as reflected by the
rapid recovery in private investment in 1981 and its steady growth since
then, as it was a function of the government's reforms in the areas of
trade, industry, energy, and finance.
Second, Korea's success in adjustment benefited from a favorable ex-
ternal environment. The country's remarkable achievement in controlling
and lowering inflation depended significantly on the decline in import
prices, including oil prices, which declined from an average of US$40 a
barrel in the early 1980s to US$15 in 1988.
Third, the sequencing of policy measures seems to have been important.
Korea began with stabilization efforts that laid the foundation for sub-
sequent implementation of industrial restructuring and trade liberaliza-
tion, which in turn facilitated the liberalization of the financial sector.
This seems to have been a desirable ordering of reform measures.
Finally, the 1981-85 period during which the SAL-supported programs
were formulated and implemented was characterized by rapid economic
growth, low inflation, high savings rates, and vigorous capital formation.
Also, the balance of payments adjustment undertaken during this period
relied on export expansion and not on import compression.
Notes
1. According to the World Bank's appraisal report for the industrial finance
project in 1983, the capacity-utilization rate declined from a high of 88 percent
in 1978 to 73 percent in 1980 for the manufacturing sector as a whole and from
75 percent in 1979 to less than 50 percent in 1980 for the machinery subsector.
2. Between 1980 and 1984 production in the automobile industry tripled;
production was projected to more than triple by 1990 and reach 1.3 million
vehicles. Similarly, automobile exports grew more than fourfold between 1980
and 1985. Of US$1.7 billion total sales revenue in 1986, more than US$500
million was due to exports.
3. Because of a combination of factors, including weather, space-heating tech-
niques, and low coal prices, energy consumption per household in the late 1970s
was higher in Korea than in Japan. In 1979, for example, the residential and
commercial sectors accounted for about 38 percent of total energy consumption
in Korea compared with 19 percent in Japan.
4. For a detailed description of Korea's recent liberalization of the financial
sector, see Cho and Cole (forthcoming) and Cho (1988).
References
Bank of Korea. Various years. Economic Statistics Yearbook. Seoul.
Cho, Y. J. 1988. "Liberalization of Korea's Financial Market in the 1980s."
Paper presented at the East Rock Institute Conference on the Dynamics of
U.S.-Korea Trade Relations, Yale University, November 10-12.



Korea: Successful Adjustment 405
Cho, Y. J., and D. C. Cole. Forthcoming. "The Role of the Financial Sector in
Korea's Structural Adjustment." In Vittorio Corbo and Sang-Mok Suh, eds.,
Structural Adjustment in a Newly Industrialized Country: Lessons from Korea.
Corbo, Vittorio, and Sang-Woo Nam. Forthcoming. "The Recent Macroeco-
nomic Evolution of the Republic of Korea: An Overview." In Vittorio Corbo
and Sang-Mok Suh, eds., Structural Adjustment in a Newly Industrialized
Country: Lessons from Korea.
IMF. 1988. International Financial Statistics Yearbook. Washington, D.C.: In-
ternational Monetary Fund.
Korea Economic Planning Board. Various years. Major Statistics of the Korean
Economy. Seoul.
Korea Ministry of Trade and Industry. 1988. Trade and Industry Bulletin. Seoul.
Nam, Sang-Woo. 1988. "The Role of Exchange Rate Policy in Four East Asian
Countries." Asia Regional Series, Report IDP 15. World Bank, Washington,
D.C.
OECD. Various years. Main Economic Indicators. Paris: Organisation for Eco-
nomic Co-operation and Development.
World Bank. 1987a. Korea: Managing the Industrial Transition. Vol. 2. Wash-
ington, D.C.
. 1987b. World Debt Tables. Washington, D.C.
Comments
Larry E. Westphal
MY REMARKS will concentrate on the structural adjustment aspects of the
Korean experience. There appears to be some confusion about what
"structural adjustment" means. In my view, there can be little question
that the "structure" at issue is the structure of markets, incentives, in-
stitutions, and other elements that have an important influence on re-
source allocation in both static and dynamic terms. Correspondingly,
policies aim at structural adjustment insofar as they explicitly seek fun-
damental changes in existing modes of allocative decisionmaking. Some
policy measures, by their nature, entail structural adjustment; privati-
zation, when meaningful, is an example. For some policy measures, the
relationship to structural adjustment depends on the context in which
they are undertaken. Real devaluation of the exchange rate, for instance,



406 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
is associated with structural adjustment only insofar as it accompanies
a change in development strategy toward greater openness.
The confusion about structural adjustment may result from a failure
to recognize the importance of contextual elements in determining
whether particular policy measures are or are not related to structural
adjustment. There is general agreement, however, that stabilization pol-
icies typically do not involve structural adjustment in the sense just out-
lined. Their objective is clearly understood to be the improvement of
economic performance in the short run. Structural adjustment, by con-
trast, inherently pertains to the longer run. It is concerned with phenom-
ena that are properly seen as being essentially developmental.
Mansoor Dailami's chapter discusses many of the salient ingredients
of Korea's overall experience with adjustment in the 1980s but does not
always place them in perspective. Thus it does not develop with sufficient
clarity the point that Korea's rapid recovery from macroeconomic crisis
owed far more to stabilization policies than to concurrent structural ad-
justment policies. More fundamentally, it does not emphasize that the
turnabout owed a great deal to the supply-responsiveness of the private
sector-a responsiveness engendered by a much earlier episode of struc-
tural adjustment that had put the Korean economy on the path of export-
led industrialization. Nor, in this connection, does it adequately reflect
the fact that the crisis had its genesis in the evolution of the structure of
decisionmaking that followed from the earlier episode of adjustment. It
is for these reasons that I wish to focus on the structural adjustment
aspects of the Korean experience.
Korea's first episode of structural adjustment came in the early 1960s
in response to the exhaustion of opportunities for "easy" import sub-
stitution. Two key components of this episode were the institution of a
virtually free trade regime for exporters and the establishment of an
appropriate exchange rate. For sectors in which Korea could then readily
attain international competitiveness, these measures eradicated the an-
tiexport bias of the trade policy and ensured the relative as well as ab-
solute profitability of exporting. In other words, they established the
incentives needed for export-led industrialization.
Other components of this adjustment episode included a policy change
that led to positive real interest rates and reform of government taxation
and expenditure, which greatly enlarged the magnitude of government
savings. The reform of interest rate policy was not accomplished by gen-
eral liberalization of the financial sector, however. Its primary effect was
to divert funds from the curb market into regulated financial institutions
that served as channels for government-directed credit allocation. In the
case of governmental budget reform, however, structural adjustment was
accompanied by moves toward a more liberal structure. ("Liberal" here
refers to conformance with sound economic management or with free



Korea: Successful Adjustment 407
market principles.) Among other things, public enterprises were trans-
formed into profit-maximizing agents.
Import liberalization-in the sense of substantial reductions in import
barriers-was not an element of Korea's structural adjustment in the
early 1960s. This is not to say, however, that there were no important
changes in the rules of the importing game. In fact, there were funda-
mental changes that virtually abolished unproductive rent-seeking in the
use of import licenses. Following those changes, the realization of import
quota rents was generally linked to meaningful production, whether for
export or for domestic sale. There was some liberalization in the con-
ventional sense of the word, but its effects were modest. The most sig-
nificant reform in this respect came somewhat later in the 1960s with
an administrative change from listing commodities that were approved
for import to listing those that were restricted.
The decisionmaking structure that grew out of the reforms of the early
1960s involved significant elements of government intervention that were
discretionary and selective as well as coercive in important respects. As
already mentioned, much of the credit that flowed from regulated fi-
nancial institutions was directed by the government toward sectors it
considered to have priority. The only major exception was export fi-
nance; in this case market forces clearly had the whip hand, but even
here there was rationing in the sense that the government stipulated min-
imum credit allocations in relation to export orders (letters of credit).
Inflows of foreign capital were strictly regulated, as were inflows of pro-
prietary foreign technology. In addition, infant industry protection was
a central element in the government's efforts to promote new industries.
In sum, when all of the government's interventionist policies are taken
into account, Korea's effective incentives were more widely dispersed
among industries than is the case in countries that adhere to an inward-
looking strategy of industrialization.
Just as it would be a mistake to neglect the government's key role in
micro decisionmaking, so too would it be a mistake to infer that market-
mediated, decentralized decisionmaking played no significant role in Ko-
rea's outstanding performance following the first episode of structural
adjustment. The structure of decisionmaking was a complex combination
of public direction and private enterprise in which the government's use
of both carrots and sticks was moderated by frequent, meaningful in-
teraction between public and private decisionmakers.
The critically important nuances of this subtle structure are difficult
to capture adequately in a short description. They are nowhere better
reflected than in Korea's exceptional export performance during the
1960s and early 1970s, which is not only a testament to the efficacy of
"getting the prices right" but also to the potency of export targeting as
it was practiced by the Korean government. It would be naive in the



408 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
extreme to think that government fiat was responsible for Korea's export
success, but it would be equally naive to think that the government's
close monitoring of exports and responsive fine-tuning of promotional
activities was not a major factor.
This structure of decisionmaking served Korea well as long as efficient
economic growth was the overriding objective of government interven-
tion and as long as the interventions were pragmatically flexible in the
face of changing circumstances. Matters changed radically in the mid-
1970s, however, when the government embarked on its campaign to
develop heavy industry. The desire to develop capability in armaments
production (a sensible desire in the geopolitical context of the time) and
the belief that heavy industry is the key to advanced industrial status
supplanted the objective of efficient economic growth. But far more im-
portant, the government became excessively rigid in its interventions;
pragmatic flexibility gave way to ossification as the government lost its
ability to see problems as they emerged and to take appropriate corrective
action. Thus, whatever one might say about the advisability of the con-
certed effort to develop heavy industry, there is no denying that there
was substantial, even egregious, mismanagement of the effort.
Bad management was most clearly apparent in poor project design and
implementation within heavy industry. The long-lasting result was a high
degree of excess capacity. The balance between the volume of investment
in light and heavy industry was also problematic. Owing to insufficient
investment resources, light industry was unable to pursue many profitable
opportunities for expanding capacity. Thus the crowding out of export-
oriented investment was one cause of the marked deterioration in Korea's
export performance over the second half of the 1970s. Adverse real ex-
change rate movements associated with a fixed nominal rate were another
significant cause. In addition, the economy was permitted to become
seriously overheated, as manifested by the substantial tightening of the
labor market because of the large investment in heavy industry and the
simultaneous boom in overseas construction.
The important point here is that the macroeconomic crisis associated
with the 1980 fall in real output was fundamentally due to economic
mismanagement during the 1970s-mismanagement engendered by scle-
rosis in the structure of decisionmaking. In fact, serious macroeconomic
deterioration was apparent in 1978 and 1979, accompanied by clear
evidence of microeconomic mismanagement. The association of ma-
croeconomic deterioration with microeconomic mismanagement was rec-
ognized by many contemporary observers, and a vigorous debate ensued
about the necessity for radical changes in the structure of decisionmaking.
The government embarked on an adjustment program in the first half
of 1979. Shortly thereafter the severely adverse shock in the external
environment only added to the need for adjustment. Consequently, the



Korea: Successful Adjustment 409
adjustment program was greatly strengthened in 1980. The World Bank
did not become formally involved until 1981.
The government's stabilization effort is best seen simply as the rein-
troduction of the sound macroeconomic management it had earlier prac-
ticed with great effectiveness. Likewise, the speed and magnitude of the
turnaround in macroeconomic performance should be understood in
terms of the structure of decisionmaking that had earlier served Korea
so well. As Dailami observes, there was very little reform of that structure
before the restoration of good macroeconomic performance. But there
were major changes in the nature of public intervention. Most important,
the objective of efficient economic growth was restored along with flex-
ible pragmatism in the interventions made in its pursuit. As a conse-
quence, the government abandoned the obsessive push to develop heavy
industry across a broad front. More generally, it greatly reduced its efforts
to promote new industrial activities.
Nonetheless, it is not true that the recovery was associated with a
fundamental transformation in the structure of decisionmaking, as Dai-
lami's chapter implies. If allowance is made for differences in circum-
stances, one sees that the government's role in directing economic activity
was no less than it had been earlier. The directions had changed, but not
the underlying structure. In short, with the old structure of decision-
making reestablished, the economy was able to respond vigorously to
the improvement in external conditions.
Does the success of the government's stabilization efforts within the
preexisting structure of decisionmaking imply that structural adjustment
is unnecessary? "No" is the answer of government policymakers, or so
it appears from the record of efforts undertaken since 1979. In line with
the emerging consensus about policy phasing, the government opted to
pursue stabilization before structural adjustment. Seeing its stabilization
success, it could have abandoned structural adjustment as unnecessary.
The excesses of the second half of the 1970s would then have been con-
sidered an aberration rather than indicative of a new, unhealthy tendency.
But this is not how policymakers chose to view the matter.
The excesses were seen as symptoms of deficiencies that would, with
growing probability, have adverse consequences of increasing severity as
the economy became more complex. Policymakers believed that the par-
amount deficiency was the centralization of decisionmaking with respect
to dynamic allocation, which gave the government tremendous influence
through directed credit, protection of infant industry, and licensing of
foreign capital and technology. Other key deficiencies were either prod-
ucts of past government intervention or concomitants of centralization.
An example of the former is the concentration of market power in the
chaebol, conglomerates which grew extremely large by carrying out many
of the government's favored investment projects. An example of the latter



410 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
deficiency is the many import barriers that remained in the early 1980s.
(With respect to these barriers, it is not irrelevant that Korea has been
under increasing pressure to open its markets.)
The pace of the second episode of structural adjustment has gradually
accelerated since its beginning in the late 1970s. Throughout, the objec-
tive has been to achieve a more liberal decisionmaking structure. Lib-
eralization has been pursued with respect to imports, financial inter-
mediation, and inflows of foreign capital and technology. In addition, a
great deal of ferment has attended efforts to devise a radically new policy
for industrial organization. But, except for a relaxation of restrictions
on inflows of foreign capital and technology (which are now freely ad-
mitted), liberalization is proceeding at a slow and deliberate pace. A great
deal of selective intervention continues as the old structure of decision-
making is gradually replaced by a more liberal one. Dailami makes this
point clearly in regard to the financial sector, but he greatly overstates
the pace of liberalization with respect to imports.'
The slow pace of liberalization can be justified by the desire to give
potentially responsive private enterprise sufficient time to adjust to the
new rules of the game. Specific trade liberalization measures have been
preannounced in order to foster greater productivity in advance of in-
creased import competition. The slow pace can also be justified by the
need to develop adequate processes of decisionmaking within institutions
that are expected to assume new decisionmaking roles of considerable
consequence. Regulated financial institutions are the obvious case in
point.
But, especially with respect to this justification, there is considerable
tension between the utility of delay and the imperative of learning by
doing. Tremendous interest attaches to knowing how Korean policy-
makers have managed this tension and how successful they have been.
It seems clear that the Korean government will continue for some time
its practice of selective intervention coupled with protection of infant
industry to generate import substitution. Here again there are significant
issues concerning the choice of areas for intervention and the detailed
management of intervention. The same is true with respect to the con-
tinuation of selective intervention in the restructuring of sick industries.
These issues are all related to the management of structural adjustment,
which I believe is a more important aspect of the Korean experience than
the management of stabilization, which receives great attention in Dai-
lami's presentation.
In addition, Korea's experience has several implications for the time
horizons of adjustment. In the first episode of structural adjustment, the
export response was strong and virtually instantaneous. Three atypical
elements of the reform were largely responsible for this. First, Korea
established the equivalent of a virtually free trade regime for exports;



Korea: Successful Adjustment 411
with the exception of the other Gang of Four members, I know of no
developing country that has gone this far. Second, the Korean government
was committed to export promotion with a seriousness generally un-
matched among developing countries. The lesson I draw is that efforts
to encourage exports are generally not as extensive as they need to be
to elicit a strong response. Third, in the 1980s Korea's macroeconomic
response to stabilization was strong and rapid. As already noted, this
was due to the underlying strength of the Korean economy, which in
turn was the result of its development over the previous two decades.
The lesson here is that history matters.
What about the World Bank's role in adjustment? I believe it was useful
but minimal. From the beginning of World Bank involvement, the Korean
government had the capacity to formulate and implement the requisite
policies. It also had the political will. Thus Korean policymakers generally
led the World Bank, rather than vice versa. Nonetheless, the World Bank
played a useful role in its consultations on policy; World Bank staff served
as an effective sounding board for brainstorming and analyzing the full
consequences of contemplated actions. Perhaps their most important con-
tribution was to act as a gadfly; they rather consistently emphasized
impediments to achieving Korea's targets, pointing out problems that
they felt could not be overcome. The problems were typically real enough,
but the Korean response was generally to work hard to overcome them,
as Dailami notes was the case in the automotive sector. Significantly, the
only case I am aware of in which the World Bank foresaw serious prob-
lems that were not adequately addressed was the campaign to develop
heavy industry in the mid- and late 1970s.
Note
1. See Vittorio Corbo and Sang-Mok Suh, eds., Structural Adjustment in a
Newly Industrialized Country: Lessons from Korea (forthcoming), especially the
chapter by Soo-Gil Young, "Import Liberalization and Industrial Adjustment."
Young observes that the import liberalization ratio (which Dailami presents as
evidence of actual liberalization) gives at best only part of the story. The ratio
does not incorporate several discretionary import licensing schemes that have
been used to delay import penetration until competing producers were able to
achieve effective competition with imports. Dailami's figures on average tariff
rates appear to overstate the rate of tariff reduction by a considerable margin;
Young gives averages of 24.8 percent in 1979, 23.7 percent in 1983, and 21.3
percent in 1988.



412 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Comments on Asia
Attila Karaosmanoglu
IN REVIEWS of country experiences with adjustment, considerable atten-
tion has focused on the evolution of events rather than on the funda-
mental aspects of structural adjustment. Much more needs to be known
about the cause-result relationships in the processes that are taking place.
Asia provides a rich source of material for analyzing the process of
adjustment. The cases of Indonesia and Korea reveal a great deal of the
complexities as well as the opportunities connected with adjustment.
Other countries in Asia would also make interesting case studies of ad-
justment programs-studies at least as illuminating as those presented
here. The case of the Philippines reveals missed opportunities and wide-
ranging prospects. A study of the dynamics of the adjustment process in
Thailand would be very instructive. China and India have also under-
taken structural reforms in varying degrees and speed, although they have
not obtained adjustment loans from the World Bank.
Although Asia as a whole has achieved better results in adjustment
and growth than have other regions, its experience nevertheless comprises
a range of successes and failures. Generalizations cutting across the varied
experiences are therefore necessarily problematic. Two characteristics,
however, distinguish the successful cases and the more effective aspects
of their adjustment processes.
One characteristic of Asian countries that have achieved higher growth
rates and greater efficiency in their economic structures is the continuous
nature of the structural adjustment process. Continuity in policy, cred-
ibility of reform, and the supply response to policy change are all inter-
linked. Some of the countries in the economic forefront such as Korea
have been engaged in the process for quite some time. Institutional de-
velopment has been sufficiently advanced to facilitate effective changes
in policy. Their track record in carrying out sound policies in turn has
lent credibility to their adjustment efforts. Credibility of reform turns
out to be a fundamental determinant of the degree of response of eco-
nomic agents to policy changes.
The second characteristic of the successful cases is their openness and
their considerable realism with respect to the world around them. Those
countries had, at the starting point, a greater awareness of the importance
of foreign trade and a larger volume of trade. They were better prepared
to incorporate the changing world conditions into their decisionmaking.



Korea: Successful Adjustment 413
I have often wondered whether the policymakers in the successful
Southeast Asian countries are more intelligent and more courageous than
those elsewhere or whether something else drives them to make their
decisions, to make them in time, and to act on them forcefully. One of
my conclusions is that their success is very much related to the openness
of their economies. This openness in turn has forced them to become and
remain competitive in a rapidly changing world. By reflecting the realities
of the world around them in.their own economic structures, these econ-
omies have avoided losing ground internationally. External competitive-
ness and macroeconomic stability are mutually interrelated. With con-
siderable openness, the need for macroeconomic adjustments becomes
more quickly apparent. This is one of the reasons that adjustment has
taken place in that part of the world rather dynamically, continuously,
and effectively.



19 Pakistan: Structural Adjustment
and Economic Growth
William A. McCleary
PAKISTAN EXPERIENCED a very uneven pattern of development during the
1960s and 1970s. Rapid growth during the 1960s, based mainly on pri-
vate enterprise and fairly outward-looking policies, was followed by gen-
erally poor performance and relative stagnation in the 1970s. Starting
in 1969 political disruptions led to the breakaway of Bangladesh and to
a new government in Pakistan in 1971 that reversed many of the policies
of the 1960s. Most large industries, banks, and insurance companies were
nationalized, and massive public investments with long gestation periods
were undertaken in industry (most notably steel). To improve the lot of
urban workers, consumer price controls and subsidies were instituted,
together with low producer prices for crops. Finally, a lack of fiscal and
monetary discipline led to high budget deficits, rapid monetary growth,
and inflation.
The government that came to power in 1977 sought to reverse this
trend by shifting emphasis back to the private sector and restoring fiscal
and financial discipline, but the oil price shocks of 1979 and the sub-
sequent slowdown in the world economy made its task more difficult.
Between 1979 and 1981 Pakistan's terms of trade fell sharply, and rising
interest rates and slackening world trade were additional problems. Pak-
istan received help in weathering the shocks through large inflows from
the International Monetary Fund (IMF) and other external agencies and
from migrant worker remittances, which reached US$2.9 billion at their
peak in 1983 (compared with US$600 million in 1977). These resources
helped smooth Pakistan's adjustment, but they may also have delayed
painful policy changes that would otherwise have been necessary.
Improvements in government policy continued, however, and assisted
the adjustment to the external shocks. Policies in 1977-81 emphasized
reducing the fiscal deficit and government borrowing from the banking
system, reinstituting a program of medium-term planning, improving
public sector infrastructure and institutions, and strengthening the in-
centives for private sector participation and efficiency in the key sectors
of agriculture, industry, and energy.
414



Pakistan: Structural Adjustment and Economic Growth 415
Through expenditure restraint and ad hoc tax and price increases, the
budget deficit was cut from 8.6 percent of GDP in 1977 to 5.3 percent in
1981, thereby reducing credit expansion and cutting inflation rates. The
thrust of medium-term planning was to reestablish real growth in public
expenditures for development and to shift resources from activities that
could be performed by the private sector to those that strengthened public
infrastructure, especially for energy and the social sectors. A shortage of
resources, however, suppressed real growth and delayed the shift of re-
sources to priority sectors. Steps were also taken to enlarge the role of
the private sector and of price mechanisms. Some agricultural processing
and industrial firms were denationalized, tax incentives were introduced
for exporters, and a number of joint ventures were entered into with
foreign private firms for oil and gas exploration.
Despite such policy improvements and the acceleration in growth dur-
ing the late 1970s and early 1980s, Pakistan's adjustment to past policy
mistakes and to external shocks was incomplete. The economy continued
to suffer from structural weaknesses that threatened its potential for rapid
growth under conditions of reasonable internal and external balance.
Prominent among these weaknesses were the following:
D Relatively low investment and savings rates, which were insufficient
to accommodate both the recovery of the private sector and the grow-
ing requirements for public economic and social infrastructure
* A fragile external balance, with imports double the size of exports;
exports heavily concentrated in rice, cotton, and cotton-based prod-
ucts; and foreign exchange receipts heavily dependent on remittances
from migrant workers
* Poor mobilization of resources for the public sector
* Agricultural policies that continued to emphasize low producer prices,
public sector distribution and subsidization of agricultural inputs, and
investment in major new infrastructure projects (mainly irrigation)
* Heavy protection, continued government regulation of prices and mar-
ket entry, and inefficient public enterprises in the industrial sector
* Failure to exploit the country's considerable potential in hydropower
and hydrocarbons, which led to periodic energy shortages.
The Adjustment Program
From 1981 through 1987 Pakistan pursued a structural adjustment pro-
gram addressing these problems. Its efforts were supported by an IMF
Extended Fund Facility program in 1980-83 and World Bank loans (for
structural adjustment in 1982, for energy in 1985, and for export de-
velopment in 1986). The IMF program and the World Bank's structural
adjustment loan (SAL) were drawn up and implemented by the two agen-



416 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
cies in close consultation with each other, and although their emphases
differed, many conditions overlapped. The IMF program emphasized a
continuation and intensification of the government's stabilization efforts,
phased implementation of import liberalization, and improved manage-
ment of the exchange rate. It also included several issues later covered
by the SAL (for example, consumer prices for petroleum products, agri-
cultural subsidies, and tariff reform). Because of the IMF's role the early
World Bank adjustment operations were able to take the importance of
stabilization programs as given and to emphasize issues such as the res-
toration of medium-term development planning, rationalization of prices
in key productive sectors, establishment of a more balanced set of in-
centives for exports and import substitution, deregulation of industrial
prices and investments, and improved institutional performance.
Stabilization Programs
Under the economic stabilization program, fiscal and credit restraints
were expected to contain the level of aggregate demand and inflation
while exchange rate policies were used to shift resources toward the pro-
duction of tradables, especially exports. Under the IMF program the gov-
ernment budget deficit was to be held below 5 percent of GDP by curbing
expenditure, reducing subsidies, and increasing revenue. Government
borrowing from the banking system was to be kept under 2 percent of
GDP, thereby helping to hold annual growth in the domestic money supply
to 14 to 15 percent, or slightly below the expected nominal growth rate
of GDP.
The IMF also monitored the exchange rate. At the urging of the IMF
(and the World Bank), the rupee was delinked from the U.S. dollar in
January 1982 and subsequently managed flexibly with respect to a basket
of currencies of key trading partners. The immediate impact was to re-
verse the real appreciation of the rupee that had begun in 1979, but the
rupee continued to fall in value by more than 40 percent in real terms
by 1987 against the basket of currencies.
Until 1983 Pakistan met the fiscal and credit targets of the IMF. In that
year, however, an unexpected surge in foreign exchange receipts and
government borrowing for budget and commodity operations caused
monetary growth to exceed the program limits by a considerable margin.
Since the government's budget for the following year already included
measures to mobilize resources equivalent to 1.5 percent of GDP, the
government was reluctant to undertake the additional measures the IMF
was demanding to shrink the size of the budget deficit and reduce the
monetary overhang. In November 1983 the IMF credit lapsed, and the
final portion was never drawn.



Pakistan: Structural Adjustment and Economic Growth 417
Policy Reforms
The government's adjustment program emphasized a major shift of re-
sources toward the key productive sectors and increased reliance on mar-
ket-determined prices to provide incentives and encourage efficiency. The
major elements in the program were improved resource mobilization and
restoration of medium-term planning, trade policy reform, decontrol of
input and output pricing, deregulation of private activities, and increased
efficiency of public enterprises.
BUDGETARY POLICY AND. MEDIUM-TERM PLANNING. The IMF program
and various World Bank operations called for improved resource mo-
bilization through improvements in the coverage and elasticity of the tax
system and reductions in subsidies. Such policies would have permitted
an expansion of government expenditures and a shift toward high-prior-
ity sectors, while overall budget deficits remained at acceptable levels.
However, the government made relatively little progress in this area either
under the IMF program or subsequently. Revenue measures consisted
largely of ad hoc changes introduced with each budget (for example,
import surcharges and changes in public utility rates or oil and gas prices).
No systematic effort was made to introduce new taxes (such as a value
added tax or broad-based sales tax) or to expand the coverage of existing
taxes. Although some subsidies were eliminated, others were expanded
to make up the difference. As a consequence of these failures, budget
deficits were a continuing problem through the 1980s.
Increased resource mobilization was also intended to permit directing
public expenditures more toward the development of agriculture, water,
energy, and social services. Reductions in fertilizer subsidies and in public
investments in industry were expected to free additional resources. De-
velopment expenditures grew at a real average annual rate of 3 to 4
percent during the 1980s, and substantial progress was made in re-
orienting expenditures. About 85 percent of expenditure targets under
the sixth five-year plan (fiscal 1984-88) were implemented, and public
resources were shifted markedly toward power, education, health, ag-
riculture, and rural development. The shift in resources would have been
greater but for budgetary constraints and the failure to reduce subsidies
(especially for fertilizer) as called for in the plan.'
Public investment programs in agriculture and energy received strong
support in the adjustment program. In agriculture there was a shift away
from activities that could be handled by the private sector (for example,
storage and mechanization) and from major new investments to expen-
ditures that would raise the productivity of existing investments. Thus
allocations for operations and maintenance in irrigation rose significantly
in real terms, extension and research received a growing share of total



418 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
agricultural expenditures, and irrigation projects emphasized water man-
agement, rehabilitation, and drainage in saline areas. In energy the goals
were to increase private production and to rationalize investment in and
improve financing for public sector programs in oil, gas, and electric
power. These efforts substantially increased exploration activity and oil
and gas output from private and public sources. The private sector was
particularly responsive to the adoption of a new formula for gas producer
prices.
TRADE POLICY. Pakistan's trade regime was substantially biased toward
import substitution, with pervasive quantitative restrictions and tariffs
ranging from zero to 350 percent dominating a duty drawback system
and outright subsidies for exports (the "compensatory" rebate). The IMF
and World Bank called for substantial trade liberalization, involving a
reduction in the coverage of nontariff barriers followed by tariff reform.
Coverage by quantitative restrictions dropped from the equivalent of
about 66 percent of manufacturing value added to 27 percent, and a
negative list of prohibited imports replaced a more restrictive positive
list of allowed imports. The tariff reform was not introduced during this
period, however, because the study on which it was to be based was not
completed until December 1983 and did not contain specific proposals
for the government to consider.
Trade policy reform continued after the IMF program and the SAL,
through the government's initiative and the World Bank's support under
the export development loan. Flexible management of the exchange rate
continued as well. Some 136 items were removed from the negative or
restricted lists, although about 400 four-digit commodity headings
(mostly consumer goods but some important capital goods as well) re-
mained subject to some form of quantitative restriction. The number of
tariff rates was reduced from seventeen to ten, while the (unweighted)
average tariff rate was reduced from 77 percent to 66 percent. Improve-
ments in the duty drawback and open-bonded manufacturing systems
were introduced to move exports toward a free trade basis. These
changes, combined with the elimination of export subsidies, have prob-
ably improved relative incentives for exports somewhat, but they have
left the antiexport bias of the trade regime largely intact.
DEREGULATION. Through regulation, the government had sought to
foster efficient development while avoiding the adverse effects of relying
on the price mechanism in the absence of foreign competition. The main
forms of regulation were investment sanctions, price controls, and cost-
plus pricing. During the 1980s several changes were introduced. The size
limit for projects requiring government sanction of investment plans was
increased substantially; the list of industries requiring government sanc-



Pakistan: Structural Adjustment and Economic Growth 419
tions was shortened (changes in 1987 reduced the list from the equivalent
of 40 percent of manufacturing value added to 7 percent); and cost-plus
pricing or subsidized pricing was eliminated in key industries such as
cement, nitrogenous fertilizer, and vegetable ghee.
By and large, there was considerable progress. The role of government
in investment and pricing decisions was significantly reduced, and the
economy became more market-oriented. Additional improvements are
still needed, however, in particular the removal of investment sanctions
based on project size and import content, further shrinkage of the use
of price controls and cost-plus arrangements, and increased use of price
taxes (rather than location licenses) to ration infrastructure and handle
environmental concerns.
PRICING POLICIES. Pricing policies in energy and agriculture shifted their
emphasis toward encouraging efficient consumption and production de-
cisions and providing appropriate incentives for domestic producers. Ex-
cept in the case of certain agricultural inputs (fertilizers and irrigation
water), the policy changes were significant. On the consumption side,
price increases averaging 24 percent a year during 1981-86 brought
natural gas prices to two-thirds of the fuel oil parity level by mid-1985
(three years ahead of schedule). As a result of subsequent declines in
world oil prices, gas prices came to exceed fuel oil parity. Domestic oil
prices continued to exceed world prices because increases in world prices
were fully passed through to domestic consumers but were only partly
reversed as oil prices fell. Reversing a long-standing policy of keeping
power cheap for social reasons, the government also agreed to adjust
power tariffs so that 40 percent of an agreed-on power investment pro-
gram could be self-financed; subsequent increases in power tariffs have
met this objective.
With respect to producer prices for energy, the government responded
flexibly in providing appropriate incentives for private sector partici-
pation. Through much of the 1980s producer prices for oil discovered
under joint ventures were set equal to the official prices for market crude
(Arabian light), with an adjustment for quality and a prenegotiated dis-
count. As international oil prices declined, continuation of this practice
would have threatened exploration and development. Instead, the gov-
ernment announced a new formula that included sliding-scale discounts,
which vary inversely with comparator prices. This revised formula was
successful in attracting new companies and maintaining incentives for
exploration and development.
Government policy was somewhat slower to change on producer prices
for gas. Under the SAL the government initiated a number of joint ventures
for gas exploration and development. According to the government's
plan, prices, based on a cost-plus formula, would not be agreed on until



420 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
after discovery. The World Bank had serious doubts about this system
because of the considerable uncertainty involved. When after several
years no agreements had been reached on this basis, a new formula was
introduced under the energy loan. It tied the price to fuel oil parity, less
a discount to be negotiated in advance on the basis of such factors as
location and geological risk. This formula has been successful in attract-
ing private response.
During the 1980s Pakistan also sought to revise two key agricultural
sector policies: support prices and input subsidies. Support prices had
been significantly below international prices and constituted a severe tax
on agriculture. Under the SAL the government conducted annual price
reviews and established appropriate incentives for farmers through price
support recommendations based on trends in international prices and in
local costs. On the basis of such annual reviews, the prices established
for most major crops remained close to world price trends, and the use
of support prices to tax agriculture was virtually eliminated. In the case
of input subsidies progress was more mixed. Pesticide subsidies were
eliminated, but targets for eliminating fertilizer subsidies and increasing
the cost recovery of irrigation operations and maintenance were not met.
PERFORMANCE OF PUBLIC ENTERPRISES. Public enterprises are important
in the manufacturing sector less for their overall size (15 percent of the
sector's value added) than for their dominant role in key subsectors such
as steel, cement, fertilizer, and basic chemicals. Government efforts to
improve the performance of public enterprises during the 1980s centered
on making such firms behave more like those in the private sector and
reducing their reliance on budgetary resources. The government imple-
mented a "public enterprise signaling system" for public manufacturing
firms, involving information management, and established targets and
rewards for performance. Evidence suggests that the system did improve
performance and helped to focus government officials' and managers'
attention on the need to improve efficiency. It also revealed a number of
impediments to improving efficiency, such as insufficient management
autonomy, the need for skilled and experienced managers, and excessive
government regulation.
In some cases the government has undertaken privatization measures,
including disinvestment of shares and denationalization. Progress in this
area has been slow, however. A disinvestment committee was established,
a few firms were offered for outright sale, and plans were being made
to sell shares in a number of important enterprises, such as Pakistan
International Airways and Sui Gas Transmission. To date, however, little
disinvestment has actually taken place.



Pakistan: Structural Adjustment and Economic Growth 421
Economic Performance
Pakistan is one of the few developing countries whose growth was higher
during the 1980s than during 1973-80, to a great extent because of the
large scope for reforming the policies that had restrained growth during
the 1970s. Improvements in policies affecting stabilization and efficiency
had already begun to accelerate growth before the 1979 shock, and policy
reforms adopted since then helped to reconfirm the process. Despite sub-
stantial progress, however, Pakistan's adjustment is far from complete.
Many areas remain in need of considerable improvement, especially pub-
lic resource mobilization and import liberalization.
Growth, Employment, and Wages
Pakistan grew at an annual rate of 6.5 percent (or 3.3 percent on a per
capita basis) during 1980-87 (see table 19-1). This rapid growth was
not concentrated in a few sectors but characterized virtually all key sec-
tors. Agricultural expansion of 3.1 percent a year, though somewhat less
than government authorities were targeting, was nonetheless respectable.
It allowed self-sufficiency in some key commodities (for example, wheat
and sugar) and provided surpluses for export in others (for example, rice,
cotton, and cotton-based textiles). Although production is still heavily
concentrated in these major crops (just over half of value added), some
diversification has occurred into minor crops (such as oilseeds) and into
livestock and fisheries. Some of the improvement in agricultural perfor-
mance resulted from favorable weather and private initiative, but gov-
ernment efforts also contributed-for example, better output prices, im-
provement in the amount and timing of irrigation water, distribution of
improved seeds, increased coverage of agricultural credit, and improved
balance between the private and public sectors in the distribution of
fertilizer and pesticides.
Industrial growth of 9.1 percent a year during 1980-87 was encour-
aged by improvements in the policy environment for private initiative.
The freeing up of prices and investment as well as a reduced role for the
public sector were helpful in this regard. The composition of investment
shifted heavily toward the private sector: private investment expanded
at an annual rate of 23 percent in real terms during fiscal 1984-87,
compared with 7 percent for total industrial investment. Overall, 72 per-
cent of industrial investment during that period was contributed by the
private sector. Public sector investment grew slowly and was restricted
to ongoing projects and some rehabilitation projects. Major contributors
to growth in the large-scale industrial sector were cotton yarn, sugar,
cement, fertilizer, chemicals, iron and steel, and automobiles. Also im-
portant was the rapid expansion of small-scale manufacturing, which



Table 19-1. Economic Indicators for Pakistan
Growth rates (annual
Share of GDP (current prices)          percent in real terms)
1980     1985      1986      1987      1988      1973-80    1980-88
Gross domestic product                   100.0    100.0    100.0    100.0    100.0            5.4        6.5
Agriculture                             26.5      22.8      22.0      23.6      23.5        2.8        4.3
Industry                                22.4      24.6      25.3      21.6      21.7        7.0        7.2
Services                                40.9      42.8      42.6      44.8      44.0        6.4        7.4
Net indirect taxes                       10.2      9.9      10.1      10.0      10.7        -          -
Total expenditure                        111.6    112.1    109.5    105.3    107.5            6.3        5.5
Total consumption                       93.1      95.3      92.8      86.1      89.4        6.5        5.3
Private consumption                   83.1      83.1      80.5      72.6      75.5        6.8        4.4
General government                     10.0      12.2     12.2      13.5      13.9        4.0       11.0
4s                 Gross domestic investment                18.5      16.8     16.7      19.1      18.1        5.8        6.4
Fixed investment                       17.6      15.0     15.1      17.5      16.6        6.7        6.1
Changes in stock                       0.9        1.8      1.6       1.6       1.5        -           -
Resource balance                        - 11.6    - 12.1     - 9.5     - 5.3     -7.5         -           -
Export of goods and nonfactory services  12.5      10.4     12.1      13.8      14.4        4.2        8.9
Import of goods and nonfactory services  24.1     22.5      21.6      19.1      21.9        9.1        3.0
Gross domestic savings                     6.9       4.7       7.2      13.9      10.6     -11.6        16.0
Gross national savings                    13.6      11.7      13.8      19.6      14.5       4.0         9.5
Overall budgetary deficit                - 6.3     -7.4      - 8.1     - 8.2     - 8.6        -           -
Current account deficit (b.o.p.)          -4.8      -5.4     -3.9      - 2.2     -4.4         -           -
Debt ratios
Debt outstanding and disbursed/GDP        42.0      42.9      46.7     50.1      44.5         -           -
Debt outstanding and disbursed/exports   208.9    228.1    228.9    246.0    231.0            -           -
Debt service/exports                      18.0      24.2      25.3      26.5      24.7        -           -
-Not applicable.
Source: World Bank data.



Pakistan: Structural Adjustment and Economic Growth 423
also contributed to export growth (for example, garments, towels, ho-
siery, leather products, and surgical instruments).
In the energy sector, as a result of public and private activities, a record
number of wells were drilled for exploration and development. Oil pro-
duction more than tripled, and Pakistan now supplies about 25 percent
of its petroleum requirements (compared with 10 percent in the early
1980s). Expansion in natural gas was more modest, but as investments
began to come on-stream growth was expected to average 4.8 percent
annually in the sixth plan period. Expansion of electric power, while
impressive, is considerably short of projections; even with the acceler-
ation of investment in recent years, capacity will not meet demand before
the early 1990s.
With the help of the Middle East oil boom, the domestic labor market
was able to absorb, at rising real wages in key sectors, the increases in
the labor force. By the early 1980s outmigration to the Middle East was
accommodating some 2 million Pakistani workers (the equivalent of
about 7 percent of the labor force). Since the labor force was growing
by 2.3 percent a year during 1979-85 while employment was growing
at 2.2 percent, open unemployment grew slightly from 3.5 percent to 3.7
percent of the labor force. As is typical in developing countries, urban
unemployment (5 percent) was higher than rural (3.1 percent), and youth
unemployment (7 percent) higher than the national average. A fairly
recent phenomenon is the rise of unemployment among the educated
urban population, with degree holders experiencing unemployment rates
roughly double those of illiterates. Consistent with the strong expansion
of the economy and the relatively slow growth of the labor force, real
wages have continued to increase. The wages of agricultural laborers
grew at 11 percent a year and those of employees in small-scale and
large-scale manufacturing grew at 3 to 4 percent a year between 1979
and 1987.
Macroeconomic Balance
DOMESTIC ABSORPTION. Pakistan responded to the crisis of 1979 by
reducing the real growth of domestic absorption from an average of 6.3
percent a year in 1973-80 to 5.9 percent in 1980-87, which was below
the average real growth of the economy (6.5 percent) in the latter period.
The external counterpart to this decrease in absorption showed up in the
resource balance, as real growth of exports of goods and nonfactor ser-
vices accelerated to 9.2 percent a year, while import growth fell sharply
to about 2.8 percent a year. Contributing to this performance were a
significant fall in the real effective exchange rate, improved policies in
the agriculture and energy sectors, and the slow growth of external fi-
nancing.



424 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
The reduced growth in domestic absorption resulted from a fairly sharp
shift in the structure of demand. Compared with the 1970s, private con-
sumption grew more slowly while private investment, government con-
sumption, and government investment grew more rapidly. The deceler-
ation of private consumption, though it cannot be explained with
certainty, could be the result of rising real rates of return on financial
assets such as deposits or government savings schemes, as well as a policy
environment that was considerably more hospitable than before to pri-
vate investment. The acceleration of government expenditures, since it
was not accompanied by increased resource mobilization, led to increas-
ingly large budget deficits.
Pakistan's investment rate of 16 to 17 percent of GDP was very low,
especially given the economy's relatively high growth rate. It seems im-
plausible that efficiency was as high as an incremental capital-output
ratio (ICOR) of 2.6 for 1980-88 would imply. A more plausible expla-
nation is that improved policies made for improved utilization of existing
capital and that existing capital was being depleted as the maintenance
and the replacement of aging assets were being neglected (as evidenced,
for example, in the deteriorating highways and irrigation systems and
the neglect of social sector investment). It appears highly likely, therefore,
that the ICOR will be much higher in the future.
There was very little improvement in the national savings rate. Im-
provements in private savings were virtually offset by deterioration in
public sector savings. The share of public investment financed from public
savings fell to extremely low levels (less than 20 percent in 1985-88 and
less than 5 percent in two of those three years). Although the savings
performance of public enterprises was far from satisfactory, the deteri-
oration in overall public sector performance emanated from the budget.
BUDGETARY PERFORMANCE. After some improvement during the IMF
program, Pakistan's budget situation deteriorated markedly. Overall def-
icits rose from 5.3 percent of GDP in fiscal 1981 to over 8 percent of GDP
in fiscal 1986 and 1987. With the rising deficits came a growing depen-
dence on borrowing, especially from domestic sources. The growth of
public domestic indebtedness has been accelerating, and debt-service pay-
ments are absorbing a growing share of government expenditures. The
budgetary performance of recent years is clearly not sustainable.
The major factors contributing to the budgetary deterioration were the
rapid rise of current expenditures and the failure to generate sufficient
additional revenue. The growth of development expenditures was rela-
tively slow, and their share in GDP fell. After fiscal 1981 current expen-
ditures rose at an annual rate of 18.8 percent (11.1 percent in real terms),
with their share of GDP rising from 15 percent to 19 percent. Defense
and interest payments accounted for a large part of that increase. In 1988



Pakistan: Structural Adjustment and Economic Growth 425
Pakistan was spending 25 percent of government expenditures on de-
fense, a significantly larger proportion than the averages in 1985 for low-
income (18.6 percent), lower-middle-income (14.2 percent), and indus-
trial countries (16.8 percent). Interest payments as a share of GDP more
than doubled in the 1980s. The only bright spot on the current expen-
diture side was the positive real growth in outlays for the chronically
underfunded economic and social sectors.
Over the 1981-88 period revenue remained at 17 to 18 percent of
GDP. That this stability was achieved through ad hoc measures introduced
yearly indicates a basic inelasticity of the revenue system. Since the early
1980s the IMF and the World Bank have been recommending that the
government increase the coverage and elasticity of its revenue-generation
system by introducing a broad-based tax on consumption (sales or value
added), shifting excise taxes from specific to ad valorem rates and ex-
tending the range of goods covered, expanding coverage of the income
tax system, and improving cost recovery for public goods and services.
Very little was done on these measures, however, before the fiscal 1989
budget.
Both foreign and domestic borrowing for budgetary purposes in-
creased, with domestic borrowing, especially from nonbank sources,
growing more rapidly. Domestic interest payments grew at 27 percent a
year until they absorbed 14 percent of current expenditure, compared
with 8 percent in the early 1980s. The government was able to tap private
savings through a number of savings schemes with attractive interest rates
and other innovative features. Although tapping private savings enabled
the government to postpone tax reform and other resource mobilization
measures, the postponement came at a high price in the form of rising
interest costs and constraints on more essential government expenditures.
The budgetary performance of the late 1980s was inconsistent with
the government's targets for inflation, increased foreign exchange hold-
ings, growth in the private sector, and expansion of essential public eco-
nomic and social services. Continuation along recent lines would mean
that one or more of the government's targets would have to be aban-
doned. The sustainable long-term deficit appears to be just under 5 per-
cent of GDP, a reduction of 3.0 to 3.5 percentage points from the deficits
of recent years.2
MONEY, CREDIT, AND PRICES. In Pakistan the objectives of monetary
policy are articulated in an annual credit plan, which places ceilings on
credit expansion by commercial banks and allocates mandatory mini-
mum credit levels for priority sectors (for example, agriculture, small-
scale industry, exports) and indicative levels for other sectors. The gov-
ernment's credit needs and expected changes in net foreign-held assets
are also taken into account. After fiscal 1980 the money supply grew at



426 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
an annual rate of 15 percent and prices at 6.2 percent, roughly in line
with the government's targets for real growth and inflation. These in-
creases were a marked improvement over the 1973-78 period when the
money supply grew at 22 percent a year and prices at 12 percent.
Generally, the annual plans have called for some modest expansion in
net foreign assets, limited public sector reliance on the monetary system,
and a growing share for the private sector in the additional credit. Al-
though the private sector has indeed been obtaining a growing share of
the additional credit (nearly two-thirds of the increase in recent years,
compared with 40 to 50 percent in the late 1970s), credit expansion to
both the private and public sectors has often exceeded targets. Thus,
although growth of the money supply has generally been in line with
government targets, the accommodating factor has been net foreign as-
sets, which have declined in most years since 1980. This is clearly not a
sustainable policy. The need to increase official foreign exchange holdings
(building up reserve holdings while paying off short-term debt) will raise
issues about priorities between the private and public sectors and the
possible crowding out of the private sector by public sector credit de-
mands.
The External Sector
BALANCE OF PAYMENTS. During the 1980s Pakistan's balance of pay-
ments was affected by several factors over which the government had
little control: the instability in international commodity markets (espe-
cially rice, cotton, and petroleum products); the sluggish recovery of the
world economy; the shrinking markets for migrant workers, especially
in the Middle East; and the war in Afghanistan, which affected import
levels (to assist refugees) and aid flows. The balance of payments was
also affected by factors over which the government did have influence,
such as the real devaluation of the rupee after it was delinked from the
U.S. dollar, gradual reductions in nontariff barriers and tariff rates, and
improvements in sectoral policies (concerning agricultural and energy
pricing and investment programs) and policies affecting capital flows (for
example, aid disbursements and the creation of new instruments to raise
short-term funds from abroad).
The major forces driving Pakistan's balance of payments performance
after 1980 were the rapid growth (10 percent a year) in the volume of
merchandise exports coupled with a slow growth (4 percent) in mer-
chandise imports; a rise in Pakistan's share in world trade for important
export products (such as rice, cotton, cotton textiles and clothing, surgical
instruments, and leather goods); the rapid rise in migrant worker re-
mittances following the oil price boom and the subsequent decline; rela-
tively flat (net) long-term capital inflows (mostly official); and increased



Pakistan: Structural Adjustment and Economic Growth 427
reliance on short-term capital inflows to slow the decline in (gross) official
reserve holdings. As a result of these and other factors, Pakistan's current
account deficit declined from US$1.2 billion in fiscal 1980 (4.8 percent
of GDP) to US$700 million (or 2.2 percent of GDP) in fiscal 1987.
Despite this improvement, however, Pakistan's external situation at
the end of the decade remained fragile. Since delinkage of the rupee from
the dollar, Pakistan's real effective exchange rate had depreciated some
40 percent. Although this contributed to improved export performance,
there was still no export diversification. Merchandise exports consisted
mainly of rice, cotton, and cotton textiles whose share had remained
roughly unchanged at 60 percent since 1980. Rice and cotton were subject
to substantial year-to-year price fluctuations, and cotton textiles were at
the mercy of growing protectionism in the United States and the European
Economic Community (EEC). Remittances from migrant workers re-
mained an important source of foreign exchange earnings. Although past
declines in remittances were compensated for somewhat by lower oil
prices, Pakistan depends on the Middle East as a market for its products
and its labor and, on balance, probably benefited from the higher oil
prices.
By the end of fiscal 1987 Pakistan's short-term position had deterio-
rated sharply. Gross official reserves were US$500 million, down from
their peak of US$2 billion in 1983 and not much higher than they were
in 1980 when trade was substantially smaller. In addition, Pakistan's
short-term liabilities stood at US$2.1 billion as a result of increasing
reliance on short-term instruments for raising external funds. Although
these vehicles have attracted substantial resources, they are highly volatile
since they depend on holders' perceptions of exchange rate movements
and other economic developments in Pakistan and abroad. This could
limit the options for raising additional resources, should this become
necessary.
EXTERNAL DEBT AND DEBT SERVICE. Between 1977 and 1982 Pakistan's
indebtedness and debt-servicing capacity showed marked improvements
because of increased exports, a more than fourfold increase in migrant
worker remittances, and relatively conservative financial management.
Since then, there has been some deterioration because of falling remit-
tances, increased debt service payments to the IMF and commercial lend-
ers, and a looser fiscal policy stance. Of particular concern, however, is
the buildup of all forms of debt since 1982 and the decline in official
reserve holdings. With it has come a substantial increase in the debt
service/exports ratio from 18 percent in 1980 to 27 percent in 1987.3
Reserves, which were no higher at the end of 1987 than in 1980, could
cover only one-quarter of debt service requirements in 1987 versus two-
thirds in 1980. This deterioration suggests that continued progress in



428 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Pakistan's adjustment program will be essential to restoring the viability
of its external position and improving its creditworthiness.
Assessment of Conditionality
Adjustment loan conditionality in Pakistan went through two distinct
phases. During the first phase the IMF and World Bank programs over-
lapped; the macroeconomic conditionality of the SAL mirrored that of the
IMF. With the IMF and the World Bank in substantial agreement on ap-
propriate stabilization policies, the World Bank concentrated on medium-
term planning, the level and structure of the public sector investment
program, and sectoral policies-such as appropriate pricing for inputs
and outputs, deregulation of prices and investment decisions, and insti-
tutional reforms. In the second phase the World Bank's sectoral emphasis
remained broadly the same, but it had to take on the monitoring of
macroeconomic policy as well because of the absence of the IMF and a
failure to reach agreement on a follow-up SAL.
Table 19-2 lists the major conditions that accompanied adjustment
loans to Pakistan by degree of compliance. Implementation of some con-
ditions was relatively easy because no readily identifiable major group
was harmed by the policy change (for example, medium-term planning,
reorientation of public development programs). In some cases positive
external shocks eased the policy change; for example, the great drop in
international prices made it easy to remove price controls and subsidies
on oilseeds and nitrogenous fertilizer since the price drop rendered the
policy redundant. Often conditions were implemented after intensive pol-
icy dialogue with the World Bank (as in the case of energy pricing) or
after a long period of technical assistance and sector work (agricultural
pricing, reductions in import bans and restrictions). In cases in which
there was little or no progress, it is evident that the political will was
lacking (for example, on tariffs, budgetary improvements, irrigation
charges) or that the World Bank did not give the government concrete
advice in time (the study on effective protection, for example, was late,
flawed, and lacking in concrete policy prescriptions).
Sustainability
An interesting and unusual aspect of Pakistan's adjustment program is
the lack of policy reversals once reform had begun.4 Several factors help
to explain why reforms were so sustainable. The government's approach
to policymaking was very cautious: there were no abrupt changes, and
the policy measures adopted were flexible enough to leave the govern-
ment room for maneuver should the results not be as expected. Moreover,
in most instances the government was convinced of the rightness of its



Pakistan: Structural Adjustment and Economic Growth 429
Table 19-2. Implementation of Adjustment Loan Conditionality in
Pakistan
Significant progress               Satisfactory progress              Little or no progress
Exchange rate                Power tariffs                      Budget deficit
management                 Elimination of cost-plus pricing  Tax reform
Consumer prices of oil and     (for example, of oilseeds,       Budget subsidies
gas                          nitrogenous fertilizers, cement)   Tariff reform
Producer prices of oil and   Elimination of pesticide subsidy   Irrigation cost recovery
gas                        Allocation for operation and      Non-nitrogenous fertilizer
Agricultural output prices     maintenance of irrigation          subsidies
Medium-term planning           systems                          Disinvestment of public
Reduction in requirements for       enterprises
investment sanctioning
Incentive system for
performance in public sector
manufacturing
Reduction in import bans and
restrictions
Restructuring of public
development budget (overall,
and especially for energy,
agriculture, irrigation)
Revision of duty drawback
system for exportersa
Open-bonded warehousing for
exporters'
Improved credit for exporters'
Note: "Significant" indicates that progress exceeds World Bank-IMF conditions and was not reversed in a
succeeding period; "satisfactory" indicates that progress meets World Bank-iMF conditions and was not sub-
sequently reversed.
a. Substantial progress under way, but condition not yet met.
approach in advance (for example, deregulation, agricultural output pric-
ing, gas producer prices) or as a result of the benefits that quickly became
apparent (for example, the favorable effects of exchange rate delinkage
on exports and migrant worker remittances). In a number of cases (for
example, energy, deregulation), the government's confidence was the
product of extensive dialogue with the IMF or the World Bank. The sta-
bility of Pakistan's growth and the absence of further adverse shocks also
certainly aided the reform process.
Tranching
Both the SAL and the energy loan were disbursed in two tranches; the
export development loan had only a single tranche, given its small size
and the World Bank's expectation that progress in implementation could
lead quickly to an industrial sector loan the following year. Release of
the second tranche of the SAL was linked to three conditions as well as



430 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
to satisfactory progress in implementing the remainder of the program:
preparation of a three-year public sector investment program, a report
on government decisions on agricultural support prices and key input
prices, and a report assessing progress in implementing the accelerated
program for developing the gas supply. Implementation was judged to
be satisfactory, and the second tranche was released on schedule. The
tranche release for the energy loan was not as straightforward. The
tranche conditions were review of the electricity tariff for fiscal 1987,
review of implementation of the core investment program for the energy
sector for 1985 and 1986, and agreement on the revised energy invest-
ment program for 1987 and 1988. Tranche release was delayed about
nine months until it could be determined that power tariffs had been
raised sufficiently to finance 40 percent of the power investment program.
World Bank adjustment lending for Pakistan has made sparing use of
"up-front" conditionality, which must be met before the loan. The SAL
had no up-front conditions because it was introduced at the same time
-that the IMF program was being successfully implemented. The energy
sector loan had two advance conditions related to increases in power
tariffs and a new producer pricing formula for natural gas. The export
development loan had only one advance condition-appointment of a
person of appropriate experience and rank to head the Industrial Incen-
tives Reform Cell-because many of the policy reforms that had been
discussed (for example, exchange rate adjustments) were enacted by the
government on its own initiative before the loan was presented to the
Executive Directors of the World Bank for approval.
Sequencing of Loans
In view of Pakistan's continuing problems with resource mobilization, it
would have been logical to follow the first SAL with a second. However,
the failure to complete the IMF program, the high visibility of SALS, and
the growing political liberalization of Pakistan made negotiation of a
follow-up SAL virtually impossible. Instead, the World Bank and the gov-
ernment of Pakistan agreed to a lending program consisting of about 75
percent project loans and 25 percent policy-based sectoral adjustment
loans (SECALS) in industry, agriculture, and energy. The policy-based loans
depended on agreement on an acceptable program of reform in the rel-
evant sectors and on the government's maintaining appropriate overall
macroeconomic management. Progress in macroeconomic management
was to be monitored in annual reviews by the World Bank and the gov-
ernment. This approach was more acceptable to the government than a
SAL because the macroeconomic conditionality would be less intrusive
and the sectoral conditionality was broken up into discrete chunks.



Pakistan: Structural Adjustment and Economic Growth 431
As it turned out, however, the procedures for ensuring an appropriate
macroeconomic policy environment were not effective. Despite annual
consultations, the budgetary situation continued to deteriorate, and the
government made little effort to reverse the process. The problem was
not that the macroeconomic environment was so unfavorable that sec-
toral policy changes would have been ineffective; indeed, relatively low
inflation rates plus continued effective exchange rate management meant
that policy changes involving relative prices and improvements in effi-
ciency would most likely have been effective. Rather, the issues were the
deteriorating creditworthiness and the ineffective use of the World Bank's
leverage to obtain policy improvements.
With the rapid buildup of debt service requirements on both foreign
and domestic debt, Pakistan was quickly approaching the limits of its
ability to service such debt. This, in turn, called into question the jus-
tification for a larger World Bank lending program in Pakistan. In ad-
dition, progress in some sectors was very slow. As a result, the World
Bank first slowed its sectoral lending considerably and then suspended
it. This outcome highlights the problem of how to ensure appropriate
stabilization policies in the absence of an IMF agreement. World Bank
consultations are infrequent, and the implied threat of cutting back a
lending program seems to lack the urgency and immediacy that accom-
panies the IMF's quarterly tranche reviews.
Several broad trends in conditionality are discernible in Pakistan's ex-
perience in moving from SALS to SECALS. First, the number of conditions
attached to any one loan dropped. This is not surprising, although per-
haps misleading, since the coverage of sectoral loans is narrower and an
array of implicit macroeconomic policy conditions was being monitored
concurrently. Second, there was a tendency to focus each loan on about
four or five items on which fairly specific-if not necessarily quantifi-
able-actions were expected. In part, this increasing precision was a
product of the World Bank's growing knowledge in a number of areas,
progress in the dialogue with the government, and a growing realization
that it is better to focus on a few issues that are important and feasible.
Third, the World Bank became more sensitive to the need to adjust
the pace of reform to what is politically and institutionally feasible, that
is, to make progress in areas where progress is feasible and to leave
recalcitrant problems to later follow-up loans. For example, the export
development loan capitalized on the government's desire to move forward
with export incentives, liberalization of imported inputs, and deregula-
tion, and it left tariff reforms for a later operation. Fourth, significant
progress was sometimes made with little specific conditionality. For ex-
ample, substantial liberalization of government controls on investment
and prices was based on a general understanding between the government
and the World Bank about deregulation priorities. The precise details,



432 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
order, and timing were left to the government. Had the World Bank
inserted itself into this politically sensitive area with demands for specific
actions by specific dates, the process could have been self-defeating.
Conclusions and Lessons
In many respects the Pakistan economy of the late 1980s differed from
the economy of 1980. A period of rapid growth and inflows of migrant
worker remittances had raised standards of living broadly across the
population. Private initiative played a leading role in this expansion, and
no longer was public investment in large-scale industry viewed as the
engine of growth. The public sector had taken on a supportive role and
was beginning to restore the country's neglected economic and social
infrastructure while initiating steps to improve the performance of public
enterprises. During this time, the economy became substantially more
outward-looking, flexible, and market-oriented. More active exchange
rate management, combined with better export incentives and reductions
in nontariff barriers and, to a lesser extent, tariffs, brought about some
improvement in the environment for exports. The government attitude
toward regulation became more relaxed: controls on market entry, en-
forced through investment controls, were significantly reduced, and price
controls and cost-plus pricing arrangements were eliminated in some key
areas. Agricultural and energy prices were adjusted in line with inter-
national prices, improving efficiency and the incentives for increased pro-
duction.
Although reforms undertaken during the 1980s improved performance
and the capacity to respond to economic change, there are some wor-
risome signs. The continued decline in remittances and the possible lev-
eling off of aid following the end of the Afghanistan War are likely to
mean that some painful adjustments can be postponed no longer. Re-
forms in some areas are still incomplete, while in others progress has
been slow or nonexistent. Controls over investment remain intrusive,
while in the crucial areas of the budget and the trade regime, major
changes are yet to come. Fundamental reform of public sector resource
mobilization remains to be undertaken, and budget deficits have been
running at unsustainable levels. Many consumer goods remain subject
to quantitative restrictions, and the average level of tariffs is still very
high, so antiexport bias remains significant.
The dialogue between the government of Pakistan and the World Bank
has contributed to the economic improvements in Pakistan. After modest
beginnings under the SAL, when the areas amenable to meaningful dia-
logue were few, the relationship has broadened through improved eco-
nomic and sector work, experience with successful reforms, and a grow-
ing mutual trust. Negotiations now include discussions and actions in



Pakistan: Structural Adjustment and Economic Growth 433
areas that could not have been broached in the early years. For example,
progress would have been difficult or unthinkable in the early 1980s on
such issues as a least-cost investment program for the power sector, an
incentive program for exports, reductions in import restrictions for raw
materials and intermediate goods, and reforms of producer pricing ar-
rangements in gas and power-areas in which actions have recently been
taken or are under way.
Several broad generalizations can be drawn about the adjustment pro-
cess and the respective roles of the government of Pakistan and the World
Bank:
. Progress has been easier and more solid in areas in which the dialogue
covered issues that were specific and concrete.
* Reforms in Pakistan have been sustained because of the government's
incremental and flexible approach and because continued strong eco-
nomic performance has obviated the need for reversals.
* When there is broad understanding between the government and the
World Bank about the direction and elements of needed change, con-
siderable progress is possible without specific conditionality. This find-
ing may be particularly important for application in areas in which
reforms are politically sensitive (deregulation is a case in point).
* There are gains to be had from being sensitive to the effects of se-
quencing and timing. In the long run, for example, it may prove to
have been desirable that Pakistan improved export incentives and per-
formance in advance of introducing a strong trade reform program
(even granting that progress has been too slow in this area). Similarly,
improving private sector performance in oil and gas before attacking
this issue in the electric power sector may have improved the prospects
for reform in the power sector.
• Trying simultaneously to monitor the macroeconomic situation and
hold a government to standards of performance under a program of
SECALS without a concurrent IMF program or a SAL has proved to be
unsuccessful. In Pakistan the SECALS are supporting useful sectoral re-
forms, but World Bank attempts to use the size of its lending program
and the proportion and number of SECALS to exercise leverage over
macroeconomic performance issues have failed to stem an extended
decline in budgetary (and credit) performance.
A Postscript for Fiscal 1988 and 1989
In fiscal 1988 Pakistan's economic performance continued to show the
effects of large fiscal deficits. Although growth continued at a respectable
7 percent rate, inflationary pressures accelerated and the balance of pay-
ments situation deteriorated. At 8.6 percent of GDP, the budget deficit



434 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
exceeded both original projections and the 1987 deficit by about 0.5
percent. Most inflation indicators jumped into the 5 to 10 percent a year
range, some 2 to 3 percentage points higher than in previous years. Al-
though export growth was strong (25 percent in current dollars), imports
also surged and remittances continued their decline, which led to a current
account deficit of 4.4 percent of GDP (more than double that of the pre-
vious year). By year's end, gross official foreign exchange reserves were
the equivalent of only three weeks of imports, which was too low to meet
any unanticipated short-term fluctuations.
The government has begun an ambitious program of reforms covering
the 1989-91 period, partly in the context of an agreement with the IMF
and a number of negotiated or potential sectoral loans from the World
Bank. Although at this writing the details remain to be worked out, some
important initiatives are aimed at reducing the budget deficit to sustain-
able levels, further liberalizing foreign trade and government regulations,
and improving cost recovery in agriculture. Among the most important
of these measures are increases in the coverage of the sales tax, a surtax
on higher income groups, substantial reductions in the imports subject
to quantitative restrictions, a reduction in the dispersion of tariff rates
(through reductions in the maximum tariff rate from 150 to 225 percent
to 100 percent and consolidation of rates on inputs into a narrower
range), further liberalization of investment controls, and reductions in
agricultural subsidies.
Notes
1. Expenditures in all the priority sectors, in real terms and as shares of total
development expenditures, were significantly higher in the sixth plan than in the
fifth. In most cases, however, the share did not reach the ambitious targets called
for in the sixth plan.
2. Deficits on the order of 5 percent of GDP would be consistent with the
government's macroeconomic targets and would keep foreign and domestic debt
roughly constant as a share of GDP. The calculation is based on the following
targets: real GDP growth of 6 percent; domestic debt/GDP of 35 percent; foreign
debt/GDP of 35 percent, base money/GDP of 16 percent; inflation of 5 percent;
real exchange rate depreciation of 4 percent; domestic real interest rate of 6
percent; and foreign real interest rate of 3 percent.
3. Fiscal years. Debt service includes the service on all forms of debt and
repurchases and interest payments on IMF programs. Exports include all goods
and nonfactor services, factor services, and current transfers.
4. Although the deterioration in the government's budgetary deficit following
the reduction in the deficit under the IMF program could be viewed as a reversal,
it can more strongly be argued that no reversal took place because there had
been no real reform. The improvement during the IMF period was achieved largely
by restraining budgetary expenditures; no important reforms in tax or subsidy
policy were introduced that would have made the low budget deficits sustainable
over time.



Pakistan: Structural Adjustment and Economic Growth 435
Comments
Mohsin S. Khan
WILLIAM MCCLEARY'S chapter provides a useful account of economic de-
velopments in Pakistan during the last twenty-five years or so and, in
particular, gives a clear and concise picture of the role of the World Bank
in Pakistan during the 1980s.
Pakistan can in certain respects be viewed as a semisuccess story. For
example, in 1972-86 real gross national product (GNP) grew at an average
annual rate of over 6 percent, and inflation was kept below 12 percent
a year on average. In comparison with many other developing countries
this is a most creditable performance. The current account deficit was
large by developing country standards, however (averaging nearly 5 per-
cent of GNP), and although a large part of it was financed by official aid
flows, it did lead to a significant buildup of foreign debt.'
McCleary's chapter goes behind these numbers to tell how and why
Pakistan was able to achieve this fairly impressive performance, especially
in the face of severe external shocks, both economic and political, that
hit the country in the 1970s and 1980s. The rendition of history is rea-
sonably accurate, and there is little one could disagree or take issue with.
But the real purpose of McCleary's chapter seems to be to make the
case that the World Bank played a significant (and certainly positive)
role in the development of the economy during the 1980s through the
June 1982 structural adjustment loan (SAL) and subsequent sectoral ad-
justment loans (SECALS) of 1985 for energy and 1986 for export devel-
opment. It is with this aspect of McCleary's analysis that I have some
difficulties and disagreements.
McCleary contends that circumstances in the 1980s were quite differ-
ent from those in the 1970s and, by implication, that the 1982 SAL and
the 1980-83 arrangement with the International Monetary Fund (IMF)
had a lot to do with Pakistan's purportedly better performance in the
second period.2 He believes that a major contributing factor to-this per-
formance was the government's promarket and outward-looking ap-
proach to development in the 1980s. I would argue, rather, that there
was not a great deal of difference in performance between the two decades
and that the major economic problems of the 1970s continued into the
1980s-and, for that matter, still remain to be addressed.
As can be seen from the data for the main macroeconomic aggregates
table 19C-1), growth rates in the two periods were almost identical.



436 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 19C-1. Macroeconomic Indicators for Pakistan, 1972-79 and
1980-86
(percent)
Fiscal             Fiscal
Indicator                                 1972-79             1980-86
Growth of real GNP                           6.3                 6.4
Inflation                                   14.4                 8.0
Current account deficit (percentage of
GNP)                                       6.0                 3.4
Debt service ratio                          23.9                24.9
External debt (billions of U.S. dollars)     8.6                12.0
Note: Fiscal year starts July 1.
Source: Pakistan Economic Survey, 1985-86 (islamabad: Government of Pakistan, 1986).
Although inflation was cut by about half, a good part of the increase in
the earlier period was a result of the oil price shock in 1973-74. If those
two years are dropped from the calculation, the average annual rate of
inflation in the 1970s falls to less than 10 percent. A similar adjustment
to the current account deficit brings that down from an average of 6
percent of GNP to 4.6 percent of GNP, as compared with 3.4 percent of
GNP during 1980-86. Debt servicing was somewhat larger in the second
period as the stock of external debt rose sharply. Certainly, it is not
obvious from the numbers that the 1970s was characterized "by generally
poor performance and relative stagnation" or that the 1980s were so
much better on all fronts.
Why was economic performance so similar in the two periods? I would
attribute this to the following three factors:
1. Same policymakers. For some twenty years the overall direction of
economic policy has been in the hands of a few individuals, who
generally are not part of the civil service but are supported by it.
This small group of technocrats has maintained its central role in
economic policymaking despite changes in political regime.
2. Same political and institutional constraints. The continuing tensions
with neighboring countries have kept defense needs high and ab-
sorbed considerable resources. In addition, the power of the agri-
cultural lobby, which has prevented the imposition of an agricul-
tural income tax, was as strong in the 1980s as it was in the 1970s.
3. Same policies. The policymakers continued with the same mixture
of orthodox policies and direct intervention. The only major change
was in the active use of exchange rate policy beginning in 1982.
Despite the urging of the World Bank and the IMF, structural reform
has proceeded very slowly.



Pakistan: Structural Adjustment and Economic Growth 437
The 1982 SAL was designed to raise savings and investment, strengthen
the balance of payments, improve the fiscal situation, raise agricultural
production, and reduce protection, among other changes. How success-
fully were these goals achieved? "Significant" progress was achieved on
the exchange rate (the adoption of a managed floating rate in January
1982), pricing policies, and the restoration of medium-term planning.
"Little or no" progress was achieved in fiscal reform, tariff reform, and
privatization of public enterprises.
Although I do not wish to downplay the positive achievements of the
SAL, unfortunately it was unsuccessful in the two areas that seem to me
to matter the most-fiscal reform and trade reform. Correcting fiscal
imbalances has proved to be an intractable problem for successive gov-
ernments. From 1972 to the present the fiscal account has been in deficit,
and though attempts have been made on several occasions to reduce the
deficit, any success has been temporary and limited.
In contrast to the situation in many developing countries, however,
the fiscal deficit has not led to spiraling inflation or lowered growth.
Rather, the behavior of fiscal deficits is matched by current account def-
icits (see figure 19C-1). Except for one year (1982) there is almost a one-
for-one correspondence between fiscal deficits and current account def-
icits. These fiscal deficits have been financed in the main by foreign cap-
ital, although in more recent years the government has resorted to fairly
expensive domestic borrowing.3
Since it has been widely recognized that the fiscal situation has put the
balance of payments under continual strain and that this position is
clearly unsustainable, why has there been no effort to raise sufficient
revenue or, failing that, to cut expenditure on a more lasting basis? On
the revenue side, it is clear that the tax base of the economy is small.
The agricultural sector, which accounts for 25 percent of GDP, has been
exempt from direct income taxes. No government has yet been able to
impose an agricultural income tax despite the frequent and well-reasoned
recommendations of taxation commissions, the IMF, and the World Bank.
The problem of revenue is further compounded by the low tax compliance
rate-it has been estimated that roughly three-quarters of potential tax-
payers do not pay any income tax. The government has relied on ad hoc
measures and indirect taxes (particularly trade taxes) to raise revenue.
On the expenditure side, the major share of current expenditure (ap-
proximately 35 to 40 percent) has been going to defense. In recent years
interest payments on external and domestic debt have risen to about 20
percent of current expenditure. With another 10 percent or so going for
subsidies on essential goods, these figures indicate that close to 70 percent
of the current budget is virtually untouchable. When the need to continue
expenditure for infrastructure and social programs is added in, govern-



438 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Figure 19C-1. Current Account and Fiscal Deficits in Pakistan,
1972-85
Percentage of GNP
0
-2
Current account
6Rl          (XZ~defci
-4-
-12     X                    . X . . l 
1972    1974    1976    1978   1980    1982    1984
Source: Pakistan Economic Survey, 1985-86 (Islamabad: Government of Pakistan,
1986).
ments find themselves needing to increase expenditure continually while
revenue is stagnating.
This budget crunch was a serious problem in the 1970s, and it has, if
anything, gotten worse in the 1980s, despite the best intentions of the
SAL. The Pakistani government, with the support of the World Bank and
the IMF, has now begun to move on revenue generation, but agricultural
income taxes are still off the table.



Pakistan: Structural Adjustment and Economic Growth 439
Since fiscal policy has remained expansionary, quantitative controls
on trade and payments have been regarded as essential for the balance
of payments.4 Although there have been some attempts at liberalization,
the evidence is far too tenuous to argue, as McCleary does, that the
economy has become more open and outward-oriented in the 1980s.
Broadly speaking, the economy remains regulated, and most foreign trade
and capital transactions are still subject to government approval. The
changes that have been made have been marginal, and it is doubtful
whether those made in the 1980s have been any greater than those of
the previous decade. For example, although the government in 1983
moved from the "positive" list of items that could be imported (with all
others banned) to a "negative" list of items that were banned (all others
could be imported), it is noteworthy that this negative list included 6,000
items. McCleary's reference to the removal of 136 items from the negative
list has to be put into that context. The system still has significant barriers
to trade and an antiexport bias.
In conclusion, I would not want to say that the SAL and the structural
policies pushed by the World Bank did nothing for the economy. Rather,
I believe that McCleary has chosen the wrong counterfactual. More in-
formative would have been a comparison of Pakistan's performance
under the SAL with the performance of other similarly placed developing
countries without SALS. The approach selected here of comparing pre-SAL
performance (the 1970s) with performance under the SAL (the 1980s)
yields, as I have tried to show here, ambiguous results. The situation in
Pakistan was not as bad in the 1970s nor as great in the 1980s as
McCleary indicates, but my guess is that relative to the performance of
many other developing countries, Pakistan probably did much better than
average in the 1980s. If so, one could give some of the credit to the SAL,
as well as to the IMF and the government of Pakistan. There would be
enough credit to go around.
Notes
1. Between 1971 and 1986 Pakistan received grants and concessional loans
of more than US$20 billion. External debt in 1986 was approximately US$12
billion.
2. It is clearly very difficult to disentangle the effects of the SAL from that of
the IMF program since many of the policies overlapped.
3. For example, floating and unfunded debt of the government of Pakistan
rose from 39.5 billion rupees (17 percent of GDP) to 141.4 billion rupees (26
percent of GDP) between 1975 and 1986.
4. Exchange rate policy, of course, also played a role in the early 1970s after
the major devaluation in 1972, and in the 1980s when the rupee peg to the U.S.
dollar was eliminated.



20 Turkey: Structural Transformation
and the Threat to Sustainability
Faezeh Foroutan
DURING THE 1960s AND 1970s Turkey's trade policy was characterized
by import substitution with high protection, periodic overvaluation of
the exchange rate, and a considerable degree of antiexport bias. Despite
this inhospitable setting, however, gross national product (GNP) grew by
approximately 6.7 percent a year between 1963 and 1973 (Celasun and
Rodrik 1989). This economic growth was to a great extent stimulated
by a rapid increase in worker remittances. The remittances permitted
investment to increase almost enough to offset the deterioration occurring
in investment efficiency as measured by the incremental capital-output
ratio.'
As the country's balance of payments position worsened in 1969, the
government negotiated a stabilization program with the International
Monetary Fund (IMF). The program featured a large devaluation of the
Turkish lira and the introduction of several liberalization measures. After
the military coup in 1971, the liberalization measures were never im-
plemented, but the lira was devalued considerably in 1970. The deval-
uation, favorable world economic conditions, and the continuing, buoy-
ant inflow of remittances from Turkish workers abroad helped Turkey
expand its exports and GNP from 1971 to 1973. Despite the negative
impact of the oil price increase of 1973, the economy grew by more than
7 percent a year between 1973 and 1977, the years corresponding to the
ambitious Third Economic Plan. This growth was financed by drawing
down reserves and heavy short-term borrowing from abroad. Short-term
foreign borrowing, estimated at 7 percent of GDP in 1977, was intensified
during 1975-77 through the introduction of a new instrument, the Con-
vertible Turkish Lira Deposit (CTLD).
By 1977, as foreign lenders became aware of the magnitude and un-
stable structure of Turkey's foreign debt, foreign capital flows ceased.
As a result of this foreign debt crisis and the ensuing shortage of foreign
exchange, private investment and economic growth collapsed. Because
Turkish imports consisted mainly of intermediate and capital goods, the
foreign exchange shortage caused a shortage of essential investment
goods. Total investment fell from a peak of 25 percent of GNP in 1977
440



Turkey: Structural Transformation and the Threat to Sustainability 441
to 18.5 percent of GNP in 1978. The private sector suffered the brunt of
the adjustment.
In the absence of foreign borrowing and a lack of infrastructure to
permit domestic borrowing, the budget deficit was financed almost ex-
clusively by money creation. As a result, inflation soared to a peak of
141 percent in May 1980 (see table 20-1). Accelerating inflation had a
negative impact on both the real exchange rate and the pattern of income
distribution. A series of nominal devaluations was undertaken, but the
real exchange rate continued to appreciate and Turkey's competitiveness
in international markets continued to deteriorate. Income distribution
also worsened, with the rural sector and wage earners hit the hardest.
Real wages fell by almost the same rate as the rate of the inflation, with
a cumulative reduction of 43 percent during 1978-80. The agricultural
terms of trade in 1979 were below the level of 1970 (Celasun and Rodrik
1989).
During the 1977-79 crisis the government made no fundamental
changes in its economic philosophy and management. Two standby ar-
rangements with the IMF failed. In the face of this increasing economic
deterioration, a new government was formed and its economic reform
program was submitted in January 1980. The program represented a
radical departure from previous policies and effectively paved the way
for support from the IMF and the World Bank and for the resumption of
private lending to Turkey.
The 1980 Adjustment Program
Turkey's economic reform program was put together by a group of tech-
nocrats headed by the prime minister, Turgut Ozal. In the short run the
primary goal was to achieve a reasonable degree of price stability, even
though this would have to come at the cost of a temporary sacrifice in
growth. In the longer run the program aimed at a gradual reduction in
the extent of state intervention in production, principally by reforming
state economic enterprises; shifting to a more outward-oriented, market-
based trade regime; and removing barriers to foreign direct investment.
IMF Programs
In support of the government's 1980 program, a new three-year standby
arrangement was reached with the IMF in June 1980 involving 1.25 billion
in special drawing rights2 (approximately US$1.6 billion). Turkey was
the first country to have a three-year standby arrangement with the IMF.
The key conditions included maintenance of a realistic and flexible ex-
change rate; improvement of public sector finances, principally via price
increases in state-owned enterprises in line with inflation; strict limits on



Table 20-1. Basic Economic Indicators for Turkey, 1980-88
Indicator                      1980       1981       1982       1983       1984       1985        1986       1987T       1988a
Macroeconomic balances (percentage of GNP)
Total consumption               84.1       82.0       81.8        83.7       83.3       80.9       77.7       76.1         73.7
Public                        12.3       10.7        10.8       10.2        8.8        8.4        8.8         9.1         8.7
Private                       71.9       71.3        71.1       73.5       74.5       72.5       68.9        67.0        65.0
Fixed investment                19.5       18.9        18.9       18.8       18.0       20.1       23.6       24.1         24.2
Public                        10.9       11.7        11.5       10.6        9.9       11.7       14.0        12.9        11.5
Private                        8.5        7.2         7.3        8.2        8.2        8.4        9.6        11.2        12.7
6    Stock changes                     1.9        2.6        1.5        1.1        1.4        0.9         1.4        1.3          0.0
9      Public                          0.5        1.5        0.5      - 0.4        0.0       - 0.2        0.1        0.4        - 0.6
Private                        1.4        1.1         1.0        1.5        1.4        1.0        1.3         0.8         0.6
Current account                -5.5        -3.5       -2.2       -3.5       - 2.8      - 1.9      -2.6        -1.5          2.1
Importsb                      15.0       16.5        17.9       19.5       23.1       23.5       20.7       20.3         23.6
Exportsb                       7.2       10.9        14.5       15.7       19.5       20.7       18.0        22.3        21.9
GNP (millions of U.S. dollars)  58,327   58,925     53,737     51,237      50,111    53,238      58,082     64,790         n.a.
GNP growth (percent)           - 1.1        4.1        4.5        3.3         5.9        5.1        8.1        7.4          3.4
Gross external debt (billions of U.S. dollars)
Total                           16.3       16.9       17.6        18.2       20.8       25.5       32.5       38.3         37.7
Medium to long term           13.8       14.7        15.9       16.0       17.6       20.8       25.6        29.6        30.0
Short term                     2.5        2.2         1.8        2.3        3.2        4.8        6.9         8.7         7.7
Debt/GNP                        28.0       28.6       32.8       35.6       41.5        47.9       55.9       56.1         53.1
Debt/exports'                  284.1      198.3      175.0      192.9       180.5      194.5      260.5      227.2          n.a.



Real exchange rated
Average                           100.0        98.4        107.7       107.4       111.3       110.0       131.9        141.8         141.9
End of period                     100.0       106.3        110.0       109.4       103.0       114.1       135.4        143.7          n.a.
Nominal exchange rate (Turkish lira to U.S. dollars)
Average                            76.0       111.2        162.6      225.5        366.7       522.0       674.5        857.2       1,422.3
End of period                      90.2       133.6        186.8      282.8        444.7       576.9       757.8      1,020.9       1,814.8
Monetary aggregates (percentage of GNP)
Base money                          8.1         8.5          9.7        10.2         9.1         8.7         8.2          7.2'         n.a.
M2                                 14.9        17.4         22.1        23.0        21.7        23.2        23.5         21.4'         n.a.
M2X                                n.a.        n.a.         n.a.       23.0         22.8        26.0        28.4         28.3'         n.a.
Inflation rate
cpff (year average)               110.9        36.8         23.1        31.4        48.4        44.0        34.6         39.8         75.4
cPi (year end)                     89.6        28.3         26.2        37.1        49.7        44.2        30.7         55.1         75.2
n.a. Not available.
a. Estimated.
b. Goods and nonfactor services.
c. Exports of goods, services, and workers' remittances.
d. 1980 export weights.
e. June 1987.
f. Consumer price index.
Source: World Bank data.



444 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 20-2. Standby Arrangement between the IMF and Turkey,
1978-85
Millions of special drawing
rights
Period                 Committed    Disbursed          Status
April 1978-April 1980     300           90       Cancelled July 1979
July 1979-July 1980       250          230       Cancelled June 1980
June 1980-June 1983      1,250        1,250      Executed as planned
June 1983-June 1984       225           56       Cancelled April 1984
April 1984-April 1985     225          169       Expired April 1985
Total                    2,250        1,795
Source: Celasun (1989).
central bank lending to the public sector; and adjustment of interest rates
to market levels.
Two more one-year standby arrangements were agreed with the IMF
in June 1983 and April 1984 (see table 20-2). The 1983 agreement was
canceled early, and the 1984 agreement expired with the funds not yet
fully disbursed. Even with the two new standbys, Turkey's net position
with the IMF became negative in 1984, which added significantly to the
country's debt service burden. Moreover, the IMF'S supervisory role ended
at the same time as World Bank structural adjustment loans (SALS). Some
World Bank audit reports argued that it was premature for the IMF to
end its standby arrangements with Turkey in 1985 before stabilization
had been achieved. Other people have argued that five years should have
been sufficient time for Turkey to reach its stabilization goals; any IMF
presence beyond that period would simply have delayed even longer the
surfacing of the economic problems that Turkey needed to resolve-and
is currently facing.
The OECD Consortium and Debt Restructuring by Commercial
Banks
The Organisation for Economic Co-operation and Development (OECD)
assisted Turkey in the management of its foreign debt crisis by launching
a special rescue operation in 1979 and by arranging for the rescheduling
of Turkey's debt.
As part of the four-year rescue operation, the OECD consortium ar-
ranged four pledging conferences to increase the volume of bilateral con-
tributions to Turkey above the relatively low levels (US$100 million to
US$200 million a year) of the early 1970s. Almost US$4 billion was
pledged during the four conferences in the form of program loans, project



Turkey: Structural Transformation and the Threat to Sustainability 445
Table 20-3. Pledges to Turkey during the OECD Rescue Operation,
1979-82
(millions of U.S. dollars)
Category         1979       1980        1981      1982
Program loans   625.1        705.0     649.4      425.0
Project loans    46.2        230.0     124.1      200.0
Export credits  318.0        225.5     190.3      225.0
Total            989.3      1,160.5    973.8      850.0
Source: Wolff (1987).
loans, and export credits (table 20-3). The largest commitments came
from the United States and the Federal Republic of Germany. The timing
of disbursements and the nature of the loan conditions varied consid-
erably. Conditions ranged from soft terms on long-term program loans
to export credits tied to specific purchases at prevailing market rates.
After the last pledging conference in 1982, bilateral commitments fell
back to their pre-1978 level.
A working group within the OECD consortium arranged for the re-
scheduling of Turkey's official debt. In all, three rescheduling agreements
were signed with Turkey, in 1978, 1979, and 1980. The consolidated
debt rescheduled in the three agreements totaled US$5 billion.
Through its pledging conferences and rescheduling operations, the
OECD made a significant contribution to Turkey's financing requirements
during 1979-82. In all, the rescue operation resulted in a gross capital
inflow of at least US$3 billion. If the loan reschedulings are added to
this figure, the OECD'S contribution exceeded that of all other official
agencies. The intervention by the OECD-motivated in part by Turkey's
important geopolitical position-makes Turkey's experience atypical
among problem debtors. No other country among the highly indebted
nations received so much assistance from the international community.
This is an important explanatory factor for Turkey's outperforming the
other highly indebted nations in economic growth during the 1980s.
The commercial banks also helped Turkey meet its external financing
needs by rescheduling accumulated debt and supplying new money. The
banks' commitments were conditioned on the conclusion of an agreement
with the IMF. Turkey reached agreement with the banks in July 1979,
after the OECD rescue operation had begun and a second agreement had
been arranged with the IMF.
The rescheduling covered Turkey's short-term liabilities, amounting to
US$2.8 billion, mostly in the form of Convertible Turkish Lira Deposits.
The agreement provided for the repayment of the debt and the interest
arrears in four yearly installments beginning in 1982, and the interest



446 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
rate was to remain at 1.75 percentage points above the London interbank
offer rate. Turkey was not satisfied with this agreement, and in 1982 the
banks, under increasing pressure from official lenders, modified the orig-
inal terms by extending the life of the loan to ten years and the grace
period to five years.
After the rescheduling, the commercial banks showed a renewed in-
terest in providing Turkey with fresh funds. In 1983 Turkey received its
first new medium-term syndicated loan, and by 1985 commercial banks
had provided Turkey with US$1.4 billion in new loans in addition to the
US$2.8 billion of rescheduled debt. Although commercial banks clearly
made a significant contribution toward easing Turkey's external debt
burden, net transfers from commercial banks turned negative in 1981
and remained so until 1986.
The Role of the World Bank
The first adjustment loan to Turkey from the World Bank was in March
1980. The loan agreement was concluded only two months after the
Turkish government's announcement of its new economic policy pack-
age. By 1988 Turkey had received four more SALS and three sectoral
adjustment loans (SECALS) for a total of US$2.5 billion (see table 20-4).
These loans supported a comprehensive structural adjustment process
that began in January 1980. Although their emphases differed slightly,
all the loans addressed basically the same policy objectives-mainly those
in the government's 1980 program. The major objectives were:
* Export promotion, through direct incentives and institutional reform,
to increase foreign exchange earnings so that Turkey could meet its
growing need for imports
* Import liberalization to reduce antiexport bias and the cost of im-
porting the inputs necessary for the production of exports
Table 20-4. World Bank Adjustment Loans to Turkey, 1980-87
Amount
Loan                        Year          (millions of U.S. dollars)
SAL I and Supplement       1980                   275.0
SAL 11                     1981                   300.0
SAL III                    1982                   304.5
SAL IV                     1983                   300.8
SAL V                      1984                   376.0
SECAL, agriculture         1985                   300.0
SECAL, financial sector    1986                   300.0
SECAL, energy              1987                   325.0
Total                                           2,481.3



Turkey: Structural Transformation and the Threat to Sustainability 447
* Rationalization of public investment by reducing it to levels compat-
ible with revenue growth, shifting public investment toward infra-
structure while leaving private investment a greater role in manufac-
turing, and concentrating public investment on fewer projects with
higher priority and higher rates of return
* Reform of state enterprises by improving pricing policy, management,
and finance so as to enhance their efficiency and productivity and
increase their contribution to economic growth
* Development of domestic energy resources by adjusting domestic
prices to reflect international prices, by increasing the share of public
investment in energy infrastructure, and by rationalizing demand and
supply and reducing Turkey's dependence on imported fuel
* Reform of agriculture by modifying the structure of incentives and
institutions so as to increase agricultural output and exports
* Financial reform, through fiscal measures affecting financial trans-
actions and through institutional reform, to increase the stability of
financial institutions and improve their intermediation capabilities.
Other measures concerned development of a framework for medium-
term planning, management of external debt, and reduction in the rate
of inflation.
Performance under Structural Adjustment
Turkey has made enormous progress toward many of the policy objec-
tives supported by the SALs and SECALS, with the most spectacular achieve-
ments coming in trade. The measures adopted since the beginning of the
adjustment program have transformed the country from an inward-ori-
ented, almost closed economy into an outward-looking, open economy.
However, other unresolved problems, such as the worsening income dis-
tribution and fiscal deficits, could jeopardize the sustainability of the
progress made so far if corrective actions are not taken. The following
sections examine the extent of Turkey's fulfillment of the policy objectives
and conditions of the SALs and SECALS.
Export Promotion
Export promotion was central to the 1980 government program and to
all the SALS. It was achieved in three ways: maintenance of a competitive
real exchange rate, provision of direct subsidies, and simplification of
administrative and bureaucratic procedures (see Milanovic 1986; For-
outan 1987).
The lira was devalued by more than 50 percent in nominal terms in
1980. At the same time, except for a few agricultural inputs, the multiple



448 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 20-S. Turkish Export Subsidies, 1980-86
(percentage of total exports)
Subsidy           1980    1981    1982    1983    1984    1985    1986
Export tax rebate    0.0   3.6    10.1    11.5    11.1      9.7     7.6
Duty-free imports    5.5   4.4     4.2      5.4    2.9      4.9     6.2
Export credit     16.6    12.5     6.4      6.5     1.1     0.0     0.0
RUSFa              0.0     0.0     0.0      0.0     0.0     4.0     2.0
Total             22.1    20.5    20.6    23.4    15.1    18.6    16.0
a. Resource Utilization Support Fund.
Source: Milanovic (1986); Foroutan (1987).
exchange rate system was abolished. After the initial nominal devalua-
tion, the government pursued a policy of slow real devaluation of the
Turkish lira that resulted in a cumulative real depreciation of over 33
percent between 1980 and 1986. The government also provided direct
incentives to exporters. These measures gave exporters substantial sub-
sidies that to a large extent neutralized the antiexport bias implicit in the
import protection system (see table 20-5). A third important support to
exporters-although its results are unquantifiable-was the simplifica-
tion of procedures for obtaining export incentives and the centralization
of control over export policy in the State Planning Organization and the
Directorate for Exports.
The response of exports to these three measures was spectacular. Mer-
chandise exports rose from US$2.9 billion in 1980 to US$10.3 billion in
1987. Similarly, exports of goods and nonfactor services increased from
7 percent of GDP in 1980 to 21 percent of GDP in 1985.
Optimism about the continuation of such successful export perfor-
mance must be tempered, however, by four considerations. First, exports
started booming during the early 1980s when there was widespread ex-
cess productive capacity. This allowed exports to expand without re-
stricting domestic consumption. Today, capacity-utilization rates are
much higher (65 to 85 percent, with an average of 72 percent), and further
export expansion is possible only through investment in new capacity.
As discussed later, industrial investment in the 1980s was rather stagnant,
in large part because of very high real rates of interest. Since interest
rates are unlikely to fall without a drastic fiscal contraction, the extent
of future investment and export expansion remains uncertain.
Second, after 1981 the Iran-Iraq war was important in boosting Turk-
ish exports (see Celasun and Rodrik 1989). Given the general decline in
the purchasing power of the oil-exporting countries and the recent shift
of Turkish exports to OECD markets, it seems unlikely that Turkey could
increase its exports to these markets at the spectacular rates of growth
achieved in the early 1980s.



Turkey: Structural Transformation and the Threat to Sustainability 449
Third, because export subsidies not only are costly but also violate the
rules of the General-Agreement on Tariffs and Trade (GATT), it is doubtful
whether they can be sustained and consequently whether exports can
continue to grow at the same rate as in the early 1980s.3 Fourth, real
wages have been squeezed so hard that it is not clear that the policy of
real devaluation can be sustained much longer.
Import Liberalization
The liberalization of imports constituted another focal point of the Turk-
ish government's 1980 policy package and of the sALs. The principal
measures were carried out in two steps. First, in mid-1981 the quota list
was eliminated. Second, and more far reaching, at the end of 1983 the
positive list of imports, which allowed in only the listed commodities,
was replaced by a negative list, which allowed the importation of all
commodities not explicitly prohibited. By 1988 most other quantitative
restrictions in the form of prohibitions and licensing requirements were
also abolished.
In addition, a large-scale rationalization of tariffs brought about a
reduction in the average rate and the dispersion of rates. The objective
of the tariff reform, as agreed with the World Bank, was to lower the
average rate of nominal tariff protection to 7 percent by the end of the
fifth five-year plan (1989). After 1984, however, import liberalization
slowed to some extent because of the growing importance of import levies
administered by the extrabudgetary funds. In 1987 levies stood at 6.1
percent of the value of imports, or about 60 percent of the weighted
average tariff rate.
Fiscal Stability and Domestic Resource Mobilization
A reduction in the public sector deficit was a principal requirement of
the IMF stabilization program and the World Bank's adjustment lending.
Little progress was made on this front, however, because of the govern-
ment's failure to reduce inflation and because of high real interest rates
in Turkey.
Inflation can be viewed as a residual tax that allows the government
to finance the difference between its total expenditures and all other
sources of revenue at its disposal. A similar equilibrating function is
played by the real rate of interest. In the absence of perfect capital mo-
bility, because of controls on capital flows or imperfect substitution be-
tween domestic and foreign interest-bearing assets, and given the current
account target, the interest rate must be high enough to elicit sufficient
net savings from the private sector to close the gap between the current
account surplus and the public sector deficit. Alternatively, the real rate



450 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
of interest in Turkey can be viewed as being determined by the real world
interest rate and the expected rate of real depreciation of the Turkish lira
plus a risk premium. Moreover, on the assumption that the risk premium
increases with the stock of government debt, a direct link is again es-
tablished between the government deficit and the domestic real interest
rate.
The public sector deficit, as measured by the public sector borrowing
requirement in current prices, improved considerably in 1981 and then
remained at about 5 percent of GNP until 1985 (table 20-6). The deficit
worsened considerably thereafter as a result of the growing burden of
interest payments on domestic and foreign debt and the government's
inability to cut expenditures elsewhere. (The high deficit figure in 1984
reflects a shortfall in tax revenue in that year.) Table 20-6 points also to
considerable changes in the sources of financing of the deficit. As the
government sought greater control over inflation, recourse to central
bank financing was curbed and replaced by domestic borrowing. As men-
tioned, this growing recourse to debt financing substantially worsened
the deficit because of high interest payments (these are included in the
net transfer figures in table 20-6).
Table 20-6. Public Sector Revenue and Expenditure in Turkey,
1980-87
(percentage of GNP, based on current prices)
Item                  1980  1981  1982  1983  1984  1985  1986  1987a
Revenueb               19.8   22.2  22.9  22.4  21.8  23.8   28.7  28.0
Current transfersc     2.3    2.9   3.2   5.0   5.2   6.1    9.3  11.0
Total disposable income  17.5   19.3  19.7  17.4  16.6  17.7   19.4  16.9
Investment expenditure
(including stock
changes)           11.5   13.2  12.0  11.5  10.0  11.0   13.6  13.5
Consumption            12.2   10.7  10.8  10.1   9.0   8.5    9.1   9.4
Total expenditure     23.7   23.9  22.8  21.6  19.0  19.5   22.7  22.9
Public sector borrowing
requirement
(estimate)          9.9    3.7  5.0   5.3   7.9   4.9    6.0   8.2
Net financing
External borrowing    3.2    2.5   1.0   1.4   2.7   0.5    2.8   4.1
Domestic borrowing    1.1    1.4  1.5   0.9   2.3   2.7    2.9   2.9
Central bank          3.5    2.0   0.3   0.6   0.7   1.3    0.7   1.0
Other                 2.2  -2.2   2.2   2.3   2.1   0.4  -0.4   0.2
a. Provisional estimates.
b. Includes factor income from property and the net surplus of social security institutions;
excludes wealth tax and capital flows.
c. Includes subsidies and interest payments; excludes capital transfers.
Source: Celasun (1989).



Turkey: Structural Transformation and the Threat to Sustainability 451
The increasing burden of interest on the foreign debt reflects mainly
the increasing share of nonconcessional loans after 1984. The interest
payment on the central government's and the central bank's foreign ex-
change liabilities increased from approximately 1.6 percent of GNP before
1984 to 3.5 percent in 1987. The interest payment on domestic debt also
climbed dramatically. For the central government alone, interest pay-
ments rose from less than 1 percent of GNP before 1984 to an estimated
2.5 percent in 1987. The real interest on treasury bills was about 12
percent in 1987. This high real rate of interest is related partly to the
generally high real interest rates in Turkey and partly to the way the bills
are auctioned. The bills compete with commercial bank loans rather than
commercial bank deposits and must therefore offer a comparable interest
rate to be attractive to the banks.
Clearly, continued reliance on domestic debt issue is unsustainable at
such high rates of interest. These rates are far in excess of the rate of
growth of GNP, so the increase in the interest expense of domestic debt
exceeds by a wide margin the rate of increase in the debt itself. Thus the
government must borrow ever increasing amounts just to finance its in-
terest payments.
What has been the composition of public sector expenditure? How has
it affected consumption and investment expenditure in the private sector?
These questions are addressed in the following subsection.
Public and Private Investment
In the 1980s the increase in public sector expenditure was due exclusively
to investment expenditure. Between 1980 and 1988 government con-
sumption as a share of GNP fell from 12 percent to 9 percent, whereas
its investment expenditure increased from 11 percent of GNP in 1980 to
14 percent in 1986 (see table 20-1). Public investment in 1987 accounted
for 60 percent of government expenditure, up from 47 percent in 1980.
This increase in investment occurred against the advice of the World
Bank, which recommended that the overall level of government invest-
ment be reduced to match available resources. Although public invest-
ment expenditure remained at high levels, the government did shift its
composition, in accordance with World Bank guidelines, away from man-
ufacturing and into infrastructure. By 1987 public investment in man-
ufacturing represented only 6.5 percent of total public fixed investment
compared with almost 29 percent in 1980.
It has been argued that the shift in public investment in favor of in-
frastructure may have helped to "crowd in" (as opposed to crowd out)
private investment (World Bank 1988b). In a recent study, however,
Conway (1989) finds that any crowding-in effect of public investment
on private investment relates to private investment in the housing sector;



452 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
in other sectors, although public investment may have encouraged private
investment initially, this positive impact seems to have reversed itself later
with lagged negative effects. Given Conway's additional findings that
real currency depreciation, high inflation, and high interest rates have
all had a detrimental effect on private investment, it can be concluded
that, overall, the developments in the public sector deficit have been a
hindrance to private investment in the productive, export-oriented sectors
of the economy.
Implementation of Other Objectives
The government's performance in implementing the other major objec-
tives of the SALS, except for the control of inflation, was generally sat-
isfactory, although a number of concerns remain. The failure to lower
inflation permanently was strictly related to developments in the public
sector deficit. Moreover, reform of the financial sector, by reducing the
inflation tax base, contributed to an increase in the average rate of in-
flation for each level of the deficit.
Reform of the state economic enterprises, particularly price deregu-
lation, considerably improved their operating profits, which became pos-
itive in 1984 and accounted for an estimated 1.5 percent of GNP in 1985.
Because of heavy investment by the enterprises, however, their overall
surplus remained negative even though it improved from - 11 percent
of GNP in 1980 to - 4.5 percent in 1985. Some concern has been expressed
that the improvement in the profitability of the state economic enterprises
may owe more to their freedom to increase prices than to greater effi-
ciency and reduced costs.
Sectoral Performance
After the completion of the fifth SAL, Turkey received a series of SECALS
for agriculture, energy, and finance, the sectors that had earlier received
particular attention in the five SALs.
AGRICULTURE. During the 1970s the agricultural sector, much like the
industrial sector, grew by following inward-looking policies that stressed
food self-sufficiency through subsidized input prices. As part of the struc-
tural adjustment program begun in 1980 and supported by the various
SALs and an agricultural SECAL, many of these incentive policies were
dismantled and more market-oriented policies were introduced. Input
subsidies and production price supports were reduced, preferential credit
was curtailed, and exports were encouraged through exchange rate de-
valuation and the relaxation of export-licensing restrictions. Agriculture
as a share of public investment increased from 7 percent in 1980 to 10



Turkey: Structural Transformation and the Threat to Sustainability 453
percent in 1985. As a result of improved criteria for project selection,
completion rates for agricultural projects also rose considerably. Com-
modity support prices were replaced by a system of protective floor prices.
In addition, the importation and distribution of fertilizers and seeds were
liberalized.
The results of these reforms were generally positive once the initial
shock of the subsidy reduction was overcome. The rate of growth of
agricultural GDP had fallen to less than 1.5 percent a year during the late
1970s and to 0.9 percent a year in 1980 and 1981. Thereafter, sectoral
growth recovered to an average 4 percent a year from 1983 through
1987. Exports of unprocessed agricultural products increased in volume
by over 6 percent a year after 1980, but they decreased in value because
of adverse world price movements. Exports of processed agro-based
products quadrupled after 1980 and now represent about one-third of
the combined total for agricultural and agroindustrial exports (about 10
percent in 1980).
ENERGY. The government's energy policy during the 1970s was dom-
inated by supply considerations, with pricing policy, demand manage-
ment, and energy conservation relegated to a low priority. Until the 1973
oil crisis Turkey had relied almost exclusively on oil imports for its energy
needs and had neglected its domestic sources of supply. The situation
changed in the second half of the 1970s as the government embarked on
a program to curtail energy imports through the substitution of domestic
supplies. Considerable resources were invested in the development of
indigenous sources of energy, especially lignite and hydropower. By 1984
the share of energy had risen to 40 percent of total government investment
and 3 percent of GNP (from 1.5 percent in the early 1970s).
Despite these efforts, the program failed to meet its ambitious targets,
mainly because inefficient management of the investment program spread
both human and financial resources too thinly across too many projects.
The result was frequent and protracted breakdowns in service that were
extremely detrimental to industrial users of energy. By the early 1980s
problems in the energy sector had reached crisis proportions. They were
deep rooted and almost all were linked to inefficiencies in planning, im-
plementation, finance, and operation in the government agencies con-
cerned with energy.
To address these problems the government adopted a new development
plan in 1984. It was supported by the World Bank which, after extensive
discussions with energy agencies, prepared an energy sector strategy
paper. The paper supplemented the government plan by recommending
actions to remedy the most pressing problems. The collaboration between
the World Bank and the government eventually resulted in the Energy
Sector Adjustment Loan of 1987. The government program supported



454 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
by the loan centered on three areas: institutional reform, including greater
private sector participation in the development and supply of energy and
greater autonomy for public agencies; economic investment, including
measures to minimize the cost of present and future public investment;
and improved efficiency, induding measures to maximize the operating
efficiency of all energy agencies and pricing policies to encourage energy
conservation.
FiNANCE. The most problematic of the three sectors under examination
was the financial sector. The government reform program introduced in
January 1980 contained few measures affecting the financial sector other
than those for deregulating commercial bank interest rates. The bank-
ruptcy of a major nonbank financial institution in late June 1982 refo-
cused attention on the structural ailments of the financial system. The
World Bank supported the new government initiatives by launching a
major financial sector study.
The study was the basis for measures adopted under SALS iv and v and
helped the government initiate several important financial reforms. The
reforms were directed toward reducing the undercapitalization of banks,
loosening the interlocking relationships between banks and corporations,
increasing the supervisory role of the central bank, and introducing a
deposit insurance scheme. The government also took steps to lower the
cost of bank intermediation by reducing various taxes on intermediation
and established the Capital Market Board and an auctioning system for
treasury bills. Financial reform was further advanced by measures intro-
duced during the first Financial Sector Adjustment Loan in 1986. In 1988
a second financial SECAL followed the path laid by the first. According
to recent reviews within the World Bank, however, the measures taken
have been insufficient to resolve the fundamental instability of the fi-
nancial sector in Turkey.
The instability of the banking sector and the financial distress of the
banks were reviewed extensively in the initiating memorandum for the
second financial SECAL (February 1988). The instability was attributed
to the large proportion of nonperforming loans in the portfolios of com-
mercial banks and the lack of an appropriate regulatory framework to
prevent banks from rolling these loans over indefinitely by capitalizing
the interest payments of their problem borrowers. The full extent of the
problem is unknown, but current estimates put the stock of bad loans
as high as one-third of the banking system's total credits. This means
that for some banks, potential losses would exceed the book value of
their capital and reserves.
In part, the large number of nonperforming loans was an outcome of
the structural reforms undertaken by the government beginning in 1980.
These reforms changed the profitability of firms operating under the



Turkey: Structural Transformation and the Threat to Sustainability 45S
previous inward-looking policy environment. The problem was aggra-
vated by the extremely high real rates of interest, which in turn were
related to the large budget deficits and the policy of continuing real de-
valuation of the Turkish lira.
In summary, the financial sector requires urgent intervention on several
fronts. It needs a stronger regulatory framework, a more effective system
of external bank audits, and a restructuring of insolvent banks. In the
long run, actions to reduce the very high real rates of interest are vital,
since high interest rates have been one of the principal contributing fac-
tors to the problem of nonperforming debt.
Income Distribution
Although income distribution was not an explicit objective of the gov-
ernment's 1980 policy reform package, the impact of the reforms on the
distribution of income in Turkey was profound. Real wages fell sub-
stantially, and the rural sector was adversely affected by the considerable
decline in the agricultural terms of trade. Despite the importance of in-
come redistribution to the sustainability of the adjustment policies, the
issue was never explicitly addressed in Turkey's SAL program.
Reliable data on employment and income distribution in Turkey are
scarce. A World Bank study (1988a) focusing on formal workers in the
nonagricultural sector (roughly 23 percent of the domestic civilian labor
force) estimated that real wages declined by more than 50 percent be-
tween 1977 and 1987. By 1985 real wages in manufacturing had dropped
to their 1965 level. Real wages and salaries as a share of nonagricultural
factor income were estimated to have fallen from 52 percent in 1977 to
22 percent in 1986.
The decline in real wages changed the factor distribution of income in
favor of capital. In the manufacturing industry, for example, it was es-
timated that the share of interest, rents, and profits increased to 70 per-
cent of manufacturing GDP from approximately 35 percent in the 1960s
and 1970s (World Bank 1988a, table 3.10). Although several changes
in the fiscal system (changes in income tax brackets and the tax rebate
system) since 1984 mitigated to some extent the decline in pretax labor
income, wages appear to have absorbed the brunt of the adjustment
burden. Moreover, despite the large decline in labor costs, the employ-
ment response was not dramatic. The growth in output seems to have
been based more on greater utilization of existing capacity and labor
than on an expansion of capacity or labor.
Agricultural labor experienced a similar deteriorating trend in income.
According to recent estimates (Boratav 1989), the agricultural terms of
trade (as measured by the implicit GDP deflator) deteriorated by 53 per-



456  COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 19 0s
cent during 1976-86. Such a dramatic relative price change is unique in
the postwar history of Turkey.
The dramatic squeeze on real wages together with the partial resto-
ration of civil liberties, including the right to strike, cast doubt on the
economic and political feasibility of curbing public sector expenditure
through a further reduction in the wage bill or a reduction in certain
popular programs, such as the low-cost housing subsidy. Similarly, since
real wages have already been squeezed severely, it is hard to believe that
the slow but systematic nominal depreciations of the currency will con-
tinue to be as easily translated into real devaluations through a further
squeezing of labor income.
The redistributional aspects of the adjustment process in Turkey have
received insufficient attention in the SAL program. The World Bank's view
that the deterioration in labor income in Turkey is a necessary, albeit
undesirable, side effect of the adjustment process fails to consider its
potentially adverse effect on sustainability.
Conditionality
For each SAL, the government announced the policies it intended to pursue
in a Statement of Development Policies. The loan was then granted in
support of these enunciated policies, and the time framework was spec-
ified. The commitments presented in the Statement of Development Pol-
icies can be treated as the true loan conditions even if they were not
specifically mentioned as conditions in the loan agreements or as con-
ditions for tranche release. The SALs and SECALS made infrequent and
rather imprecise mention of conditions for loan effectiveness or for
tranche release. SALS III to v, for example, used such phrases as "satis-
factory progress in carrying out" or "adequate progress achieved in."
SAL I was released in three tranches (the other SALS had only two) and
had the greatest number of precise conditions for tranche release, that
is, conditions requiring implementation of a specific action within a spe-
cific time frame. These conditions called for the initiation or completion
by specified dates of a review of the feasibility of an export credit and
risk facility, a study on the rationalization of the industrial protection
system, a study on external debt management, and a review of measures
to increase household savings. The results of these studies were expected
to guide future action in those areas.
SAL V was the only other SAL that included any precise tranche-release
conditions, namely, the completion of the long-awaited fifth five-year
plan. Of the SECALS, only the first Financial Sector Adjustment Loan
contained any specific conditions; one was that an external audit be
required for commercial banks and corporations issuing securities by the
beginning of fiscal 1987.



Turkey: Structural Transformation and the Threat to Sustainability 457
Table 20-7 summarizes the main conditions in each SAL and the extent
to which the conditions were fulfilled by the end of each sAL period.
Several observations can be made. First, each SAL seems to have so many
conditions that it is difficult to pinpoint the main ones. Second, sAL con-
ditions seem to focus mainly on trade liberalization rather than on fun-
damental macroeconomic balance. The World Bank program apparently
focused more closely on measures to increase the efficiency of the econ-
omy and left the surveillance of the macroeconomic picture to the IMF
stabilization program. Third, the delay in implementing some of the SAL
conditions, such as the introduction of a value added tax, was due to the
unrealistic time frame specified. This problem was apparently fully under-
stood by the World Bank, since none of the tranche releases was withheld.
In summary, the true SAL conditions were the government commitments
specified in its Statement of Development Policy. In this sense, one may
speak of a front loading of conditions. With the exception of SAL 1, con-
ditions for loan effectiveness or for tranche release were few and flexible;
it was left to the negotiators to determine whether satisfactory progress
had been made. This flexibility was useful in several respects: it kept the
dialogue open with the Turkish authorities; it was politically palatable
because it avoided the appearance of being imposed by the World Bank;
and it helped the Bank correct its overoptimistic evaluation of the fea-
sibility of carrying out certain reforms within the agreed-on time frame.
World Bank-IMF Relations
Since the first SAL, the World Bank and the IMF have collaborated closely.
They have not only exchanged information and participated in joint mis-
sions, but also reached agreement on the spheres of involvement. From
the beginning, the IMF had primary responsibility for providing advice
on macroeconomic issues such as exchange rate, monetary, and fiscal
policies. Other policy reforms were of interest to both institutions.
With the expiration of the last IMF standby arrangement with Turkey
in 1985, the supervisory role of the IMF ended. Recent argument suggests
that the IMF acted prematurely in ending its standby arrangements with
Turkey just as the last SAL was also expiring. The World Bank seems to
have had difficulty enforcing a sustainable macroeconomic framework
through the SECALS alone. The deterioration of the macroeconomic en-
vironment in Turkey since 1987 appears to lend indirect support to this
argument. It is also possible, however, to argue that if the IMF stabilization
program and the World Bank SAL program did not succeed in bringing
about a fundamental and lasting change in the macroeconomic picture
after five years, then continuation of the programs would not have made
much difference. It would only have delayed the surfacing of the economic



458 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Table 20-7. Turkey's Progress in Implementing Structural Adjustment
Loan Conditions by the End of Each Loan Period
Significant progress          Satisfactory progress      Little or no progress
SAL r and Supplement (1980, US$245 million)
Exchange rate management  Reduction in budget deficit  Establishment of export
Export promotion: priority   Autonomy of state          credit insurance scheme
credit to exporters;      enterprises in setting    Reduction in public
improvement of             prices                     investment
institutional            Public investment shifted  Reduction in subsidized-
arrangements for           toward infrastructure     interest loans to state
exporters; financial                                  enterprises
incentives to exporters;                            External debt management
access to duty-free
imports
Partial liberalization of
interest rates
SAL II (1 981, US$30 million)
Exchange rate management  Steps toward creation of a   Establishment of export
Import liberalization:       stock market               credit
abolition of quotas;     Further state enterprise   Introduction of a value
reduction of import        reform: constraints on     added tax
guarantee deposits;        employment; reduced      Reform of state enterprise
partial relaxation of      access to central bank    management
quantitative restrictions    Further rationalization of    Preparation of a medium-
Further export promotion     public investment         term economic
Reform of the direct tax   Incentives for agricultural  framework
system                     exports                 External debt management
Increase in prices of energy   Rationalization of input
to world market levels    subsidies for agriculture
SAL nr (1982, US$204.5 million)
Continued exchange rate    Continued rationalization    Development of a medium-
management                 of public investment      term economic
Continued import           Continued reform of state   framework
liberalization             enterprises: limit on   Increased efficiency of state
Reduction in bank            budgetary transfers;      enterprises
intermediation tax to     limit on hiring of new   External debt management
reduce cost of loans      personnel
Continued increase in      Continued reform of the
energy prices to reflect   agricultural sector:
market conditions         reorganization of the
Ministry of Agriculture;
extension of some
export incentives to
agriculture; phasing out
of subsidy on certain
inputs



Turkey: Structural Transformation and the Threat to Sustainability 459
SAL IV (1983, US$300.8 million)
Reform of the tariff system   Continued rationalization    Development of medium-
Further simplification of   of public investment       term economic
import procedures       Continued improvement in    framework
Continued export            the financial            Control of inflation
promotion                 performance of state     Reduction in real interest
enterprises               rates on nonpreferential
Reduction in public deficit    loans
to 0.6 percent of GNP   Law for regulating the
Replacement of              auditing and accounting
agricultural producer     profession
price subsidy with floor   External debt management
prices
Continued rationalization
of energy pricing and
investment
SAL V (1984, US$300 million)
Substantial import        Preparation of medium-     Export credit insurance
liberalization: reduction  term economic           Diversification of channels
in average tariffs to 7   framework (five-year       for marketing treasury
percent by the end of     plan)                      bills
five-year plan;         Review of public           Managerial reform of state
elimination of            investment program         enterprise
quantitative restrictions    Issue of treasury bonds for   Inflation control
Elimination of financial    the public               Balanced budget
transactions tax         Opening of the stock      Privatization of state
Continued autonomy of       exchange market            enterprises
state enterprises in     Continued reform of the
setting prices             agricultural sector
Continued policy of       Strengthening of energy
flexible exchange rate    conservation programs
Increased role of market in
setting interest rates
Source: World Bank President's Reports.
problems currently facing Turkey. After all, no country can be supervised
forever by multinational agencies.
Conclusions and Lessons
The case of Turkey is often cited as proof that there is life after debt.
Turkey has shown that it is indeed possible to reverse the orientation of
an economy from an inward-looking, strictly controlled one to an out-



460 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
ward-looking, more market-oriented one in a relatively short time when
the domestic and international environments are favorable and policy-
makers are committed to reform and adopt effective policy tools. A closer
look at Turkey's experience, however, shows that unless fundamental
macroeconomic aggregates are also stabilized on a sustainable path, eco-
nomic growth may be short-lived and progress may be reversed. In fact,
despite the remarkable success achieved during the 1980s, especially in
the areas of foreign trade and economic growth, Turkey today is facing
some of the same pressing problems that afflicted its economy over ten
years ago during the height of its debt crisis. Turkey's public sector deficit
is high and growing, inflation is high and accelerating, and the foreign
and domestic debt service burden is mounting while investment in key
sectors is stagnating.
That these conditions exist does not minimize the remarkable success
Turkey has achieved. But current conditions must be examined in order
to understand what has gone wrong and what can be done to preserve
these achievements. A clear understanding of the policy successes and
failures in Turkey is important not only for the case of Turkey, but also
for an overall evaluation of the effectiveness of the World Bank's SAL
program.
A look at the areas in which success has clearly been achieved, namely
high export and economic growth, reveals several lessons and observa-
tions. Of primary importance was the ability and willingness of the po-
litical authorities to carry out the necessary reforms. It must be stressed
that it was the Turkish government rather than any outside economic
agent that started the adjustment program in 1980. The international
community, including the World Bank and the IMF, simply lent their seal
of approval to the program through their financial assistance. The other
side of the coin, however, is that the financial assistance of these insti-
tutions and the rest of the international community was essential to Tur-
key's success. The debt reschedulings of the late 1970s and the early 198 Os
with the OECD consortium and the private commercial banks, as well as
the new funds provided by the World Bank, the IMF, and the private
lenders in the 1980s, enabled Turkey to spread over time the burden of
the necessary adjustment measures. Because sufficient financial resources
were available, Turkey was able to avoid the more painful alternative of
suppressing economic growth during adjustment, at least until now.
These same two factors-political commitment and external re-
sources-also explain the difficulty Turkey is currently facing in sus-
taining its growth momentum. Whereas the military regime in 1980-83
was able to implement its adjustment program, including the massive
redistribution of income, with relative ease, the newly restored demo-
cratic government has had more difficulty traveling along the same path.
Similarly, the end of concessional loans in 1984 and the continuation of



Turkey: Structural Transformation and the Threat to Sustainability 461
the policy of real devaluation have meant that the debt burden is growing
again and with it the public sector deficit, whose financing has accelerated
the inflationary pressures.
An examination of the areas in which the adjustment program has not
been so successful also offers valuable lessons. For example, the less
successful effort to reform the financial sector raises the issue of the
feasibility of sectoral reform in the presence of severe macroeconomic
imbalances. It also calls into question some of the World Bank's policy
prescriptions related to liberalization of the financial sector or the trade
sector before stabilization was clearly achieved. The Turkish experience
thus shows that financial deregulation and liberalization of the capital
account are inherently destabilizing in an economy that is plagued by
inflation, fiscal crisis, and continuous real depreciation (Aricanli and
Rodrik 1989).
In sum, despite Turkey's substantial adherence to World Bank policy
advice in the 1980s and the remarkable structural transformation of the
Turkish economy, the sustainability of progress achieved so far is un-
certain. Sustainability is principally threatened by the large and growing
public sector budget deficit and its negative impact on inflation, interest
rates, private investment, and the stability of financial institutions. Sus-
tainability is also threatened by the sharp deterioration in the purchasing
power of wage and salary earners and the rural classes over the past
decade. In retrospect, it seems both that the government did not pay
sufficient attention to the importance of macroeconomic stability as a
prerequisite for sustainable growth and that the SAL program failed to
incorporate a mechanism that would guarantee the maintenance of ma-
croeconomic stability. In particular, the SAL program focused more on
liberalization than on stabilization and remained largely silent on income
redistribution. Although liberalization has borne substantial fruit, deci-
sive additional corrective action is now required on the fiscal front to
preserve these hard-won benefits.
Notes
Many thanks to Bela Balassa, Jeffrey Balkind, Jaime de Melo, Francis Geary,
Robert Hunt, William McCleary, and Dani Rodrik for their very useful comments
on early drafts.
1. The ratio had risen from 1.6 in 1963-67 to 2.4 in 1968-72 (Balassa 1981).
2. IMF funds are denominated in special drawing rights (SDRS), based on a
basket of currencies.
3. Considering only the export tax rebate and cash grants, two practices that
are most likely to be in violation of the GAIT rules, the cost of export promotion
can be roughly estimated at US$516 million in 1986, or roughly 5 percent of
the central government's revenues. This figure, obtained by multiplying the total
value of industrial exports by the average subsidy rate of the two measures,



462 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
represents an upper limit because exports of certain manufactures to western
markets are not eligible for the cash subsidy.
References
Aricanli, Torsun, and Dani Rodrik, eds. 1989. The Political Economy of Turkey:
Debt, Adjustment, and Sustainability. London: Macmillan.
Balassa, Bela. 1981. "Policies for Stable Economic Growth in Turkey." In The
Newly Industrializing Countries in the World Economy. Elmsford, N.Y.: Per-
gamon Press.
Boratav, K. 1989. "Inter-Class and Intra-Class Relations of Distribution under
Structural Adjustment: Turkey During the 1980s." In Aricanli and Rodrik
(1989).
Celasun, M. 1989. "Fiscal Aspects of Adjustment in the 1980s." In Aricanli and
Rodrik (1989).
Celasun, M., and Dani Rodrik. 1989. "Debt, Adjustment, and Growth: Turkey."
In Jeffrey Sachs, ed., Developing Country Debt and Economic Performance:
Country Studies. Chicago, Ill.: University of Chicago Press.
Conway, P. 1989. "A Record on Private Investment in Turkey." In Aricanli and
Rodrik (1989).
Foroutan, Faezeh. 1987. "Trade Policy in Turkey in 1986 and 1987: An Eval-
uation." Trade Policy Division, World Bank, Washington, D.C.
Milanovic, Branko. 1986. Export Incentives and Turkish Manufactured Exports,
1980-1984. World Bank Staff Working Paper 768. Washington, D.C.
Wolff, P. 1987. Stabilization Policy and Structural Adjustment in Turkey, 1980-
198S. Berlin: German Development Institute.
World Bank. 1988a. "Turkey Country Economic Memorandum: Towards Sus-
tainable Growth." Report 7378-TU. Washington, D.C.
*= . 1988b. "Turkey: External Debt, Fiscal Policy and Sustainable Growth."
Washington, D.C.
Comments
Dani Rodrik
As THESE WORDS were written (April 1989), Turkey appeared to be headed
toward political instability and economic chaos. The government of Tur-
gut Ozal, the architect of the economic reform program pursued since



Turkey: Structural Transformation and the Threat to Sustainability 463
January 1980, had just been badly beaten in local elections. The defeat
was due at least in part to Ozal's inability to bring down inflation despite
applying a rather strong squeeze to the economy during the second half
of 1988.
Is Turkey at the end of another one of its infamous ten-year cycles of
boom followed by military bust? That would be too hasty a conclusion.
After all, Turkey's 1988 current account was in surplus by US$1.5 billion,
and exports stood at an all-time high. There was no immediate foreign
exchange crisis, although government officials nervously watched foreign
creditors and the black-market premium for the dollar.
But after nine years of adjustment, why does the Turkish recovery look
so fragile? The answer requires dismantling some of the myths that sur-
round post-1980 Turkish economic policies. Unlike Chile, another fre-
quently cited "success" case, Turkey has followed a path that was quite
unorthodox in certain key respects. The chapter by Faezeh Foroutan aptly
discusses many of these unconventional approaches: the emphasis on
public investment early on, despite World Bank advice to the contrary;
the generous use of export subsidies; the substitution of off-budget import
levies for the phased-out quantitative restrictions; and, most critically,
the lack of any sustained fiscal retrenchment. None of these policies sits
well with the idea that Turkey's success derives from liberal, market-
oriented policies. Nor does the exchange rate policy look like the product
of orthodox prescriptions after nine years of pronounced downward
crawl in the real exchange rate-with the expected effect on real wages.
In the only area in which Ozal has acted as a genuine liberalizer, namely
in financial markets, the results have been nothing short of disastrous.'
So the first step in understanding the Turkish outcome is to shed pre-
conceived notions about what an outward orientation has meant in terms
of day-to-day policy. The Turkish state remains by and large as dirigiste
as ever. The terms "liberal" and "market-oriented" are quite misleading
when used in connection with the Turkish reforms of the 1980s. In view
of experience elsewhere, this should come as no surprise: the Republic
of Korea's example amply demonstrates the affinity between outward
orientation and the visible hand of government. But in contrast to the
case in Korea, greater export orientation has not reduced the Turkish
economy's proclivity to rent-seeking; it may have actually increased it,
thanks to the highly arbitrary and discretionary export subsidy and im-
port levy systems. Ozal's great insight was his recognition very early on
that the only way Turkey would regain its creditworthiness was by in-
creasing its export base. And that is, above all, what he accomplished.
The second myth to dispel has to do with fiscal adjustment. According
to accepted wisdom, fiscal retrenchment is essential to successful ad-
justment. How does this square with Turkey's conspicuously undistin-
guished fiscal performance to date? Looking at the borrowing require-



464 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
ments of the public sector, one is struck by the fact that 1981 is the only
year with a moderately respectable deficit-and the revenue figures for
that year are greatly assisted by the one-time realignment in public en-
terprise prices and the positive effects on tax collection of the reduction
in inflation. The fact is that Turkey has so far not paid the fiscal price
of adjustment. In its first few years the Turkish program was underwritten
by substantial inflows from multilateral and official sources (this is well
described by Foroutan). Since 1983 the export boom has allowed Turkey
to turn to private sources of finance: current account deficits have been
financed increasingly from private sources' and by Turkish workers in
Germany. Such inflows did more than provide Turkey with the oppor-
tunity "to spread over time the burden of the necessary adjustment mea-
sures," as Foroutan puts it; they virtually eliminated the need for fiscal
adjustment.
In this light, the success with inflation control and economic growth-
at least until very recently-does not fook so surprising. The stagflation
of the late 1970s was the direct consequence of the shortage of foreign
exchange in the aftermath of the mid-1977 debt crisis: the lack of im-
ported inputs had squeezed domestic activity, while the curtailment of
foreign borrowing had necessitated recourse to inflationary finance.3
When the OECD consortium lifted the foreign exchange constraint in
1979-for reasons related to Turkey's strategic location in the Middle
East and on the eastern flank of NATO-inflation could be lowered and
growth could be restored. Ozal's exchange rate policy and generous ex-
port subsidies, together with the Iran-Iraq war, ensured that the ensuing
spurt in output would be export oriented. This in turn prepared the
ground for Turkey's reentry into private capital markets just as official
flows were beginning to be phased out.
The present fragility of the Turkish economy can therefore be viewed
as the consequence of the long delay in fiscal adjustment. External debt
service now exerts a suffocating pressure on fiscal balances; foreign debt
stands at 60 percent of GDP, twice its level at the start of the adjustment
in early 1980. And now there is internal debt to contend with as well.
Moreover, the prospects for a serious fiscal effort in the near future are
not encouraging. The Ozal government is at present weaker politically
than it has ever been, while the cumulative deterioration in income dis-
tribution leaves it little room to maneuver.
Could the World Bank's lending policies have improved things? Turkey
is a critical test case for the Bank's adjustment lending. Not only has
Turkey received the largest number of structural adjustment loans of any
country to date, but it is also the one important case for which the World
Bank can claim some success. As Foroutan's discussion implies, however,
the World Bank's main contribution to Turkey lay in the commitment
of financial resources rather than in the area of conditionality and policy



Turkey: Structural Transformation and the Threat to Sustainability 465
advice. The conditionality element in the adjustment loans was excep-
tionally weak, and Turkey was left (in the words of Lance Taylor) on a
"long leash." The outline and sequencing of Turkish policies were home-
grown, which partly accounts for the heterodox elements in the program.
Credit must be given here to Ozal's reluctance to undertake a deep cut
in public investment early on and to his delay of substantial import policy
liberalization until 1984, both of which must have gone against the in-
stincts of World Bank staff. When Ozal yielded to foreign pressure, as
with his attempted financial liberalization of 1980, the consequences were
occasionally disastrous.
The World Bank saw itself primarily as a promoter of economic lib-
eralization, or of what the loan documents called "a development path
placing more reliance on market forces." The Turkish economic crisis of
the 1970s was attributed to an inward-oriented strategy rather than to
a simple mismanagement of macroeconomic policy. But, although the
two frequently go hand in hand, it was wrong to think that liberalization
on its own could contribute much to the process of macroeconomic sta-
bilization. As a consequence of the emphasis on liberalization, the purely
macroeconomic aspects of the program received too little attention and
analysis-perhaps, as Foroutan suggests, because the accepted division
of labor made them the responsibility of the International Monetary
Fund. Whatever the reason, the World bank paid insufficient attention
to the fiscal consequences of trade liberalization. (This was later pointed
out in a performance audit report.) The World Bank's scheme had the
balance between stabilization and liberalization skewed too much toward
the latter. This misplaced emphasis on liberalization created other blind
spots as well, a key example being the current privatization drive: it is
hard to see what the sale of state equity at bargain prices accomplishes
other than a worsening of the government's intertemporal budget con-
straint. It is fortunate that the Turkish government, for the most part,
paid little attention to this advice.
Perhaps the most alarming aspect of the Turkish experience is that
private investment in tradables has still not revived after nine years of
"structural adjustment." One need go no further than the highly unstable
macroeconomic and policy environment to understand why. The most
important lesson to be learned from the Turkish case is this: one of the
chief goals of structural adjustment programs must be to achieve overall
stability in the economic environment.4 Accomplishing that requires put-
ting first things first, in particular, putting macroeconomic stabilization
ahead of economic liberalization. Turkish policymakers have wasted too
much of their credibility on unsustainable liberalization attempts and in
the process have engendered too much instability. The price is now being
paid in the form of a dormant private sector.



466 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Notes
1. For an excellent account of efforts to reform the financial market in Turkey,
see the chapters by Yilmaz Akyiiz and Izak Atiyas in Aricanli and Rodrik (1989).
2. In 1988 Turkey raised US$3.4 billion from private capital markets abroad:
US$2.6 billion in bank credits and US$0.8 billion in bond issues (as reported in
Cumburiyet, February 7, 1989, p. 10).
3. For an account of this period, see Celasun and Rodrik (1989).
4. For an extended argument along these lines, see Rodrik (1989).
References
Aricanli, Torsun, and Dani Rodrik, eds. 1989. The Political Economy of Turkey:
Debt, Adjustment, and Sustainability. London: Macmillan.
Celasun, M., and Dani Rodrik. 1989. "Debt, Adjustment and Growth: Turkey."
In Jeffrey Sachs, ed., Developing Country Debt and Economic Performance:
Country Studies. Chicago, Ill.: University of Chicago Press.
Rodrik, Dani. 1989. "Liberalization, Sustainability, and the Design of Structural
Adjustment Programs." John F. Kennedy School of Government Discussion
Paper 177D. Harvard University, Cambridge, Mass.
Comments on Pakistan and Turkey
Hans-Eberhard Kopp
I WOULD LIKE to comment on three themes that have emerged in the
preceding chapters: the political economy of adjustment, the time horizon
of adjustment, and the role of the World Bank in adjustment.
First, as Shahid Husain, in his concluding remarks on Latin America,
has reminded us, the political component is crucial to the adjustment
equation. In this respect, the experiences of both Pakistan and Turkey
in the 1980s have been extremely instructive, and the political and eco-
nomic lessons they teach have an important bearing on the consideration
of adjustment processes. Along with sociopolitical and economic changes,
there has nevertheless been growth throughout-real growth, one hopes,
and not just growth conjured up by statisticians.
Second, adjustment turns out to be a drawn-out process that takes
longer than was first expected. The presentations here have also con-
firmed the difficulty of pinning down just when the adjustment process



Turkey: Structural Transformation and the Threat to Sustainability 467
starts and the perhaps even greater difficulty of knowing when it ends.
Both Pakistan and Turkey, with all the problems that remain or have
reemerged, have shown that adjustment is a gradual process and an al-
most permanent task. Countries make some progress, they move from
here to there, and then they start again. Given the realities of the political
environment, the gradual approach is necessary, at least to some extent.
Third, there remains much ambiguity regarding the role of the World
Bank, and more generally the external financing agencies, in producing
positive results. Despite real difficulties, in both Pakistan and Turkey
tremendous progress has been made. Whether that has been due to the
World Bank or the International Monetary Fund (IMF) is not really the
important issue. Sometimes it is meant as a compliment to say that the
government "owns" a program and sometimes it is meant as a criticism-
suggesting that the IMF and the World Bank did not play a role. We
would all prefer that the government play the guiding role and own the
program. If, under those conditions, the World Bank and the IMF appear
simply to follow along, in the interest of the countries involved, I would
consider that a positive outcome.
Pakistan and Turkey are also good examples of the different ways the
World Bank has been involved in adjustment. In Turkey the World Bank
provided five structural adjustment loans as well as several sectoral ad-
justment loans. In Pakistan it undertook just one structural adjustment
loan and then entered the sectoral adjustment phase. Since Pakistan's
program was built up more from a sectoral base, it is not accidental that
the country is again receiving IMF assistance. The long and continuing
process of adjustment there led to the need to pull things together under
a new IMF program. The question has been raised many times with regard
to Turkey whether the World Bank should have stopped with the five
structural adjustment loans. Five was indeed an arbitrary number. It
needs to be kept in mind that the adjustment process is drawn out, and
the effort needs to be continued.



21 Chile: Economic Crisis and
Recovery
Cristian Moran
DURING THE MID-1970s Chile introduced a series of reforms that radically
changed its economic structure and orientation. Price and interest rate
controls were eliminated, the economy was opened to international trans-
actions, and the import-substitution policies that had prevailed during
the previous three decades were reversed. Free-market-oriented policies
were introduced, and both the importance and the size of the public sector
were greatly diminished. Simultaneously, the country pursued a set of
stabilization policies designed to control inflation (which was running at
1,000 percent in annual terms by the end of 1973) and to correct the
external imbalance, which had widened considerably after the first oil
shock. Despite certain key improvements (such as fiscal and trade reforms
and elimination of price controls), the changes in the economic structure
that were introduced during 1975-80 created a highly vulnerable econ-
omy.'
The Crisis
The fragility of the economy became increasingly evident as the negative
external shocks of the early 1980s unfolded and eventually set off an
enormous crisis. Decisions in four policy areas in particular helped build
a rigid economic structure that was unable to respond to changes in the
external environment. These policies and the financial crisis that followed
are described in this section.
Domestic Economic Policies
The main policies that set the stage for Chile's economic problems were:
(1) the simultaneous adoption of a fixed exchange rate and a wage in-
dexation formula linking nominal wages to past changes in inflation, (2)
a privatization process that facilitated the emergence of highly leveraged
financial conglomerates, (3) financial liberalization with excessively per-
missive banking legislation, and (4) the government's delayed response
to the events preceding the crisis.
468



Chile: Economic Crisis and Recovery 469
EXCHANGE RATE AND WAGE POLICIES. Despite the adoption of con-
servative fiscal and monetary policies during 1975-78, inflation proved
more stubborn than envisioned. After dropping from about 500 percent
in 1973 to 35 percent in 1978, inflation remained at a high level with
no sign of abating in the near future. Because the policy of "preannounced
exchange rate changes" initiated in 1976 had failed to control inflation,
the government decided in mid-1979 to fix the exchange rate in an effort
to set a ceiling rate of inflation and to reverse inflationary expectations.
In that same year, however, a new labor law was also enacted, which
introduced a floor to wage changes by forcing wage adjustments to cover
at least the past changes in inflation.
The wage and exchange rate policies introduced excessive rigidity to
the economy, however. Lags in the adjustment of domestic inflation to
international levels, and a significant increase in capital inflows-
prompted by the financial liberalization-resulted in a drastic appreci-
ation of the peso during 1979-81 (table 21-1). At the same time, declining
terms of trade and increases in international interest rates during 1980-
82 necessitated a real depreciation of the peso in order to limit the increase
in the current account deficit. Given Chile's fixed exchange rate policy
and the relatively low level of international inflation, achieving a real
depreciation of the peso required (under plausible assumptions) a fall in
nominal wages.2 Nominal wages, however, could not fall because of the
wage indexation formula. In 1982 a severe balance of payments crisis,
following a current account deficit of 14.5 percent of GDP in 1981, forced
the government to abandon both the fixed exchange rate and the wage
indexation formula.
THE PRIVATIZATION PROCESS. The privatization program of the mid-
1970s was carried out in two phases. During the first phase-completed
mostly during 1974-75-the firms and agricultural land that had been
seized illegally by the government during the previous administration
were returned to the private sector. During the second phase-covering
1974-80-the government began to divest firms controlled by the State
Development Corporation, CORFO. Most of these firms had required large
capital investment and had received state support because the firms had
been considered important to the national interest.3 In 1974 the military
government started to transfer these firms to the private sector through
a series of public auctions, followed by direct negotiations. This strategy
was flawed, however. The firms were generally purchased with only a
small down payment (typically between 10 and 20 percent of the sale
value), with the rest financed through credit extended by CORFO. The
screening procedures, which were intended to ascertain the strength of
prospective buyers and any cross-ownership, were not enforced. As a
result, this privatization program tended to concentrate assets in indus-



Table 21-1. Basic Macroeconomic Data for Chile, 1978-87
Indices (1978-80 = 100)
Indicator                 1978     1979    1980      1981      1982      1983      1984      1985     1986      1987'    1970       1973     1983     1987
National accounts (annual percentage change at 1977 prices)
GDP                        8.2     8.3    7.8         5.5    -14.1      -0.7       6.3       2,4       5.7      5.4     83.9      85.3      97.0    117.6
GNP per capita             6.3     6.0    5.6         2.2    - 18.1     -2.3       2.3       2.6       4.8      3.8     98.7      96.6      86.3     98.5
Domestic expenditures      9.7    10.5    9.3        11.6    -24.1      -4.6       8.5     -1.9        5.4      7.3     89.8      93.4      88.4    106.4
Private consumption
per capita           5.7     4.7    5.0         8.9    -13.6      -5.0      -0.4     -3.3        3.5      3.5    106.8      114.9     93.6     96.9
Gross fixed
investment          17.4    16.8   21.9        16.8    -33.9    -14.9        9.0      14.8       7.1     15.2    107.2       78.7     78.1    120.5
Exports"                  11.2    14.1   14.3       -9.0       4.7       0.6       6.8       6.9       9.8      6.9     42.8      37.7    109.0    146.0
Importsb                  17.6    22.7   18.7        15.7    -36.3    -15.1       16.5   -11.0         9.7     16.2     73.0      77.3      74.2     98.1
Prices (annual percentage change)
4^     Consumer prices           40.1    33.4   35.1        19.7       9.9      27.3      19.9     30.7      19.5      19.9      0.03      0.3    218.9    491.4
14     Real effective exchange
rate                  -16.6      1.5   16.1       18.0      -9.6    -18.7       -1.7    -19.3    -15.6    -7.4    245.9        167.2     96.0      59.5
Terms of trade           -1.0     11.3    0.6    -15.7        -4.7       8.9    -4.9       -5.7        4.5      4.8    210.0     212.8      91.2     93.4
Real wages                6.4      8.2    8.6        9.0       0.3    -10.9        0.2     -4.5        2.0    -0.2    121.2       94.1    105.3    102.7
Realminimumwage          18.1   -3.1       0.2      -0.8      -2.0    -19.5    -14.6       -5.1      -3.4    -6.1    103.1        75.6     83.4      60.4
Unemployment (percentage of labor force)
Open unemployment         14.1    13.6   10.4        11.3     19.6      14.6      14.0      11.9       8.8      7.9    27.6       33.9    114.9    62.2
Emergency programs         3.9     4.0    5.8        4.8        7.4     13.7       8.3       7.4       5.1      2.7      0.0       0.0   300.0       59.1
Total unemployment        18.0    17.6   16.2        16.1     27.0      28.3      22.3      19.3      13.9     10.6     20.3      24.9    163.9      61.4
a. Preliminary estimates,
b. Goods and nonfactor services.
Sources: National accounts data and terms of trade: Central Bank of Chile.
Consumer prices: Instituto Nacional de Estadistica (INE), period average.
Real effective exchange rate: 1970-76, World Bank data; 1976-87, IMF (trade weighted, period average). An increase in the index reflects an appreciation of the Chilean peso.
Real (minimum) wages = nominal wages (INE), deflated by the revised consumer price index (Arellano 1988b, table 3, p. 8). Minimum wages were obtained as the average of industrial minimum
wages and urban wages and exclude dependent benefits.
Open unemployment rates: INE. Figures for 1978-87 are for the last quarter. The estimate for 1973 is based on the surveys for the Greater Santiago Area conducted by the University of Chile.
Emergency employment programs: Central Bank of Chile. The figures given are those for September of each year.



Chile: Economic Crisis and Recovery 471
trial and financial conglomerates with a weak capital base, which made
them highly vulnerable to financial collapse.4
FINANCIAL LIBERALIZATION. Financial liberalization was also carried
out in the second half of the 1970s. Although the liberalization of do-
mestic financial markets was implemented rapidly and was basically com-
plete by 1977, external financial flows remained restricted until mid-
1979. These restrictions were gradually lifted during 1979-81, first by
eliminating global limits on borrowing, then by restricting monthly in-
flows, and finally by liberalizing short-term flows.
During the financial liberalization process, prudential regulation and
supervision were lax, a problem that was aggravated by the heavy con-
centration of assets in conglomerates. These conditions permitted con-
tinued bank lending to risky projects and to insolvent firms associated
with the same groups that owned the banks. As a result, credit was
channeled to unprofitable but affiliated firms, and less credit was avail-
able for profitable but independent firms (Galvez and Tybout 1985).
Furthermore, inadequate supervision under conditions of excess demand
for credit encouraged an "adverse selection process"; that is, high-risk
borrowers who could temporarily afford the high real interest rates re-
ceived credit while low-risk low-return borrowers were driven out of the
market. In addition, since the government had previously bailed out fail-
ing financial institutions, the presumption was that it would do so again.
Such implicit deposit insurance provided an incentive for undue risk-
taking (Corbo and de Melo 1987).
THE "AUTOMATIC ADJUSTMENT PROCESS." The large real appreciation
of the peso during 1979-81 exacerbated the domestic and external im-
balances. This problem had serious economic consequences because the
government waited several years before initiating the necessary corrective
actions. The first sectors to experience significant declines in economic
activity were agriculture and industry (and with them, domestic trade
and transportation) during the fourth quarter of 1981. At the time, the
government insisted that an "automatic adjustment process" would soon
reverse the excess demand for tradables and that changes in economic
policies were therefore not required. As the decline deepened, however,
spreading to construction, finance, and other services and depleting the
reserves of the Central Bank, the government was forced to devalue the
peso in mid-1982.
With the devaluation, the price of domestic assets in dollar terms fell
drastically, and large capital losses occurred. This aggravated the crisis
and prompted further speculation against the peso. When commercial
bank lending to Latin American countries virtually stopped in the fourth
quarter of 1982, after the crisis in Mexico, Chile was in a deep recession.



472 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Real GDP had fallen 14 percent in 1982 and fell another 1 percent in
1983. Unemployment increased to record levels, reaching 28 percent in
1983 (table 21-1)-including those enrolled in the government's two
emergency employment programs, the Minimum Employment Program
and the Program for Heads of Households.
Despite these signs of economic crisis, the government was still not
convinced that major economic policy adjustments were needed. It did
intervene in the financial sector, however, and partially reversed the trade
policies it had implemented in the 1970s (see below). Not until early
1985, several years after the economy had collapsed, did the government
realize that drastic changes in economic policies were needed.
External Shocks
The negative external shocks Chile experienced during the early 1980s
triggered the crisis and amplified its economic costs. The deterioration
of Chile's terms of trade, in particular, and the increase in international
interest rates imposed a heavy burden on the economy. The terms of
trade deteriorated by an accumulated 20 percent between 1980 and 1982
(table 21-1), and real interest rates increased from 2.3 percent in 1978-
79 to 5.9 percent in 1980-82. The income losses Chile suffered during
1982-83 as a result of these shocks have been estimated at 12.2 percent
of GDP.5
These figures suggest the magnitude of the external shocks, but they
do not explain Chile's economic performance. It is very difficult to sep-
arate the strictly exogenous events from the endogenous policy response.
Thus, to translate the interest rate effect into a GDP IOSS (which contrib-
uted 60 percent of Chile's total loss), the change in international interest
rates during the period analyzed was multiplied by the ratio of debt to
GDP. But Chile had a high debt/GDP ratio before the crisis (48 percent,
compared with 17 percent for Argentina and 19 percent for Uruguay),
which was to a large extent policy induced. Moreover, several studies
analyzing the links between recent external shocks and economic per-
formance in developing countries (see, for example, Balassa and Mc-
Carthy 1984) have not found a significant correlation between the mag-
nitude of the external shocks and indicators of economic performance.
This finding suggests that domestic policy responses, as well as country
characteristics such as level of development, are paramount in explaining
these links-a conclusion that has been confirmed by more detailed anal-
yses of the case of Chile (Balassa 1985; Corbo and de Melo 1987).
The Financial Crisis and Its Immediate Aftermath
To avoid a complete breakdown of the economy, the Central Bank in-
tervened in the five largest private banks in January 1983, liquidated a



Chile: Economic Crisis and Recovery 473
few other financial institutions, and provided massive financial support.
Bank deposits were publicly guaranteed, upper limits were imposed on
passive nominal interest rates, and several emergency credit programs
were established. Two of these programs-one providing subsidized
credit lines to reschedule peso-denominated private debts and one offer-
ing a preferential exchange rate for U.S. dollar-denominated debts-
involved significant increases in Central Bank credit to financial inter-
mediaries. Somewhat later, the Central Bank also purchased the non-
performing portfolios of commercial banks.
In early 1984 the economic authorities reached a debt-rescheduling
agreement with commercial creditors that provided some new funds in
return for a government guarantee of the amounts rescheduled. Chile's
new economic team then launched a significant public investment pro-
gram. GDP increased by 6.3 percent in 1984, and unemployment declined
to 22 percent even though the emergency programs were drastically re-
duced (table 21-1). The recovery was unsustainable over the longer term,
however. The trade surplus dropped considerably, and the current ac-
count deficit widened from 5.7 percent of GDP in 1983 to 10.7 percent
of GDP in 1984. To cope with this imbalance, the government imposed
new limits on imports. A 15 percent tariff surtax on selected import items
was added in March 1984 on top of a 1983 general increase in tariffs
from 10 to 20 percent. Tariffs were finally leveled off at 35 percent at
the end of 1984.
The Adjustment Program
By 1985 the severity of Chile's economic problems had become evident
to everyone, and the authorities realized that a longer-term strategy was
required to address them. The government initiated a three-year adjust-
ment program that emphasized three key areas: export incentives and
the balance of payments, domestic resource mobilization, and rehabili-
tation of the financial and corporate sectors. This program was com-
plemented by an orthodox stabilization package and an explicit attempt
to restructure the external debt.
The government's adjustment program was supported by a three-year
arrangement with the International Monetary Fund (IMF) approved in
August 1985 and totaling 750 million in special drawing rights (SDR)6
and three consecutive World Bank structural adjustment loans (SALs) to-
taling US$750 million, approved in October 1985, 1986, and 1987. Both
financial packages shared the goals of a more efficient public sector, better
budgetary and institutional controls, an open and export-oriented econ-
omy, and a revitalized financial sector.
Besides specific fiscal, monetary, and reserve targets, the IMF program
stressed the phasing out of the preferential exchange rate system for U.S.



474 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
dollar debtors, a lower budget deficit, and a major reduction in Chile's
need for new commercial funds. The SALS emphasized the restructuring
of the corporate and financial sectors and support for the government's
trade, fiscal, and financial reforms.
Macroeconomic Stabilization Policies
After significantly devaluing the peso in September 1984, the government
made a commitment to maintain a real exchange rate that would promote
new (noncopper) exports. This implied adopting a passive crawling peg
system and continuously adjusting the nominal exchange rate according
to the difference between domestic and foreign rates of inflation. In prac-
tice, devaluations exceeded these differences, and the real effective ex-
change rate depreciated substantially.
To complement the exchange rate policy, the government adopted an
austere fiscal and monetary program. The fiscal policy was designed to
reduce the overall nonfinancial public sector deficit by containing fiscal
expenditures-particularly wage expenditures. Limits on the Central
Bank's creation of credit were also established in order to conform to
specific monetary targets set in the context of the IMF program.
Management of External Debt
Two new debt-rescheduling agreements were reached during this period.
In 1985 Chile and its foreign creditors agreed to restructure US$5.7 bil-
lion of maturities due in 1985-87, and the banks agreed to provide
US$1.1 billion of new money for 1985-86. Approval of the first SAL,
partial World Bank guarantee of a US$300 million loan with cofinancing,
and Chile's participation in the IMF program were instrumental in Chile's
obtaining this agreement. In 1987 Chile reached a new agreement with
its commercial creditors to postpone payments for US$12.4 billion of
maturities falling due in 1988-91, leaving an anticipated financing gap
of US$650 million for 1987-88. Two elements negotiated in this last
package provided additional relief: repricing (a reduction in the spreads
on amounts previously rescheduled in 1983, 1984, and 1985) and retim-
ing (a reduction in the frequency of interest payments). These measures
were complemented by agreements with creditor governments to post-
pone repayments of principal during 1985-88.
Despite the substantial short-term relief provided by these renegotiated
agreements, they imposed a heavy burden on the Chilean government.
The government agreed to guarantee all rescheduled amounts, including
previously unguaranteed foreign debt, thereby eliminating the possibility
of major write-offs in case-by-case private negotiations. Through this
process, the government effectively "socialized" the huge private losses



Chile: Economic Crisis and Recovery 475
that could not be serviced, an action whose consequences would become
clear during the process of financial rehabilitation.
External debt restructuring was complemented in 1985 by the intro-
duction of two innovative debt-conversion mechanisms. The first, known
as a buyback scheme, is open to both foreign and domestic investors and
is regulated under chapter 18 of the Central Bank's Rules on International
Exchange. Under the buyback scheme, external debt instruments are con-
verted into domestic debt certificates that can be sold in the domestic
capital market and used in a variety of domestic transactions. The second
mechanism, known as a debt-equity swap, is open only to foreign inves-
tors and is regulated under chapter 19. It permits external debt instru-
ments to be converted into equity investment. In both cases, the con-
versions are made at the official exchange rate. The external debt
certificates are bought at a discount in the international secondary market
but are recognized at near par value in Chile.'
Trade Policy
At the beginning of 1985 the government initiated a series of actions to
improve export incentives. It eliminated a stamp tax on export activities
and the value added tax on investment for export production and reduced
the commercial bank reserves needed for export lending. Import tariffs
were lowered from 35 to 30 percent in March 1985, to 20 percent in
July 1985, and to 15 percent in January 1988. In addition, the peso was
further devalued.
The World Bank's adjustment package emphasized complementary
measures to minimize the disruptive effects of fluctuations in interna-
tional copper prices and to improve incentives for small and medium-
size exporters. A pilot copper price stabilization fund was created in early
1986 to smooth out government revenues and avoid short-term pressure
on the real exchange rate.8 Additional incentives to small exporters in-
cluded temporary fiscal credits in amounts up to 10 percent of the value
(f.o.b.) of exports and the allowance of wider margins than those stip-
ulated in the banking law for personal credit for the promotion of new
exports. Other measures improved export incentives by offsetting the 20
percent uniform tariff on imported capital goods and intermediate inputs
for both direct and indirect exporters, establishing a privately run and
unsubsidized export credit insurance program, and expanding institu-
tional and technical assistance to small and medium-size exporters
through the state's export development agency, PROCHILE, which also en-
couraged foreign direct investment in export-oriented activities.
Resource Mobilization
Much of the initial effort to mobilize savings was concentrated in the
public sector. Additional savings were needed to cover the transitory



476 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
imbalance generated by the privatization of the social security system,
interest on the public debt, and the losses resulting from the financial
crisis. A 1981 reform of the social security system had transformed the
publicly supported system into one based on the private investment of
individual savings. During the transition, the government continued to
pay pensions to those who had already retired and to maintain pension
obligations for those who were about to retire, but it no longer received
contributions from those who had transferred to the private system. As
a result of this reform alone, the overall nonfinancial public sector deficit
increased by an estimated 4 percent of GDP in 1982, which accounted for
most of the deterioration in public sector savings during that year.9 In
addition, the govemment had to absorb part of the losses originating
from private debts that could not be serviced. Since it had already guar-
anteed these debts, the government was forced to take direct responsi-
bility for an increasing proportion of Chile's external debt. To cope with
these needs, specific targets were adopted to increase public savings to
4.5 percent of GDP in 1986,4.7 percent in 1987, and 5.5 percent by 1990,
compared with a surplus of only 0.4 percent in 1984.
Fiscal reforms initiated in 1984 were intended to encourage private
savings and investment. The maximum marginal tax rates were to be
decreased gradually from 56 percent to 50 percent, and income tax brack-
ets were enlarged. Retained earnings were exempted from personal in-
come tax and taxed at a flat rate of 10 percent. Other measures gradually
phased out the surtax on corporate profits and allowed selected invest-
ments to be partially deducted (up to a maximum of 20 percent of the
amounts invested) from the personal income tax base. In a later stage,
the government announced steps to ease restrictions on institutional
investors, including the private pension system, to permit the mobiliza-
tion of these resources into productive investments.
Rehabilitation of the Financial and Corporate Sectors
Government intervention in Chile's financial sector prevented chaotic
bankruptcies, but it led to ownership problems affecting production and
investment in the corporate and financial sectors. To address this issue,
the government created special commissions to manage the firms in which
it had intervened and to define financial programs for their reprivati-
zation. It then pursued the recapitalization of banks and firms through
the provision of heavy subsidies and strengthened the supervisory and
legal framework of the Superintendency of Banks to prevent future crises.
In 1985 the government initiated a program to recapitalize the banks
it had taken over. Their substandard portfolios were purchased by the
Central Bank, which paid for them by issuing four-year promissory notes
with a 7 percent real interest rate. In return, the banks paid a 5 percent



Chile: Economic Crisis and Recovery 477
real interest rate for the outstanding portfolios sold to the Central Bank
and committed themselves to repurchase these portfolios out of future
profits by limiting the distribution of dividends-except in the case of
new shareholders, to whom a total of 30 percent of earnings would be
distributed as dividends. This mechanism not only avoided a massive
expansion of credit and a complete bailout of private shareholders, but
also allowed the banks to spread the losses ovei several years. The Central
Bank notes were then capitalized, and the shares were sold to the public.
The corporate sector was recapitalized through Central Bank-financed
debt reschedulings, which permitted many firms to stretch their financial
servicing costs over a long period. The economic recovery then allowed
viable firms to generate sufficient profits to service their debts while
improving their overall debt-equity ratios. As soon as these firms regained
financial strength and their legal status was settled, the government began
to reprivatize them.
To limit the losses arising from the various subsidy schemes, the gov-
ernment announced that no further generalized reprogramming of loans
would be made after July 1, 1985. It also announced a gradual phasing
out (by early 1987) of the preferential exchange rate for U.S. dollar debt-
ors and reaffirmed its commitment to enforce the repurchase by com-
mercial banks of the substandard portfolios bought by the Central Bank.
Despite these policies, the Central Bank incurred heavy losses.
In 1987 a new banking law established greater accountability in bank-
ing activities and enhanced the supervisory power of the Superintendency
of Banks. The law also provided for the gradual replacement of the state's
general deposit guarantee by a limited guarantee system that spread the
risk among shareholders, bank creditors, depositors, and the state.
Outcome of Adjustment Policies
Chile's adjustment program permitted a resumption of growth and a
continued strengthening of the current account position. It also enabled
the country to improve its external debt situation, increase exports, and
mobilize resources. These gains, however, have had both financial and
social costs.
Macroeconomic Stabilization
GDP growth averaged 5.5 percent a year during 1985-87, led by increases
in exports and investment.10 With strong output growth, unemployment
decreased at a significant pace from 22.3 percent in 1984 to 10.6 percent
in 1987, its lowest level since 1974. The current account deficit fell from
US$2.1 billion in 1984 (10.7 percent of GDP) to US$1.0 billion in 1987



478 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
(4.8 percent of GDP) because of significant increases in exports, reductions
in international interest rates, and in 1987 improvements in copper prices.
Most of the targets set in the adjustment program were met. Targets
for international reserves, for net domestic assets of the Central Bank,
and for short-term indebtedness of the public sector at the end of 1987
were even exceeded by wide margins. The public sector deficit also nar-
rowed more rapidly than anticipated, dropping from 4.4 percent of GDP
in 1984 to 0.8 percent in 1987 (compared with the target of 1.6 percent).
Inflation, however, was marginally above its target in 1987. After in-
creasing to 30.7 percent in 1985, inflation dropped to 19.5 percent in
1986 and then increased slightly to 19.9 percent in 1987 (compared with
the target rate of 18.1 percent).
Despite these achievements, concerns about the social impact of the
adjustment program remain (see discussion below). Average per capita
income has not yet reached its previous level (table 21-1), while con-
sumption per capita in 1987 was still 5 percent below the average level
for 1978-80 and 12 percent below the level in 1970.1 Although un-
employment in 1987 was below the average in 1978-80, it was still more
than twice the level that prevailed during 1965-73.
External Debt
By following a consistent and austere adjustment program that stressed
improvement in the balance of payments, Chile was able to reschedule
its external debt successfully. These efforts were complemented by re-
ductions in the spreads on new and rescheduled debt and, more impor-
tant, by significant declines in international interest rates in both nominal
and real terms.
The total level of external debt increased only slightly (from US$19.7
billion in 1984 to US$20.8 billion in 1987, although it declined relative
to exports) as the new borrowing was partly compensated for by the
external debt-reduction schemes introduced in 1985 (table 21-2). These
schemes proved useful in limiting the growth of external debt, but their
benefits were limited.
At the end of 1987 Chile had retired US$2.2 billion in external debt
through these mechanisms.'2 Slightly more than half this amount (US$1.2
billion) was retired through buybacks under chapter 18, which do not
carry remittance rights. The chief benefits under this scheme are that
Chile receives not only a reduction in future interest payments (in this
case, worth about US$110 million) but also the part of the external
discount that is appropriated by the Central Bank (through auctions for
the right to participate in these operations). This discount averaged 15
percent during 1985-87 (or about half the discount rate prevailing in
international secondary markets), for which Chile received US$190 mil-



Table 21-2. Chile's External Debt, 1970, 1973, and 1978-87
Indicator                     1970       1973      1978       1979      1980       1981      1982       1983      1984       1985       1986      1987'
Total external debt (billions
of U.S. dollars)            n.a.       n.a.     7.011      8.663    11.207    15.591    17.159    18.037    19.666    20.404    20.670    20.761
Long-term debt              2.767      3.261      5.923      7.507     9.413    12.553    13.815    14.832      16.963     17.650     17.814    17.356
Public and publicly
guaranteed              2.218      2.862      4.353      4.771     4.720      4.415     5.157      6.689    10.601     12.515     14.379    14.729
Private nonguaranteed     0.549      0.399      1.570      2.736     4.693      8.138     8.658      8.143     6.362      5.135      3.435     2.627
Short-term debt             n.a.       n.a.     0.741      0.977     1.671     2.989      3.338      2,599     1.914      1.668      1.574     1.945
IMF net credit            0.002      0.095      0.347      0.179     0.123      0.049     0.006      0.606     0.743      1.085      1.328     1.460
Debt ratios (percent)
.s.    Private nonguaranteed
debt/long-term debt        19.8       12.2       26.5       36.5      49.9       64.8      62.7       54.9      37.5       31.1       27.2       15.1
Total external debt/GDP       n.a.       n.a.      45.5       41.8      40.6      47.8       70.5       91.3     102.5      127.6      122.9     110.8
Total external debt/exports  217.2b    228.7b     235.0      183.5     178.6      277.7     332.9     373.4      408.5      437.0     404.0      343.3
Debt service ratio'          24.4b      16.1b      49.0       42.3      38.8       61.8      62.9       56.5      62.1       49.7       44.4      42.8
Average interest rated          6.8        5.8       10.6       13.0      13.8       15.0      13.2       11.8      12.5        9.5        8.6       n.a.
LIBORe                          8.5        9.2        9.1       11.9      14.0       16.7      13.6        9.9      11.3        8.6        6.8       7.2
Real LIBOR                      2.8        2.4        1.7        2.9       4.5       6.5        6.7        5.9       6.9        5.1        4.0       3.2
n.a. Not available.
a. Preliminary estimates.
b. Debt ratios for 1970 and 1973 exclude short-term debt and interest payments on short-term debt.
c. Total debt service payments as a percentage of exports of goods and services.
d. Average cost of new commitments on public and publicly guaranteed long-term debt, obtained from the World Bank's Debtor Reporting System.
e. London interbank offer rate (LIBOR) on six-month U.S. dollar deposits.
f. LIBOR deflated by the U.S. GNP deflator.
Source: Central Bank of Chile.



480 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
lion (Ffrench-Davis 1987). The remaining portion of the external debt
retired during 1985-87 (US$1.0 billion) was processed as debt-equity
swaps through chapter 19. The benefits Chile obtained under this scheme
may have been even smaller than under the buyback scheme. Under debt-
equity swaps, the reduction in interest payments may be neutralized by
increases in future profit remittances, and the discounts do not accrue
to the Central Bank but are divided between foreign investors-who
effectively receive a preferential exchange rate-and commercial banks
and other intermediaries.
Three other comments need to be made about the mechanisms for
reducing external debt. First, from an accounting point of view, a debt-
equity swap has no effect on a country's net liability position with the
rest of the world when the debt is redeemed at face value and no new
investment requirements are introduced at the time of the swap operation,
as is the case in Chile. However, since debt ratios include repayment
obligations on loans but exclude interest and dividends on foreign in-
vestment, debt ratios are reduced as a consequence of the swap operation.
Second, debt-equity swaps can also have significant monetary and fiscal
implications because they often increase the domestic money supply and
may increase the domestic-currency debt service obligations of the gov-
ernment (when domestic interest rates are higher than the interest rate
on the foreign loans). There is little evidence, however, that these negative
effects have been important in the case of Chile. Third, although the
debt-equity schemes in Chile increased the flow of foreign funds-most
of which were used to reprivatize firms (see below)-this increase was
partly reversed by a reduction in net foreign direct investment (from an
average of US$220 million a year in 1980-84 to US$60 million in 1985-
86).
External Trade
Exports benefited greatly from the improved incentives. The volume of
merchandise exports increased 7.9 percent a year during 1984-87, com-
pared with 3.7 percent during 1980-84. Agricultural and industrial ex-
ports increased even more (by 19 and 15 percent a year during 1984-
86, respectively, compared with 8.4 and 4.2 percent a year during 1980-
84), helped by a favorable exchange rate and other incentives that ef-
fectively redressed the mild antiexport bias prevailing during 1974-81.
Although no direct evidence is available, there are indications that in-
vestments in export-oriented activities-particularly in fruit, forestry,
agroindustry, and mining activities-increased considerably in the late
1980s.
Imports also increased substantially during 1985-87, despite the sig-
nificant depreciation of the real effective exchange rate. Fueled by the



Chile: Economic Crisis and Recovery 481
economic recovery, increases in investment, and the decline in tariff rates,
the volume of imports rose by 15 percent a year during 1985-87. Because
of the relatively low import base (after a significant decline in 1985),
however, the trade balance improved from 1.5 percent of GDP in 1984
to 5.8 percent in 1987, despite unchanged terms of trade (table 21-3).
Resource Mobilization
Public sector savings grew significantly, from 0.6 percent of GDP in 1984
to 5.3 percent in 1987, surpassing the 4.7 percent target set in the ad-
justment program. This performance was due to both increases in revenue
and reductions in expenditure. Improvements in the administration of
the value added tax, in particular, increased the collection of indirect
taxes from an average of 15.0 percent of GDP during 1981-84 to 16.9
percent during 1985-87 (table 21-4). These gains, however, were partly
offset by the decline in direct tax collections (from 5.3 percent of GDP in
1981-84 to 3.8 percent in 1985-87) that resulted from the 1984 tax
reform-a process that intensified the changes introduced in the previous
fiscal reform and increased its regressive distributional consequences.13
Other changes that complemented the government's savings efforts in-
cluded an increased surplus in the operation of public enterprises, lower
expenditures because of reductions in the emergency employment pro-
grams and international interest rates, and continued wage restraint in
the public sector.
Private savings also increased from 2.3 percent of GDP in 1984 to 6.8
percent in 1987 but with a significant drop in 1985 (table 21-3). Thus
despite significant increases in incentives, private savings averaged only
3.7 percent of GDP in 1985-87 compared with 3.1 percent in 1980-84.
Increased savings (due in part to a cyclical recovery and in part to the
adjustment program) and declines in domestic interest rates permitted a
recovery of investment. Gross fixed investment increased by 11 percent
a year in real terms during 1985-87. As a percentage of GDP, fixed in-
vestment increased from 12.4 percent in 1984 to 15.5 percent in 1987,
and public investment increased from 6.4 percent to 7.3 percent. The
value of these achievements was enhanced because the government re-
stricted investment to the most profitable projects and dropped or scaled
down nonessential investments.
Rehabilitation of the Financial and Corporate Sectors
The Central Bank's intervention in failing banks and businesses and the
various subsidy schemes it introduced allowed many banks and firms to
continue operating. It therefore prevented widespread bankruptcies and
their devastating economic consequences, but its losses were enormous.



Table 21-3. Balance of Payments, Savings, and Investment in Chile, 1970, 1973, and 1978-87
(percentage of GDP based on current prices)
Item                        1970     1973      1978    1979    1980        1981     1982    1983       1984       1985      1986      1987'
Balance of payments
Trade balance                 4.7     -0.3    -2.8    -1.7    -2.8         -8.2       0.3      5.0        1.5       5.3        6.5      5.8
Current account balance     - 1.4     -4.2    -7.1    -5.7    - 7.1    -14.5    -9.5    -5.7           - 10.7     - 8.3      -6.5    -4.8
Savings and investment
External savingsb           -0.7        1.9      3.3      2.8      4.2      10.3      1.9    -2.7         1.1     -2.8       -3.8    -3.0
b   Gross domestic savings       17.1       6.1     14.5    15.0      16.8       12.4      9.4     12.5      12.6       16.5      18.4     19.9
,   Net factor income            -2.2      - 1.1    -2.2    -3.0    -3.4        -4.5    -4.5      - 7.8     - 10.2    -11.9      -11.2    - 8.2
Net current transfers        0.3        0.2      0.5      0.4      0.4       0.3      0.5      0.5        0.5       0.4        0.4      0.4
Gross national savings      15.2        5.2     12.8    12.4      13.9       8.2      2.1      4.4       2.9        5.0        7.7     12.1
Private savings            7.9       25.0      4.0      3.4      3.3       2.7      2.9      4.2       2.3        1.2        3.1      6.8
Public savings             7.3    -19.8        8.8      9.0     10.6       5.5    - 0.8      0.2       0.6        3.8        4.6      5.3
Gross domestic investment   16.4        7.9     17.8    17.8      21.0      22.7    11.3       9.8      13.6       13.7       14.6     16.9
Gross fixed investment    15.0       12.8     14.7    14.9      16.7      18.6    14.6      12.0      12.4       14.2      14.6      15.5
Change in stocks            1.4     -4.9       3.1      2.9      4.3       4.1    -3.3    -2.2          1.2     -0.5         0.0      1.4
a. Preliminary estimates.
b. Net imports of goods and nonfactor services.
Sources: Central Bank of Chile and World Bank estimates.



Table 21-4. Public Sector Finances in Chile, 1970, 1973, and 1978-87
(percentage of GDP based on current prices)
Item                          1970        1973       1978       1979       1980      1981       1982       1983        1984        1985         1986        1987'
Nonfinancial public sector
Current revenue              38.2        21.3      36.7       33.8       35.0       31.8       31.4       30.5        31.2         33.4        32.0        32.6
Direct taxes                7.7         5.7        &1        72        7.3        5.8        5.7        5.0         4.6          3.7         3.7         4.1
Indirect taxes,            10.8        10.1      13.7       13.3       13.4       14.8       13.8       14.6        16.6         17.0        17.0        16.8
Other revenue              19.7         5.5       14.9      13.3       14.3       11.2       11.9       10.9        10.0         12.7        11.3        11.7
Current expenditure          30.9        41.1      27.9       24.8       24.4       26.3       32.2       30.3        30.6         29.6        27.4        27.3
Current surplbs               7.3      -19.8         8.8       9.0       10.6        S.5      -0.8         0.2          0.6         3.8         4.6         5.3
0c     Capital formation             10.4         8X4       6.7        5.2        S.4       5.0         4.7        4.7         6.4          7.0         7.5         7.3
Overall surplus             -6.7       -30.5         1.S       4A8        5.5        0.8      -34        -2.8         -4.4        -2.6         -1.9       -0.8
Central Bank (net income)b      n.a.        n.a.      n.a.       n.a.       1.0        1.5        0.5      -6.4         -6.2       -12.6         -9.5         n.a.
Consolidated public sector'     n.a.        n.a.      n.a.       n.a.       6.5        2.3      -2.9       -9.2       -10.6        -15.2       -11.4          n.a.
n.a. Not available.
a. Preliminary estimates.
b. These figures were obtained from the Central Bank's statements of income and expenses. Since 1982 they have included the subsidies given to commercial banks (to finance losses originating
from the purchase of bad loans) and to individuals and corporations (to finance the restructuring of domestic debts and the preferential exchange rate system). Price adjustments that are due to
indexed operations are omitted.
c. Includes Central Bank and nonfinancial public sector.
Sources: Ministry of Finance, Central Bank of Chile, and World Bank estimates.



484 COUNTRY EXPERIENCE WITH ADJUSTMEN-T LENDING IN THE 1980s
These losses are estimated to have averaged over 6 percent of GDP during
1983-84 and 11 percent of GDP during 1985-86-for an accumulated
total of US$6.0 billion (table 21-4).? Moreover, these estimates omit the
losses incurred by the State Bank (Banco del Estado), which absorbed
the debts of two failing banks valued at US$360 million (Larrain 1988).
The losses for 1987 and beyond are expected to be substantially lower
than in previous years, however, since most of the subsidy schemes have
been phased out, but the extent of the losses will depend to a significant
extent on continued economic recovery.
The reprivatization of these banks and firms began shortly after they
had been recapitalized. In this new privatization effort-which was
mostly completed during 1986-the government deliberately distributed
ownership among a large number of proprietors and ensured more careful
screening of prospective buyers. Several methods were used to distribute
ownership, but the one that attracted the most attention was the system
of "popular capitalism." It imposed limits on the maximum stock own-
ership of individual agents and gave credits to small investors to enable
them to participate. But this system also provided tax breaks favoring
wealthy individuals and therefore involved heavy subsidies with regres-
sive distributional consequences.
In a later phase, beginning in 1985 and still ongoing, the government
continued its attempts to privatize the remaining state enterprises, a pro-
cess that had been interrupted during the 1982-83 crisis. As of mid-
1986, forty state enterprises remained, of which twenty-five were in the
hands of CORFO. Three main forms of divestiture have been used: sales
in the stock exchange; sales to institutional investors, dominated by pri-
vate social security corporations; and sales to the workers of the com-
panies being privatized.
The program has raised several issues. First, the government has not
developed a phased and carefully planned program but has chosen instead
to privatize these firms in a very short time. Second, the government has
not clarified the regulatory framework for the privatized companies
(some of which are natural monopolies) or the final role that institutional
investors will be allowed to play. Both these factors may have depressed
the prices offered for the companies. Third, and perhaps most important,
the government has not announced how it will use the proceeds from
the sales of these firms. If they are used to finance current expenditures
rather than investment, particularly in infrastructure and human capital,
the net effect on government finances will likely be negative (Larrain
1988).
The Social Costs of Adjustment
Despite significant economic improvements in the late 1980s, there are
indications that the social costs of adjustment may have fallen dispro-



Chile: Economic Crisis and Recovery 485
portionately on those at the lower end of the income scale. Two factors
in particular suggest this outcome. First, the rewards to domestic factors
of production benefited capital against labor. Profit rates increased sig-
nificantly during 1980-87 while real wages remained depressed. Average
wages declined by 5 percent in real terms during 1980-87 and by 13
percent during 1982-87 (table 21-1). Second, changes in wage distri-
bution adversely affected low-paying jobs. For example, the minimum
wage declined by 44 percent in real terms during 1981-87. Additional
indicators suggest that wages for unskilled workers declined much more
than did other wages: wages for domestic employees and construction
workers declined by 25 and 33 percent, respectively, during 1981-87,
compared with declines of 12 and 13 percent, respectively, for wages in
the financial and modern sectors of the economy (Arellano 1988a, table
6).
It is likely that the behavior of real wages, along with the economic
recovery, encouraged the substantial increase in employment. But real
income for low-paying jobs declined substantially. For example, real in-
come for domestic employees and construction workers declined by mar-
gins ranging from 3 to 50 percent during 1981-87, while average real
income for the economy as a whole (average wages times employment)
increased by 3 percent during the same period. Real income for those
earning minimum wages also declined markedly.
The apparent deterioration in income distribution is consistent with
explicit policy choices adopted by the government. The tax policy, in
particular, became more regressive as indirect taxes increased as a pro-
portion of total tax revenue from 58 percent in 1970 to 72 percent in
1980-84 and 82 percent in 1985-87. By comparison, indirect taxes as
a proportion of total tax revenue averaged 61 percent for a sample of
developing countries with income levels similar to Chile's (Tanzi 1987).
This bias was partly reversed in mid-1988 when the government reduced
the uniform rate on the value added tax from 20 to 16 percent. Other
biases, however, resulted from the financial rehabilitation and repriva-
tization (which favored U.S. dollar debtors and wealthy individuals) and
may also have contributed to the deterioration of income distribution.
The deterioration of real incomes for low-paying jobs has not been
accompanied by a deterioration in traditional indicators of mortality and
nutrition, however. Improvements in the efficiency of government pro-
grams and better targeting of social programs for poor and high-risk
groups have resulted in a continued decline in these indicators, although
at a slower pace than before (Cornia, Jolly, and Stewart 1987). None-
theless, a deterioration in health infrastructure and an increase in the
incidence of poverty-related health problems followed the significant de-
cline in health expenditures during 1981-86.



486 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Recently, the government increased expenditures in the social sectors,
particularly in health. A structural adjustment loan supports a program
to maintain the present per capita level of food distribution in maternal
and child health programs and to expand preventive care services such
as health education, screening, and community education in order to
decrease the incidence of chronic illnesses among adults. Other plans call
for improvements in health infrastructure, medical equipment, and the
administration of the public health system and reinforcement of the pri-
vate health insurance scheme.
Evaluation of Loan Conditionality
All three SALs were well designed and clearly focused on the main prob-
lems facing Chile after the 1982-83 crisis. Appropriately, the emphasis
of the loans shifted gradually from the more urgent needs requiring swift
and strong stabilization efforts to longer-term, more specific structural
reforms. SAL I focused on the balance of payments (export incentives),
savings mobilization, and an orderly phasing-out of generalized subsidies
such as the preferential exchange rate and interest rate subsidies. SAL 11
focused on the rehabilitation of the financial and corporate sectors and
included measures to strengthen public sector finances and the control
of the Superintendency of Banks over financial institutions. SAL 11 stressed
social sector policies, particularly in health, and improvements in the
mobilization and allocation of long-term savings. Conditions in SAL 1II
had none of the vagueness that sometimes afflicted SALS I and ii (for
example, when they merely called for a review of progress in the
bank recapitalization program or "satisfactory progress" on particular
studies).
All the loans had two tranches, and a significant proportion of the
loan conditions (particularly in SALS I and II) had to be met before release
of the second tranche. This allowed the World Bank to maintain a close
dialogue with the government and to monitor advances in key areas.
The implementation of loan conditions was generally quite successful,
although a few areas of concern remain. Policy implementation in export
development and public finance, in particular, was very successful. In-
deed, the government exceeded the conditions it had agreed to in these
areas. The implementation of financial sector reforms was also compre-
hensive and successful, although some problems remain (particularly con-
cerning the new banking law and the ability of commercial banks to
repurchase the portfolios sold to the Central Bank). In one instance the
government seems to have backed away from its commitment to end
generalized subsidies; in early 1988 it provided a new subsidy to allow
banks to reschedule housing loans.



Chile: Economic Crisis and Recovery 487
hnplications for World Bank Policy
Chile's adjustment program has clearly been successful in meeting its
objectives. The success is even more apparent when compared with the
experience of other countries that have faced severe debt-servicing prob-
lems in recent years. Concerns about the social costs of adjustment re-
main, however.
There are four main explanations for the success of the adjustment
program in Chile: (1) the introduction of a sound macroeconomic pro-
gram, (2) the elimination of major microeconomic distortions before the
establishment of the program, (3) the strength of the government's com-
mitment to policy change, and (4) the provision of adequate external
financial assistance.
After the crisis of 1982-83 and a few years of indecision, the govern-
ment in early 1985 instituted a comprehensive program for macroeco-
nomic reform that eliminated inconsistencies between exchange rate and
wage policies and strengthened export development and regulation of
the financial sector. Because microeconomic distortions (ad hoc subsidies
and price controls) had been eliminated and significant steps toward trade
liberalization and fiscal reform had already been taken in the mid-1970s,
the economy was able to respond to the new incentives. The government's
commitment to policy change and its firm control of the economy were
also important to the success of the program. Its long tenure and au-
thoritarian character allowed the military government to pursue strong
actions and implement policy changes with relative ease, although the
delay before any action was taken after the crisis may have been unne-
cessarily long. The final significant factor in the success was the external
financial assistance from multilateral institutions (the IMF, World Bank,
and Inter-American Development Bank) that provided the government
with the breathing space it needed to pursue its adjustment program and
to reschedule its external debt successfully.
Despite the success of the adjustment program, however, the economy
remains vulnerable. Chile's external debt is still substantial, and negative
external shocks, such as increases in international interest rates, declining
terms of trade, or increases in protectionism among its trading partners,
could complicate the management of its external and financial debt. More
important, Chile still needs to mobilize sufficient domestic savings to pay
its external debt and increase investment (which is still low by interna-
tional standards), an effort that will be difficult to sustain in the face of
strong pressures to increase consumption and expenditure in the social
sectors.
Notes
I am grateful to Jaime de Melo, Luis Ernesto Derbez, Richard Newfarmer, Jose
Olivares, and other World Bank colleagues for helpful comments on an early



488 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
draft of this chapter, and to Sheila Fallon for typing the manuscript. All remaining
errors and shortcomings are mine.
1. See Corbo (1985), Edwards and Cox-Edwards (1987), and Moran (1989)
for a detailed analysis of these policy changes.
2. The real exchange rate (er) can be expressed as e' = pIpn = ep*lp,, where
e is the nominal exchange rate, pt(pn) is the price of tradables (nontradables),
and p* is an index of international inflation. But p. is determined by domestic
wages (w) and the user cost of capital. Assuming the latter remains unchanged,
changes in the real exchange rate can be expressed as e' = e + fi# - w. If e =
O (fixed nominal exchange rate) and fi't  0 (small level of international inflation),
then a real depreciation can only be obtained by a fall in nominal wages.
3. CORFO was directly involved in about 50 enterprises in 1970, but by 1973
it had increased its participation to 180 firms.
4. See Marshall and Montt (1987), Larrain (1988), and Moran (1989) for a
more detailed discussion of the privatization process in Chile.
5. Compare this figure with losses of 6.7 percent for Argentina during its crisis
of 1981-82, for example, or losses of 1.0 percent for Uruguay during 1982-83
(Corbo and de Melo 1987).
6. IMF funds are denominated in SDR, which are based on a basket of currencies.
7. The external discounts fluctuated between 30 and 35 percent from 1985
to mid-1987 but increased to 40 to 45 percent in the second half of 1987 after
a significant increase in the supply of these certificates.
8. The fund proved useful and timely and was activated in 1987 as copper
prices increased significantly (mostly because of low international stocks and
continued demand pressure) to a level above their long-run trend.
9. The cost of this reform increased to 5 percent of GDP in subsequent years
and was expected to remain at this level until 1990 and then start decreasing
gradually.
10. Preliminary evidence indicates that the economy continued to grow at a
significant pace in 1988 as well, so the trend for 1985-87-which is the focus
of the chapter-would remain unaffected if 1988 were included (although the
levels of particular variables would obviously be affected).
11. The comparison of aggregate consumption levels between 1970 and 1987
may be misleading, however, because of drastic changes in the composition of
consumption (Moran 1989).
12. These figures include only the amounts entered under chapters 18 (buy-
backs) and 19 (debt-equity swaps) of the Central Bank's Rules on International
Exchange. They exclude US$0.2 billion entered under Decree Law 600, which
regulates foreign direct investment, and other operations (such as partial write-
offs or direct buybacks of private unguaranteed foreign debt) that accounted for
about US$0.8 billion.
13. Since lower-income families consume a higher proportion of their income
than wealthier families, a uniform indirect tax rate (such as the value added tax
in Chile) implies that lower-income families spend a higher proportion of their
income on taxes than do wealthier families.
14. 1 am grateful to Edgardo Barandiaran of the World Bank for providing
these figures.
References
Arellano, J. P. 1988a. "Crisis y Recuperaci6n en Chile en los Afios 80." Colecci6n
Estudios CIEPLAN 24:63-84.



Chile: Economic Crisis and Recovery 489
1988b. "La Situaci6n Social en Chile." Notas T&nicas 94. Santiago:
CIEPLAN.
Balassa, Bela. 1985. "Policy Experiments in Chile, 1973-83." In G. M. Walton,
ed., The National Economic Policies of Chile. Greenwich, Conn.: JAI Press.
Balassa, Bela, and F. Desmond McCarthy. 1984. Adjustment Policies in Devel-
oping Countries, 1979-1983. World Bank Staff Working Paper 675. Wash-
ington, D.C.
Corbo, Vittorio. 1985. "Reforms and Macroeconomic Adjustments in Chile dur-
ing 1974-83." World Development 13:893-916.
Corbo, Vittorio, and Jaime de Melo. 1987. "Lessons from the Southern Cone
Policy Reforms." World Bank Research Observer 2 (2):111-42.
Cornia, G. A., Richard Jolly, and Frances Stewart. 1987. Adjustment with a
Human Face. Oxford, U.K.: Clarendon Press for UNICEF.
Edwards, Sebastian, and A. Cox-Edwards. 1987. Monetarism and Liberalization:
the Chilean Experiment. Cambridge, Mass.: Ballinger.
Ffrench-Davis, R. 1987. "Conversi6n de Pagares de la Deuda Externa en Chile."
Colecci6n Estudios CIEPLAN 22:41-62.
Galvez,J., and James Tybout. 1985. "Microeconomic Adjustment in Chile during
1977-81: The Importance of Being a Grupo." World Development 13:969-
94.
Larrain, F. 1988. "Public Sector Behavior in a Highly Indebted Country: The
Contrasting Chilean Experience, 1970-85." Country Policy Department,
World Bank, Washington, D.C.
Marshall, J., and F. Montt. 1987. "Privatization in Chile." Country Policy De-
partment, World Bank, Washington, D.C.
Moran, Cristian. 1989. "Economic Stabilization and Structural Transformation:
Lessons from the Chilean Experience, 1973-87." World Development (April).
Tanzi, Vito. 1987. "Quantitative Characteristics of the Tax System of Developing
Countries." In David Newbery and Nicholas Stem, eds., Theory of Taxation
for Developing Countries. New York: Oxford University Press.
Comments
Sebastian Edwards
THE CHILEAN economic experience since the early 1970s has attracted
considerable attention. In the mid- to late 1970s the daring free market
reforms implemented by the so-called Chicago boys captured the imag-



490 COUNTRY EXPERIENCE WITH ADJUSTM1NT LENDING IN THE 1980s
ination of the financial media and conservative publicists. When the econ-
omy collapsed in 1982-GDP declined by almost 15 percent and unem-
ployment soared to almost 30 percent-many observers wrote off the
experiment, claiming that a free-market-oriented regime could not suc-
ceed in a small developing nation. When the economy recovered with
great vigor in the late 1980s, most experts were surprised. It is not an
exaggeration to say that Chile has become a showcase for those interested
in understanding and evaluating the economics of structural reform and
liberalization.
Cristian Moran's chapter provides an interesting analysis and a sub-
stantial amount of information on the evolution of the Chilean economy
during the 1980s. Although I disagree with several parts of the analysis,
I think this chapter is a timely and useful contribution to the literature
on the Chilean experience with structural reform. My comments cover
five topics: monetary and fiscal policies, the key role played by exchange
rate policy, the time horizon of structural reform, the role of the World
Bank in facilitating the process of structural adjustment, and prospects
and directions for the future.
One of the most important lessons of the Chilean experience has to
do with the influence of macroeconomic stability and macroeconomic
management on successful performance, Indeed, macroeconomic man-
agement was at the center of both the unleashing of the crisis in 1982
and the recovery in the late 1980s. Moran mentions four main internal
causes of the 1982 collapse. Two of them-the simultaneous adoption
of a fixed nominal exchange rate and a wage rate indexation rule and
the reliance on "automatic adjustment"-are obviously macroeconomic
in nature and clearly illustrate how policy mistakes at the level of ma-
croeconomic management can magnify negative external shocks. The
other two policy errors identified by Moran refer to the lack of super-
vision that characterized the privatization and financial liberalization
programs. He does not argue, however, that the whole concept of mi-
croeconomic reform was a mistake or inappropriate for Chile. In that
regard, Moran joins a growing number of scholars who have argued that
the Chilean economic problems of the early 1980s were rooted in external
shocks, macroeconomic mismanagement, and lack of supervision in im-
plementing some of the key microeconomic reforms.
In the same way that macroeconomic policy helped unleash the 1982
crisis, strong and aggressive macroeconomic management was funda-
mental to the recovery of the 1980s. After 1985 fiscal policy was based
on redirecting public expenditure from current expenditure to public in-
vestment. As a result, public investment as a share of GDP increased by
more than 7 percentage points between 1985 and 1989. In addition, fiscal
policy during this period was directed toward encouraging private savings
via a reduction in taxes. What was crucially important, however, was



Chile: Economic Crisis and Recovery 491
that overall fiscal balance was maintained during this period so that
outbursts of inflation and unsustainable pressures on the real exchange
rate were avoided.
The monetary policy of the post-1985 period was characterized by
active manipulation of the stock of domestic credit, mainly for the pur-
pose of controlling interest rates. Indeed, an overriding concern of the
authorities during this period was to avoid repeating the syndrome of
high real interest rates of the late 1970s. In this regard, the contrast
between the attitude of finance minister Hernan Buchi's team and that
of the "Chicago boys" of the 1970s is important. Although Buchi's team
had a clearly pragmatic view of the role of active macroeconomic-
especially monetary-policy in achieving stability, the 1970s team firmly
believed that the only way to tackle macroeconomic problems was to let
the economy adjust on its own. A number of observers have argued that
the automatic adjustment approach to macroeconomic policy that was
followed between 1979 and 1982 was one of the most important mistakes
of the time in Chile.
Of the economic policies pursued by Chile in the post-1982 period,
there is little doubt that the exchange rate policy based on periodic de-
valuations was one of the most important. Between 1982 and 1988 the
real exchange rate was devalued by approximately 90 percent. This de-
valuation not only allowed the country to improve its level of compet-
itiveness and thus greatly helped the boom in nontraditional exports, but
it also contributed to the maintenance of reasonable interest rate levels
and the avoidance of capital flight.
Chile's successful experience with exchange rate management during
this period generated a number of important lessons for other countries.
Perhaps the most important is the demonstration that by devaluing the
nominal exchange rate it is possible to engineer substantial real deval-
uations without generating very high inflation. In Chile the key to the
success of this policy was maintaining fiscal discipline and recognizing
that changes in real exchange rate "fundamentals" are important in de-
termining the extent to which the real exchange rate can be altered.
("Fundamentals" are the immediate determinants of the equilibrium real
exchange rate, such as international prices, import tariffs, export taxes,
real interest rates, capital controls, and so on.) In that regard, the Chilean
authorities followed a particularly pragmatic approach in which the pace
at which the nominal exchange rate was devalued was determined by
several considerations, including the classification of shocks as perma-
nent or. transitory, the evolution of terms of trade, and the availability
of foreign funds.
It must be emphasized that much of the success observed in Chile
during 1987-89 is rooted in the major structural reforms implemented
in the mid-1970s. For example, the boom in nontraditional exports oc-



492 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
curred as a result of investments that were undertaken almost ten years
earlier. Moreover, the flexible and rapid market response to incentives
is a direct consequence of the microeconomic reforms of the 1970s.
Chile's experience shows that it can take a considerable amount of time-
even eight to ten years-for an economy to adapt fully to the new rules
of the game and to new incentive structures. The relatively short duration
of World Bank adjustment programs may constitute a serious limitation
to the achievement of successful structural reform in some cases. Al-
though Moran does not make this point, there is no doubt that in the
case of Chile most of the structural reforms were undertaken long before
the World Bank got involved through its structural adjustment loans.
Moran's chapter provides information on the most important details
of the World Bank's structural adjustment loans to Chile, including their
amounts and the changing emphasis of their conditionality, but it does
not discuss the nature of the negotiations between the World Bank and
the Chilean authorities. This is of relevance to the general question of
structural reform for those who would like to draw lessons from what
appears to be one of the most-if not the most-successful World Bank
operation of the 1980s. Were the negotiations difficult? Did the Chilean
authorities take a strategic stance? Was it difficult to persuade them to
accept some of the more controversial aspects of the loans, or were they
willing to accept all conditions? Did the political situation underlying the
Chilean military regime simplify acceptance of loan conditions by the
authorities? The answers to these and other questions are important,
especially at a time when the effectiveness of World Bank (and IMF) con-
ditionality has come under increasing scrutiny.
Moran does discuss, however, some specific details of the adjustment
loan programs. I was surprised by his suggestion that the decline in the
minimum wage in real terms was a major failure of the Chilean adjust-
ment process. Any attempt to increase the overall efficiency of the econ-
omy should try to deal with all distorted markets, including the labor
market. In that sense, allowing the minimum wage to decline-and
thereby reducing the extent of this distortion-cannot be seen as a neg-
ative development. Although income distribution certainly represents a
(very) weak spot in the Chilean economy, a high minimum wage is not
the most efficient way of dealing with the problem. Any formal analysis
would clearly indicate that the minimum wage is, at best, an nth best
policy; subsidies, social expenditure targeting, and other measures are
under almost all circumstances better ways of improving income distri-
bution.
My final point has to do with the outlook for the future. The author
argues that despite Chile's recent solid economic performance, the Chi-
lean economy continues to be quite vulnerable. I tend to agree with this
view. Three areas of particular weakness in the economy are the levels



Chile: Economic Crisis and Recovery 493
of investment and savings and the very uneven income distribution. But
probably the most urgent issue that the newly elected government will
face in the early 1990s concerns the need to maintain fiscal discipline
and macroeconomic stability. How will the newly elected democratic
government persuade the public that it will actually maintain fiscal dis-
cipline and will not resort to the old inflationary ways? The public will
surely remember the macro-instability of the 1960s and early 1970s, and
consequently there is the danger, even in the absence of any real indis-
cipline, that public responses to these expectations will generate desta-
bilizing forces.
How then can credibility be obtained? A key step in acquiring much-
needed credibility is to make the Central Bank an independent institution.
Such an arrangement, by introducing an important "precommitment
technology," would greatly reduce the risk of irresponsible fiscal behav-
ior. Legislation to this effect would be greatly beneficial for the country
as a whole and should be pursued by all segments of the political spec-
trum.



22 Mexico: Adjustment and
Stabilization
John Nash
MEXICO'S ECONOMIC CRISES of 1982 and 1985 had their origins in the late
1970s. In 1976, in response to severe macroeconomic imbalances and a
high debt service ratio, the L6pez Portillo government reached a stabi-
lization agreement with the International Monetary Fund (IMF). Some
progress was made in 1976 and 1977 in reducing the fiscal deficit and
inflation and improving the balance of payments, but several events in
1977 undermined the momentum of the stabilization program: large oil
reserves were discovered, unemployment rose, and for the first time in
three decades gross domestic product (GDP) per capita failed to increase.
Pressures built up to soften the stabilization measures and to stimulate
growth through increased public sector expenditure financed by borrow-
ing against prospective oil revenue.
The public deficit mounted each year until it reached 18 percent of
GDP in 1982, with public sector outlays reaching 45 percent of GDP (com-
pared with 21 percent in 1970). An important contributing factor was
the expansion of state enterprises. Credit policy, which since 1973 had
served to monetize the fiscal deficit (Blanco and Garber 1986), remained
expansionary until the government attempted in 1982 to bring the fiscal
situation under control through credit restrictions, which were borne
largely by the private sector. Despite rising oil export earnings, the mer-
chandise trade balance deteriorated until 1981. Imports rose while nonoil
export earnings stagnated.
External conditions, by contrast, had on balance been favorable to
Mexico between 1976 and 1982, with improved terms of trade out-
weighing increased debt service (on the 1977 debt) by a wide margin.
The improved terms of trade increased real income by about 4.4 percent
of GDP, while increased debt service decreased it by 2.7 percent. Toward
the end of the period, however, the terms of trade were no longer im-
proving, the interest rate was increasing, and the debt that the country
had to service had increased greatly.
By 1981 foreign lenders had become cautious about providing new
money or rolling over short-term debts. The domestic private sector also
lost confidence in the authorities' ability to maintain the value of the
494



Mexico: Adjustment and Stabilization 495
peso. The nationalization of private commercial banks in September 1982
and the forced conversion of foreign currency accounts into pesos at an
artificially low exchange rate further exacerbated the crisis of confidence
and led to a massive flight of capital. The government had to resort to
large foreign borrowing, mostly short term with rising interest rate
spreads. The acute foreign exchange shortage and consequent import
restrictions, rather than any improvement in underlying conditions, re-
sulted in a large surplus in the trade balance and a sharp decline in the
current account deficit in 1982.
The crisis came to a head in August 1982 when the government of
Mexico suspended interest payments on foreign loans. It then negotiated
an agreement with the IMF for about US$3.57 billion, with drawings to
begin in January 1983 and to continue if certain performance criteria
were met. The key element of the program was a drastic reduction in the
public sector deficit from 18 percent of GDP in 1982 to 3.5 percent in
1985. The program also called for increases in the trade surplus and
foreign exchange reserves. These measures were projected to bring down
inflation to 55 percent in 1983, with further reductions in later years.
No decline in GDP during 1983 was foreseen, and a fairly robust recovery
was expected in the following years.
The Economic Crisis of 1985
The administration of President Miguel de la Madrid, which came into
office on December 1, 1982, introduced several measures to increase tax
revenue and raised the domestic prices of oil products and public utilities.
A commitment was made to reduce budgetary subsidies over time. The
1983 budget also provided for cuts in both current and capital expen-
ditures. The policy of low interest rates was reversed to encourage mo-
bilization of domestic resources and discourage capital flight. The gov-
ernment also entered into negotiations with foreign creditors to refinance
the public external debt falling due before 1985 and introduced a scheme
to cover the foreign exchange risk on the private sector debt and to
facilitate its restructuring.
A sharp reduction in the public sector borrowing requirement in 1983
permitted a large share of credit to shift to the private sector. Reinforced
by generous tax allowances for investments and growing confidence, real
private sector activity shot up during the last quarter of 1984. However,
the underlying incentive structure was distorted by an extensive system
of trade and other controls, and it exacerbated the pull of resources from
exportable sectors to protected domestic sectors.
The economy remained bottled up behind protectionist walls and vul-
nerable to modest shortfalls in meeting policy targets (World Bank
1988b). The target for the public sector borrowing requirement for the



496 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 198Os
last quarter of the year was substantially exceeded, and this intensified
the competition for domestic savings. Inflation also greatly exceeded the
program target, and as inflationary expectations rose, financial disinter-
mediation resumed. The balance of payments position also weakened as
the real exchange rate, which had been depreciated by an unprecedented
amount in 1982, began to appreciate rapidly. Nonoil exports, which had
responded strongly to the real depreciation, slowed and then turned
downward, while capital flight accelerated. By mid-1985 the rate of for-
eign exchange losses had risen to US$500 million monthly. With the
approach of congressional elections, monetary and fiscal policy became
increasingly expansive, and by July a number of ceilings in the quarterly
fiscal and balance of payments program had been surpassed. Drawings
of IMF funds were suspended.
Despite the tragic and costly earthquake of September 1985, the gov-
ernment unveiled an austere budget in November 1985. It called for
raising the primary fiscal surplus (that is, the surplus net of interest pay-
ments) from 3.3 percent of GDP in 1985 to over 9 percent of GDP in 1986.
Partly because of a collapse in international oil prices, however, revenue
increased less than noninterest expenditure, and the primary surplus fell
by 1.6 percent of GDP between 1985 and 1986. Because of the increased
cost of debt service, public sector borrowing increased by almost 6 per-
cent of GDP.
In the period leading to the 1985-86 crisis, the net effect of external
shocks was again favorable to Mexico, although the movements were in
opposite directions from those that occurred before the 1982 crisis. The
real interest rate on Mexico's external debt declined by almost half be-
tween 1982 and 1985, which reduced the cost of servicing the 1982 debt
by about 6.1 percent of GDP, while Mexico's terms of trade declined by
about 37.6 percent, the equivalent of a loss of income of about 3.0 percent
of GDP. Thus the primary reason for the 1985-86 crisis was the contin-
uation of excessive expenditure-although it must be acknowledged that
the 1985 earthquake made all fiscal goals more difficult to attain-in
combination with an incentive structure so distorted by trade and other
policies that sustainable and efficient growth of the private sector was
difficult.
In December 1985 the government of Mexico and the IMF entered into
lengthy and at times acrimonious negotiations over the terms of a new
standby agreement. Concurrent negotiations with commercial banks and
the World Bank produced in July 1986 perhaps the largest and most
complex package concluded for a major debtor nation in recent decades.
It included US$12 billion in new money, US$2.3 billion in contingent
financing (of which only US$0.5 billion was used), and rescheduling of
the service falling due on US$70.5 billion of existing external debt (World
Bank 1987). Under the IMF standby agreement, Mexico was scheduled



Mexico: Adjustment and Stabilization 497
to receive up to US$1.7 billion in quarterly disbursements through the
first quarter of 1988, contingent upon flexible criteria determined by
conditions in the oil market and Mexico's growth performance.
The World Bank played a large role in producing this package of agree-
ments. It agreed to provide new money directly and to cofinance some
of the new loans from commercial banks. About half of the new World
Bank lending commitments, averaging US$2.0 billion annually during
1986-87, were to be in the form of quick-disbursing policy-based loans.
These commitments were triple the amount of World Bank lending to
Mexico in the 1981-85 period.
The World Bank also played an important role in triggering the release
of up to US$500 million for a Growth Contingency Cofinancing Facility
from the commercial banks. If certain conditions (some linked to specific
World Bank adjustment loans) were met, and if Mexico's economic re-
covery did not materialize by March 1987, this facility was to become
available for designated investment projects; half the amounts drawn
under the facility would be guaranteed by the World Bank. The condi-
tions were met, and these funds became available to Mexico in April
1987.
The Adjustment Program
Before the 1985 crisis the World Bank had made only one policy-based
loan to Mexico, a 1983 export-development operation that had mixed
success. The loan package negotiated between Mexico and its creditors
in 1986 thrust the World Bank into a central role in the policy dialogue,
since the release of much of the commercial bank money was linked to
various prospective policy-based World Bank loans then under discus-
sion. Between late 1986 and early 1988 the World Bank's Executive
Directors approved lending operations totaling more than US$2.2 billion.
These loans were based on policy reforms with regard to agriculture,
fertilizer, and steel; export development; and changes in the import re-
gime (see table 22-1 for a summary of adjustment measures and their
implementation).
Trade Reform
Following the 1982 crisis the government instituted reforms in the trade
regime, mostly to promote exports. Controls on temporary imports (used
as inputs for exports) were eliminated, exporters were given access to
foreign exchange at the controlled rate, and in-bond facilities (for ex-
ample, customs warehouses) were provided. Administrative procedures
for exporters were simplified. Export taxes on certain (largely agricul-
tural) products were eliminated, and the number of products subject to



Table 22-1. Mexico's Implementation of Adjustment Measures
Policy area                  Significant progress                       Some progress                   Little or no progress, or regression
Trade              Reduction of coverage of quantitative   Creation of quasi-free trade status for
import restrictions*                    exporters '
Reduction of tariff protection of imports*
4,1                      Reduction of coverage of quantitative
410                        export restrictions*
Macroeconomic      Depreciation of real exchange rate      Reduction of noninterest government      Reduction of total government
Reduction of sectoral tax incentives      expenditures                             expenditure
Reduction of inflation (signs of dramatic
decline in late 1988)
Financial                                                  Reduction of directed credit             Denationalization of commercial banks
Decontrol of some types of deposits
Reduction of credit subsidies
Public sector      Privatization and liquidation of minor  Restructuring and reduction of fiscal    Privatization of major public enterprises
state enterprises                       losses by state enterprises*
Reduction of market power of the
National Food Marketing Company
(CONASUPO) 9
Some reduction in public sector
employment (after increases until 1984)



Pricing              Reduction of food subsidies                 Reduction of coverage of price controls
(until the Economic Solidarity Pact)
Linkage of agricultural prices to border
levels*
Reduction of subsidies for public services
Capital account                                                 Reduction of controls on foreign
investment
Labor market                                                                                                Reduction in labor market regulations
Administrative                                                   Revision of public sector procurement
regulations
Compensatory        Better targeting of food subsidies
measures
D Related to tranche-release conditions of adjustment loans.
Source: World Bank.
so0



500 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
export licenses was reduced. The World Bank supported these measures
in the discussions leading to the first export development loan. Although
specific loan conditions dealt mainly with export promotion measures,
the discussions touched on broader issues, including unification of the
exchange rate and maintenance of the rate at a realistic level, reduction
of credit subsidies, and replacement of import quotas with tariffs. Most
of these measures were eventually adopted by the government.
The major steps in trade policy reform were not taken until 1985. In
July the government announced a 20 percent devaluation of the con-
trolled exchange rate, followed soon by the introduction of a managed
floating exchange rate and a marked acceleration in the average nominal
rate of devaluation. Although a dual exchange rate system was retained,
the rates were managed so that they would eventually coalesce into a
unified rate. Import liberalization also began in earnest in July 1985 with
the rollback of quantitative restrictions on imports, which had formed
the basis of an implicit import protection policy for nearly forty years.
In 1986 the government reinforced this move toward greater openness
by requesting membership in the association of the General Agreement
on Tariffs and Trade (GATr), lowering the maximum tariff from 100
percent to 50 percent, announcing a three-year plan to narrow the range
of import tariffs from 0 to 100 percent to 0 to 30 percent, pledging to
eliminate all official reference prices on imports (another protective de-
vice) by the end of 1987 in concert with the introduction of a GATT-
compatible antidumping system, and pledging to further reduce coverage
of quantitative restrictions by the equivalent of 5 percent of domestic
production by December 31, 1986. Throughout the development of this
program, the World Bank and the government of Mexico remained in
close touch. Eventually, in 1986, the World Bank approved a US$500
million trade policy loan, with some of these actions as tranche-release
conditions.
During 1987 the government met its pledges and accelerated its pro-
gram of narrowing the tariff range by reducing the maximum tariff to
20 percent. In addition, it further reduced the coverage of quantitative
restrictions from 49.8 percent of 1983 national production to 43.1 per-
cent (or to 25 percent of 1986 production figures). This program was
supported by a second trade policy loan of US$500 million. These reforms
represent tremendous progress, although it is not clear that the anti-
dumping system is consistent with GATr rules, and there is concern that
it may be used to provide concealed protection.
Certain administrative reforms governing public sector procurement
that were made in January 1988 may also considerably advance import
liberalization. Before January 1988 all public entities had to procure
domestically all goods available in Mexico at reasonable prices and qual-
ity, and public imports valued at more than US$50,000 had to be licensed.



Mexico: Adjustment and Stabilization 501
Because the public sector accounted for a large share of total purchases
(30 percent) and imports (25 to 33 percent), these procedures created a
strong anti-import bias and provided additional protection to industries
producing for the public sector. It is still too early to judge whether the
new regulations will be effective in redressing this bias, however.
A number of important export-promotion measures were also adopted.
Many of these measures were encouraged by the World Bank through
two export development loans (1983 and 1987). Domestic letters of credit
now permit exporters and their suppliers to obtain export financing and
foreign exchange needed for imports. Under an import admission scheme,
exporters signing multiyear agreements with the government can import
equipment and intermediate inputs with reduced guarantee requirements
and receive rebates of duties and indirect taxes. Exporters can now retain
all of their foreign exchange earnings for future imports. Credit lines for
export financing were consolidated, and export insurance and guarantee
schemes simplified. Quantitative export controls were reduced, and the
coverage declined from 85 percent of nonoil exports in 1980 to 44 percent
in 1985. The government subsequently agreed to remove some of the
remaining export controls, which cover mostly agricultural products. For
some of these products, especially beef, the government had manipulated
the controls to ensure low internal prices and depressed producer prices.
The government's policy of strong exchange rate devaluation, which
coincided with a period of declining real wages, increased Mexico's cost
advantages over other export-oriented countries. Real manufacturing
wages and benefits declined by nearly 25 percent between 1981 and 1987
in local currency (about 40 percent in U.S. dollars), to make Mexico's
wage rates among the lowest of the major developing-country exporters.
For several reasons, the reform of trade policy was the most far-reach-
ing and successful of the economic reforms undertaken. First, the de la
Madrid administration, with strong support from the World Bank and
prompted by declining oil prices, concluded early in its term that Mexico's
development required a shift from an inward to an outward orientation.
Second, the government's commitment was reinforced by the recognition
that import liberalization could be a tool in its fight against inflation.
Third, both the World Bank and the government concluded that trade
reform was a prerequisite for stabilization and for the success of reforms
in other sectors. Fourth, the reforms, though symbolically important and
politically difficult, appear not to have been extremely painful. Nontariff
barriers and tariffs were not fully binding at the time of the reforms
because imports had been severely depressed by general economic con-
ditions. As quantitative restrictions were reduced, official reference prices
and tariffs were used to provide continuing, temporary protection before
being gradually eliminated or reduced. Depreciation of the exchange rate
cushioned the blow to import-substituting industries.



502 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Fiscal Reform
Fiscal reform centered on efforts to reduce the budget deficit and improve
the tax structure and incentives. Public sector investment fell from 10 to
5 percent of GDP between 1982 and 1987 while current spending (ex-
cluding interest) declined from 26 percent to around 20 percent of GDP.
But the effect of inflation on interest rates and, consequently, on the cost
of servicing domestic debt left total public spending in 1986 at around
47 percent of GDP, compared with a 1978-81 average of 34 percent. And
because spending cuts were concentrated more in investment than in
current expenditure (especially the wage bill), the effect on growth was
less than hoped. Also on the negative side was the decline in revenue
from the sale of already underpriced public sector services such as elec-
tricity, transportation, and fertilizer.
Changes in the tax structure and the system of fiscal incentives were
also significant. The Tax Reform Act of 1987 is expected to reverse
gradually the sharp decline since 1980 in corporate taxes as a share of
GDP and should improve collection rates. The use of generous investment
depreciation allowances, however, may act to increase subsidies to cap-
ital-intensive investment. With respect to fiscal incentives, the reduction
in tax subsidies between 1981 and 1987 brought the level of forgone
revenue to well under 0.5 percent of GDP. In 1988 all tax subsidies, except
those in agriculture, were eliminated. Other tax credits for investment
remain in place, targeted by region rather than industry in order to en-
courage investment outside the three major cities.
Reform of the Domestic Financial Sector
The banking system was nationalized in the last year of the L6pez Portillo
administration (1982) and remains under government ownership. The
government, however, permitted private brokerage houses to perform
some banking functions, which effectively put them in competition with
banks for deposits. It was largely these brokers who financed the devel-
opment of Mexico's first stock market.
Interest rates were kept low before 1983 and became highly negative
in real terms in 1982 as inflation surged. This implied large transfers,
both from savers to borrowers and from the government budget to bor-
rowers from lending institutions financed by the government. In 1983
most loan rates were raised to positive real levels, although the contin-
uation of large directed credit programs at preferred rates reduced the
funds available to nonpreferred borrowers and raised their rates to ex-
tremely high levels. Later the government relaxed many of the require-
ments that banks channel credit to certain sectors, as well as other reg-
ulations and implicit taxation of financial intermediation activities.



Mexico: Adjustment and Stabilization 503
Although this was a significant step toward limiting bureaucratic control
of the financial sector, a substantial amount of regulation remains. Of
greater concern, however, is that the public sector continues to absorb
66 percent of available credit. The World Bank, in part because of the
political sensitivity of the issue of nationalization, has been reluctant to
tackle reforms in this sector in a comprehensive way.
Public Sector Reform
Public sector policy remains a serious obstacle to improving the efficiency
of the Mexican economy. From 1970 to 1982 the number of public
enterprises increased from fewer than 100 to more than 1,000. Once
established, such enterprises are difficult to liquidate or sell, while efforts
to make their operations more efficient run up against the political power
of the public sector unions.
Despite political opposition, the de la Madrid administration under-
took some reforms of the public sector by divesting some firms and re-
structuring others. Under the current divestiture program, the govern-
ment proposes to sell 250 firms, liquidate 395, merge 80 with other
agencies, and transfer 30 to state governments. This program is one of
the most ambitious ever undertaken and will ultimately reduce by 65
percent the number of firms owned by the state. The real impact will be
far less, however, since the affected enterprises own only about 10 percent
of the sector's assets.
The restructuring effort was modestly successful. In the agricultural
and food sectors, the government took steps to improve the operation
of the giant National Food Marketing Company, CONASUPO. The federal
government assumed over 80 percent of CONASUPO's debt (US$1.8 billion)
in return for CONASUPO's pledges to streamline its operations and to pro-
mote private sector participation in some markets. CONASUPO's monopoly
on imports of maize, sorghum, rice, oilseeds, and barley was eliminated,
although it remains the sole legal importer of milk and beans. Many of
CONASUPO's retail outlets were moved to low-income neighborhoods to
improve targeting of subsidies. In addition, the government took the
politically difficult step of closing six high-cost sugar mills and is studying
the sale or closure of others. An agricultural loan from the World Bank
supported the relocation of the retail CONASUPO stores, closure of sugar
mills, and other steps to reduce the role of state agricultural enterprises.
Steps were taken to improve the efficiency of state enterprises in other
sectors as well. Plants were closed in the fertilizer and steel sectors. The
national railroad is being restructured with support from World Bank
lending, shipyards are being restructured to benefit from new technology,
and reorganization is being contemplated for the state oil company, the
petrochemical industry, and the state truck manufacturing firm. Aero-



504 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
mexico, one of two state-owned airlines, was allowed to go bankrupt
and is now being reorganized. To show the seriousness of the reform
effort, the government arrested the powerful head of the national oil
company (PEMEX), which had operated as a corrupt state-within-the-state
and thwarted all previous efforts at reform.
Still, public sector influence in the economy remains far too high. The
divestiture program notwithstanding, 650 state enterprises remain, com-
pared with between 39 and 84 in 1970 (depending on the definition used).
A cutback of 25,000 public sector staff in 1985 did little to offset the
increase in employment in the sector from 3.5 million in 1981 to 4.2
million in 1984. Reducing public enterprises' control over the economy
will require building a political consensus. Little real improvement can
occur, however, as long as the problem is seen as one of attaining tech-
nical efficiency rather than of basing the incentive structure on the de-
cisions of individuals in a market instead of on political pressures.
Pricing Reform
Control of consumer prices in the private sector was relaxed to some
extent. In December 1982, 55 percent of national production was under
strict controls, but by the end of 1986 coverage was reduced to 37 percent.
The changes affected mostly industrial products; about 55 percent of
primary agriculture, fishing, livestock, and agroindustry remained under
strict control. Under the Economic Solidarity Pact announced in Decem-
ber 1987, the strict control list was extended to all basic commodities.
The government's plans after the expiration of the pact (which was ex-
tended through July 1989) are not clear but may call for a return to the
control system in effect in 1987, followed by a gradual phaseout.
As part of the World Bank's agricultural loan, the government in 1989
agreed for the first time to link the prices of crops under the guarantee
price system to border prices. Prices would be maintained within a band
around their international levels, adjusted for differences in quality and
transport costs. Maize and beans, the most important of the crops under
the guarantee price system, were excluded because of their political sen-
sitivity and the technical difficulty of including them in the new scheme.
The government also agreed to maintain support prices in 1988 at their
real 1987 levels. Because pricing did not conform to these agreements,
however, the release of the second tranche of the loan was jeopardized.
After the 1982 crisis the government attempted to reduce the subsidies
on goods and services produced by government-owned entities by in-
creasing their real prices. Although it is hard to judge the efficiency effects
of these efforts, the budgetary effects were salutary. Raising interest rates
on preferential credit lines reduced interest rate subsidies channeled
through official development banks from 5.4 percent of GDP in 1982 to



Mexico: Adjustment and Stabilization 505
3.1 percent in 1985. The government later announced plans to eliminate
all interest rate subsidies except those to low-income agricultural bor-
rowers. Higher real retail prices on controlled basic foods were linked
with a phaseout of subsidies to processors, including CONASUPO. These
subsidies thus decreased by over 80 percent in real terms between 1983
and 1986. Global food subsidies continue only for sugar and some maize
products. Attempts to reduce subsidies for oil derivatives and state-run
transportation were complicated by the high and unpredictable inflation
of costs, with the result that many of the gains in subsidy reduction were
lost. For example, public sector prices were increased by around 80 per-
cent in November 1987, but by March 1988 the increases had been so
eroded by inflation that the prices in real terms were only about 20
percent over their November levels.
Capital Account
To mitigate the problems created by its strict rules governing foreign
investment, the government clarified the basic legislation regulating for-
eign investment, simplified the approval process, and expanded the class
of businesses in which foreign ownership is permitted. In 1986 the gov-
ernment launched a debt-equity conversion program that enables foreign
investors to acquire Mexican public sector external debt in the secondary
market at discounts of up to 30 percent. These instruments are redeemed
by the government at 75 to 100 percent of face value in domestic currency,
which is then used to finance certain equity investments and other in-
vestment-related expenses.
As a result, approvals of foreign direct investment increased from
around US$0.65 billion annually in 1982-85 to nearly US$2 billion in
1985, US$2.4 billion in 1986, and US$3.9 billion in 1987. However,
actual inflows (as opposed to approvals) declined in 1984 and rose only
slightly to US$0.5 billion in 1985. Inflows were US$1.5 billion in 1986
and US$3.3 billion in 1987, of which 24 percent in 1986 and 46 percent
in 1987 were inflows under the swap program, valued at official con-
version rates. Foreign investment is concentrated in the in-bond industrial
sector (which is required to export a part of production) because the
regulatory framework for other industries is perceived as unfriendly and
the macroeconomic situation remains uncertain. After the swap program
was suspended in November 1987, foreign investment dropped (World
Bank 1988a). Capital flight, which had averaged around US$7 billion
annually during 1979-82, was reduced to half that in 1983-85 and was
negligible in 1986. Capital flight seems to have resumed in 1988, but it
was limited to US$2 billion-US$2.5 billion by the government's reluc-
tance to keep domestic interest rates artificially low.



s06 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
Labor Market Policy
Because the government has raised the minimum wage much more slowly
than the inflation rate, real wages have eroded dramatically-by around
40 percent relative to the U.S. dollar, according to some estimates. There
have been no major changes in the cumbersome regulations governing
labor relations. These regulations, which make it costly for a firm to
reduce its labor force, may impose serious rigidities that could make the
economy slow to adjust to the new incentive structure created by trade
liberalization. Furthermore, the immobility of labor has contributed to
the deep decline of real wages; if the market had been able to reallocate
labor from declining to booming sectors, real wages would not have had
to fall so steeply in order to maintain employment.'
Compensatory Measures
General subsidies on basic foods that were begun on a massive scale in
the late 1970s were gradually reduced beginning in 1983 and eliminated
altogether in 1987, with the exception of those on sugar and some maize
products. World Bank staff estimate that the consequent increases in
prices reduced the purchasing power of the poorest 20 percent of the
urban population by about US$250 million, or by about 6.5 percent of
their income. With the exception of the landless poor, the rural popu-
lation was not adversely affected by higher food prices since they are
subsistence farmers or net producers of these products.
To replace the global subsidies, the government instituted a program
of targeted food stamps that could be used to buy tortillas, a staple food
of the poor, and expanded two other targeted programs (World Bank
1988c). World Bank staff estimate that expenditures on the targeted pro-
grams exceed US$250 million annually, which largely compensates for
the price rises. Unfortunately, the price of the tortilla stamps, which were
redeemable for a given physical quantity of tortillas, was not linked to
the price of tortillas. In a period of high inflation, this caused the subsidy
component of the program to grow rapidly and limited the program's
ability to serve more of the poor. This issue is being addressed in con-
nection with the agricultural loan.
The Outcome
The stabilization and adjustment program has been successful in pro-
ducing the surpluses on the current account that Mexico will need to
service its debt and return to commercial creditworthiness. The program
has been less successful in restructuring the economy and reducing the
government's role, although there has been progress.



Mexico: Adjustment and Stabilization 507
Macroeconomic Balance
Until 1988 the macroeconomic stabilization program notably failed to
control inflation, which increased from 20 to 25 percent in 1978-81 to
nearly 160 percent in 1987. The experience of other countries with high
rates of inflation shows that prospects are poor for eliminating infla-
tionary expectations once they have become established. Nonetheless,
events in the late 1980s suggest that the government may persevere better
with its current program than it has with past ones. The underlying causes
of the accelerating inflation since 1985 are being addressed, with an
increase in the primary fiscal surplus from 5.3 percent of GDP in 1987
to about 7.6 percent in 1988. Monthly inflation was reduced from 16
percent in January 1988 to 0.6 percent in September, as measured by
consumer prices, while wholesale prices even declined slightly in Septem-
ber and October. Thereafter, monthly inflation was variable, reaching
2.4 percent in early 1989, but then declining.
The dilemma for the government, which has relied on price controls
to break inflationary expectations, is how to remove the controls without
rekindling those expectations. Despite the large primary surplus, contin-
ued reduction of the fiscal deficit is complicated by high interest rates
abroad and a domestic interest rate of around 40 percent in real terms.
The high real domestic rate appears to be the result of fears that inflation
may reappear. Real money balances have not increased in Mexico as they
have in other countries after a dramatic decline in inflation.
The record on other macroeconomic performance indicators is mixed.
Real GDP decreased by around 6 percent in 1982-83 and fell again by
3.8 percent in 1986. There was modest growth in 1987 and 1988 (about
1 percent), but the future remains uncertain.
External Balance
The reforms were most successful with respect to external trade, and
large trade surpluses were created. In 1987 nonoil merchandise exports,
both agricultural and manufacturing, stood at 230 percent of their 1982
level. This increase helped offset the adverse impact of the collapse of
oil prices and allowed Mexico to accumulate substantial reserves. Reserve
accumulation continued through March 1988, but then higher interest
rates, declining oil prices, surging imports, and the effects of an exchange
rate frozen as part of the anti-inflation program combined to generate a
loss of US$6 billion by year's end. The loss came in spite of a continued
expansion of nonoil exports. In early 1989 the government began to
devalue the peso through a crawling peg at a 17 percent annual rate. It
is not clear, however, that this will be enough to sustain the boom in
exports.



508 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
It is too early to assess the sustainability and long-run impact of the
trade reforms. As the economy recovers and imports increase, pressure
to reimpose protection is likely-and may already be occurring. A recent
study (World Bank 1988b) concluded that the private sector's initial
belief that the reforms would be temporary, as well as the government's
sluggishness in reducing other sectoral distortions, has impeded the shift
toward export expansion and efficient import substitution. Regulations
governing foreign investment and especially the hiring and firing of work-
ers tend to reduce factor mobility. Employers must feel confident about
the future before they will expand output by increasing either capital or
labor inputs. The government's policy reforms do not yet appear to have
reached this level of credibility; consistency over a longer period will be
required.
Outlook and Political Sustainability of Adjustment
DEBT SERVICE. One can be guardedly optimistic about Mexico's pros-
pects for servicing its debt. Although gross external public and private
debt increased by US$17 billion between 1983 and 1987, official reserves
increased by US$13.5 billion, so that net debt increased by only US$3.5
billion. In 1987 the country's medium- and long-term debt was about
two and a half times its average level during 1978-81. The interest bill
was close to twice the average level during 1978-81 and would have
been much higher had it not been for the debt rescheduling and falling
interest rates since 1984. On the positive side, the ratio of interest to
exports was only slightly higher (25.7 percent, compared with 19.8 per-
cent), and the debt service ratio was lower (40.3 percent, compared with
46.8 percent). These payments will have to be sustained over a longer
period of time because of the rescheduling, but this should be possible
if international interest rates do not escalate, the oil market does not
weaken further, and policy reforms already in place are not reversed.
Not everyone accepts this guardedly optimistic scenario. Some argue
that as long as Mexico devotes a large part of its export earnings to debt
service instead of capital imports, investment and therefore GDP growth
will stagnate and generate irresistible pressure for a suspension of pay-
ments. To reduce the debt burden, Mexico has implemented a debt buy-
back program, but it has met with only limited success, and other pro-
posals have been floated (see Dornbusch and Modigliani 1989). It seems
likely, however, that although debt reduction will certainly promote some
growth, substantial short-term growth can be expected with or without
debt reduction when efficiency-enhancing reforms are made in an econ-
omy as distorted as Mexico's.
SUSTAINABILITY. Since 1982 Mexico has had three years of negative
GDP growth, and it entered 1988 with a lower standard of living than it



Mexico: Adjustment and Stabilization 509
had when the stabilization process began. Wages have fallen in real terms
by around 40 percent, and open unemployment was around 4.5 percent
in 1987, compared with 3.8 percent in 1981. Thus it is not easy to argue
that the program can be sustained. The government, however, has con-
vinced the populace of the necessity of persisting with the program. It
has also cushioned the blow for a large segment of the urban poor by
introducing targeted subsidies as it reduced global subsidies. The Salinas
administration appears to realize that Mexico's economic future depends
on integrating the economy into world markets and increasing its effi-
ciency. As an indication of this commitment, Salinas has filled key
policymaking positions with a first-rate economics team rather than with
traditional politicians.
The key question, however, remains unanswered: To what extent can
this administration overcome the ideas of economic nationalism and the
proper economic role of the state that have been entrenched among the
Mexican people ever since the Revolution? Two threats to the program
appear to be critical. First, if the economy does not soon show signs that
the program is bearing fruit, the pressure for retrenchment from vested
interests might become impossible to resist. And if the growth response
is further delayed, this pressure could jeopardize the macroeconomic sta-
bilization by making it more difficult to reduce the fiscal deficit. Second,
if the government receives a large windfall-for example, from another
oil bonanza-political pressure to increase its role in the economy may
build.
The outlook for sustained reform could probably be improved by a
negotiated reduction in the debt burden or a modest improvement in oil
revenue. Even without debt reduction, however, some of the reforms seem
to be permanent. Certainly Mexico's trade reforms possess many of the
characteristics associated with sustained trade reforms in other countries
(Michaely 1988), including a relatively stable political system, a strong
start to the reforms, a rapid initial depreciation of the real exchange rate,
and an external commitment to the GATr.
The World Bank's Role in the Adjustment Process
Before 1988 the only reforms Mexico introduced as a direct result of
World Bank loan requirements were in trade liberalization. However,
the discussions on the trade-related loans touched on issues only indi-
rectly related to trade reform and led to agreements or loans in the ag-
ricultural, steel, and financial sectors. In other areas, reforms were not
the basis for loans, but they were viewed by the World Bank as evidence
of the government's willingness to make difficult choices. The World
Bank played a supportive rather than a leading role in the adjustment
process, but by strengthening the position of reform-minded factions in



510 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
the government, World Bank support made possible reforms that were
more extensive than would otherwise have been the case.
Conditionality in World Bank Adjustment Loans
For several reasons Mexico's four trade-related loans from the World
Bank in the 1980s contained few conditions that were legal requirements
for putting the loan into effect or for tranche release. First, some con-
ditions, such as maintenance of an appropriate exchange rate, are difficult
to quantify and monitor. Second, it was clear that the government in-
tended to meet many of the trade liberalization goals in any case. Third,
some goals specified in loan documents, such as reducing the public sector
borrowing requirement to less than 4 percent in 1988, would be fulfilled
only after the loan was almost fully disbursed. Fourth, some of the goals
discussed at the time of the loan negotiations, notably reforms of interest
rates and price controls, may have been judged too risky or too tangential
to the loans' chief aims to be included as conditions.
Two "hybrid" sectoral adjustment loans, for the steel and fertilizer
sectors, contained a quick-disbursing component to finance imports,
other investment, and technical assistance. Each has tranche-release con-
ditions related to sectoral restructuring, trade liberalization, and imple-
mentation of more efficient pricing policies.
The agricultural loan includes conditions related to targeting food sub-
sidies, streamlining state agricultural enterprises, eliminating some re-
strictions on agricultural imports and exports, setting some guarantee
prices close to border prices, restructuring the Ministry of Agriculture,
and conducting studies of input and output pricing policies.
The sequencing and the scope of the trade-related loans were appro-
priate. They addressed the major trade policy distortions, beginning with
barriers to exports. It remains to be seen whether these reforms can
succeed in stimulating a supply response while so many distortions re-
mnain in other parts of the economy and whether new nontariff barriers
will be substituted for the quantitative restrictions that have been re-
moved. In the case of both the fertilizer and agricultural loans, some
conditions believed by World Bank staff to be important were dropped
during negotiations (for example, reduction of the coverage of food price
controls), and other desirable changes were not considered because they
were politically sensitive (for example, privatization of fertilizer produc-
tion).
Since the World Bank has made no structural adjustment loans to
Mexico, it has relied on the IMF to monitor macroeconomic performance.
Most of the World Bank's sectoral adjustment lending to Mexico has
been concurrent with IMF programs. The exception was the first trade
policy loan, which was approved in July 1986 after disbursements under



Mexico: Adjustment and Stabilization 511
the IMF program had been suspended. The trade loan required that Mex-
ico's external financing needs for 1986 be met, and since commercial
lending was contingent on Mexico's reaching agreement with the IMF on
an adjustment program, this became a de facto condition of effectiveness.
Some of the recent sectoral adjustment loans have included a tranche-
release condition requiring an appropriate macroeconomic policy frame-
work.
Effectiveness of Implementation
All the conditions included in the four trade-related loans were met, and
the reforms have so far been sustained. The government also met many
of the goals enumerated in the loans but not identified as requirements
for disbursement.
Release of the first tranche of the agricultural loan was delayed because
of lack of agreement on the terms of reference of a series of studies and
the failure to remove export controls on high-quality cuts of beef. This
delay in turn led to a postponement of the release of the second tranche,
which is also threatened by the government's failure to adjust producer
prices because of its concern about inflation.
The fertilizer loan never went into effect because the agreement ne-
gotiated between the Mexican Fertilizer Company and the central gov-
ernment did not call for the plant closures that were required by the loan.
Nor were the required increases in fertilizer prices introduced.
Lessons from the Adjustment Experience
The main lessons from Mexico's experience are related to the sequencing
of stabilization and adjustment and to the dialogue between the World
Bank and the country in the context of policy-based lending. Questions
remain, however, about the economy's future response to the changes
that have been made.
Sequencing
Some economists have argued that when inflation rates are high (say,
greater than 25 percent a year), priority should be given to stabilization,
with structural adjustment policies postponed until after inflation rates
have fallen (see, for example, Corbo and de Melo 1987; Fischer 1986;
and Sachs 1986). In a highly inflationary environment, it is reasoned,
relative prices among sectors are so variable that they will give uncertain
signals about where investment should occur.
Conversely, Mexico's experience between late 1983 and mid-1985
shows the danger in stabilization without adjustment. Inflation rates were



512 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
falling and economic activity was rekindling at the beginning of this
period, but the recovery occurred in an environment distorted by a pro-
tectionist trade structure. Thus resources were attracted to the protected
sectors and away from exportable sectors. The consequent balance of
payments crisis led to the crash program of trade reform that began in
July 1985. Through the end of 1987 that program was carried out during
a period of high and rising inflation. Yet in February 1988 the World
Bank concluded that the incentive structure had shifted in the expected
direction, with both import protection and the bias against exportable
sectors having been reduced. Although shifts in investment had been
impeded by regulatory and other constraints, the expected shifts in re-
source allocation were nevertheless apparently beginning (World Bank
1988b).
Inflation, it appears, adds "noise" to relative price movements, but it
does not alter the fundamental pattern of shifts in prices following
changes in trade policy. To some extent, economic agents know before-
hand what sectors will gain or lose from trade reform, and they can check
their expectations against average price movements over a fairly short
period. Inflation, therefore, should not be considered an insuperable ob-
stacle to adjustment. Other barriers to resource movement, such as reg-
ulatory constraints, may be as important. If that is so, then an adjustment
program should be wide-ranging and should give special attention to
regulatory reforms.
Mexico's second and current attempt at stabilization occurred in a
regime of significantly restructured incentives. The boom in nonoil ex-
ports, together with external resources from the World Bank, the IMF,
and commercial banks, provided reserves that made it possible for the
government to use the exchange rate as a nominal anchor in its anti-
inflationary program. The fact that the antiexport bias had been greatly
reduced while the exchange rate had reached a very competitive level at
the end of 1987 meant that the exchange rate could be frozen throughout
much of 1988. This led to a real appreciation, without tremendously
depressing exports. In Mexico's case, stabilization before adjustment was
an unqualified failure, while stabilization after adjustment may tenta-
tively be judged a success.
The Trade Reform Program
The Mexican trade reform program shows how constructive dialogue
between the World Bank and a country can be combined with policy-
based lending to achieve worthwhile objectives. The government planned
its import liberalization program and announced its schedule in advance.
When the. timing was opportune, the government did not hesitate to
accelerate the process since the World Bank had assured Mexico that



Mexico: Adjustment and Stabilization 513
"overperformance" would be credited toward a second trade loan. The
process was carried out in stages, to give time for any necessary adjust-
ment of plans. Tariffs were first substituted for nontariff protection to
make the system visible; this in turn made it easier to monitor the reform
and eventually to reduce protection. The Mexican experience with import
liberalization may provide a model for future World Bank operations in
the trade sector.
Liberalization of exports began several years before liberalization of
imports, in accordance with the generally accepted argument that imports
respond more quickly to liberalization and therefore create a balance of
trade deficit unless exports have been liberalized first. As it turned out,
imports declined in 1983, 1986, and 1987, apparently because of do-
mestic demand contraction and (in 1983) a shortage of foreign exchange.
It is not clear that any balance of payments crisis would have been pre-
cipitated had the liberalization sequence been reversed. Export-import
sequencing may not be an important determinant of the success of a
reform program (see also Michaely 1988).
The trade reform dialogue between Mexico and the World Bank owed
its success to two factors. First, the World Bank did not press for reforms
that were at odds with the government's own program. Rather, it pressed
for more far-reaching reforms than the government might otherwise have
considered. The World Bank took into account both the reform-minded
and the conservative elements in the government, and by its encourage-
ment and incentives apparently strengthened the former. Second, the dia-
logue began at a propitious time. In 1982 a new administration was
assuming power and policies were still in the formative stage. It was
helpful to the World Bank's position that this new administration was
in deep need of financial assistance and headed by technocratic elements
less steeped in Mexico's traditional politics than previous government
leaders. It was also helpful that the government was petitioning for acces-
sion to the GArr and was therefore required to fulfill conditions that
paralleled those that the World Bank judged necessary for structural
adjustment. Mexico was given full credit for reforms it carried out in
order to accede to the GATr, even when these were not preexisting loan
conditions. The Mexican experience supports Michaely's (1988) conclu-
sion that a multilateral commitment to reform, although it does not guar-
antee success, at least helps to prevent temporary aberrations in the re-
form process.
The trade liberalization process largely bypassed agricultural imports;
almost all basic food imports are still under direct or indirect government
control. The World Bank should continue to press for changes in this
area, being careful to reduce protection to agriculture only to the extent
that industrial protection is also reduced. (Although this may argue for
a rather slow reduction in agricultural tariff rates, it does not argue for



514 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
sluggishness in eliminating import licensing, government monopolies,
and other controls in these markets.) The World Bank should also pay
careful attention to any evidence that antidumping, health, safety, and
pollution regulations are being used as protection devices.
Complementary Policies
The key question remaining is whether and how quickly the economic
structure will respond to these changes in trade policy. Investment has
been slow to respond to the new price signals for several reasons (World
Bank 1988b). Initially, investors did not find the government's reforms
to be credible, and they were also deterred by an unstable real exchange
rate. In addition, investible funds were scarce and expensive because the
government had tightened monetary policy without substantially reduc-
ing its own borrowing requirements.
There has been only a slight reduction in government direction of the
economy. The government and the World Bank have only recently begun
to address issues of sectoral liberalization. There is a danger-and some
evidence of this is surfacing already-that the full gains from trade lib-
eralization will not be realized because of the delays in following through
on these other reforms. They should probably constitute the main focus
of future World Bank lending to Mexico.
Note
Many thanks to Bela Balassa, Suman Bery, Vittorio Corbo, John Johnson, Don
Keesing, Bill McCleary, Jaime de Melo, and Sreenivasa Ramachandran for helpful
comments on early drafts of this chapter. Remaining errors, of course, are my
responsibility.
1. I owe this point to Vittorio Corbo, chief of the World Bank's Macroeco-
nomic Adjustment and Growth Division.
References.
Blanco, H., and P. Garber. 1986. "Recurrent Devaluation and Speculative At-
tacks on the Mexican Peso." Journal of Political Economy (February):148-
66.
Corbo, Vittorio, and Jaime de Melo. 1987. "Lessons from the Southern Cone
Policy Reforms." World Bank Research Observer 2 (2):111-42.
Dornbusch, Rudiger, and F. Modigliani. 1989. "Easing the Mexican Interest
Rate." Wall Street Journal, January 3.
Fischer, Stanley. 1986. "Issues in Medium-Term Macroeconomic Adjustment."
World Bank Research Observer 1 (2):163-82.



Mexico: Adjustment and Stabilization 515
Michaely, Michael. 1988. "Trade Liberalization Policies: Lessons of Experience."
Paper prepared for the conference "For a New Policy Toward Foreign Trade
in Brazil." Sao Paulo, Brazil, April 11-12.
Sachs, Jeffrey. 1986. "Conditionality and the Debt Crisis: Some Thoughts for
the World Bank." Development Research Department, World Bank, Wash-
ington, D.C.
World Bank. 1987. "Mexico after the Oil Boom: Refashioning a Development
Strategy." Latin America and Caribbean Region Report 6659-ME. Washing-
ton, D.C.
. 1988a. "Direct Foreign Investment in Mexico: Past Patterns and Future
Strategy." Latin America and Caribbean Region Report. Washington, D.C.
. 1988b. "Mexico: Trade Policy Reform and Economic Adjustment."
Latin America and Caribbean Region Report 7314-ME. Washington, D.C.
. 1988c. Targeted Programs for the Poor during Structural Adjustment:
A Summary of a Symposium on Poverty and Adjustment, April 1988. Wash-
ington, D.C.
Comments
Vittorio Corbo
JOHN NASH'S chapter provides a well-balanced interpretation of Mexico's
adjustment program. Since I agree with most of it, my comments will
concentrate on certain points that require further elaboration.
Although it is true that the crisis of the early 1980s in Mexico was
caused by expansionary fiscal and monetary policies-which, given a
fixed exchange rate regime, resulted in appreciation of the real exchange
rate, capital flight, and the accumulation of a large foreign debt-one
should also recognize that Mexico made major adjustments on the fiscal
front after 1982. The primary balance (that is, net of interest payments)
improved by nearly 16.0 percent of GDP by 1988. The operational deficit
also improved in spite of the high real interest rates on government debt.
This fiscal adjustment is all the more remarkable in view of the drastic
decline in the terms of trade in 1986 as a result of the drop in the price
of oil.
Nash is right to indicate that in the first half of 1985 the elections had
a negative impact on fiscal adjustment efforts. After the elections, how-
ever, not only were fiscal adjustment efforts intensified, but a major
restructuring of trade incentives was initiated.



516 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 198OS
An aggressive policy of nominal devaluations in 1986-87 resulted in
a large surplus in the current account of the balance of payments, as well
as an acceleration of inflation. The acceleration of inflation placed the
overall adjustment program in jeopardy. To their credit, the Mexican
authorities implemented a heterodox stabilization program in late 1987.
This program included further fiscal adjustment and the use of income
policies to deal with the persistent elements of inflation. To date, Mex-
ico's stabilization program, with its large dose of fiscal adjustment, has
been quite successful in reducing inflation.
In the area of structural reform, Mexico has also made important
progress. After the onset of the debt crisis, Mexico decided to manage
the crisis by introducing a host of trade restrictions that further isolated
Mexico from the world economy. In the middle of 1985, after much
internal discussion, the Mexican government initiated a major restruc-
turing of its trade regime that resulted in a maximum 20 percent tariff
rate and eliminated most quantitative restrictions.
Mexico has made major progress in stabilizing and restructuring its
very distorted domestic economy. To reap the benefits of its structural
reforms, however, Mexico needs to succeed in stabilizing its economy.
It also needs to complement its trade policy reforms with a liberalization
of the domestic financial system and the rationalization of its highly
inefficient system of domestic regulation.
In conclusion, in spite of the heavy burden of debt accumulated mainly
to finance capital flight and unproductive public investment, Mexico has
made much progress in adjusting its economy. Ultimate success will be
achieved when stabilization is more permanent and growth is sustainable.
Mexico is well on the way to achieving these objectives.
Comments on Latin America
S. Shahid Husain
IN SUMMING UP the discussion on Latin America, I would like to add my
own thoughts on some often neglected, but critical, aspects of the ad-
justment process: the political economy, including the role of govern-



Mexico: Adjustment and Stabilization 517
ment, and the social dimensions. These sociopolitical considerations bear
on the success of economic reforms. They determine how much adjust-
ment is carried out, how long the process takes, and what the results will
be.
The two countries discussed here, Chile and Mexico, are relatively large
economies that have made substantial progress in adjustment. In the case
of Chile adjustment has in recent years been accompanied by rapid
growth. No one, however, should ignore the fact that both countries
have had a tremendous continuity of government and governmental pro-
cesses. In the case of Chile continuity came through an authoritarian,
dictatorial regime. It is not coincidental that adjustment in Chile is a
process that goes back to the mid-1970s and did not begin to bear fruit
until well into the middle 1980s. Mexico's government, too, although
nominally democratic, has been authoritarian and has had a much greater
integration among the civil service, the political process, and labor than
do most Latin American countries. So the importance of the continuity
of policies is inescapable.
The crisis of economics and economic development in Latin America
is fundamentally a crisis of government. The process of adjustment in-
volves, to a great extent, determining how to reduce to reasonable pro-
portions the tremendous amount of rent creation and rent receiving that
has proliferated around the government. The question to consider is
whether in large, diverse countries such as Brazil and Argentina, with
volatile political processes, results similar to those achieved in Mexico
and Chile can be expected.
We should not be deluded by the relative smoothness of this process
in Chile and Mexico into thinking that it can be duplicated with ease in
other countries. The process of adjustment is going to be very discon-
tinuous, in part because of factors related to the cost of adjustment and
the society's ability to withstand it. Although Chile and Mexico were
able to withstand some of the pain that accompanies adjustment, such
as the tremendous decline in real wages in Mexico, a similar decline over
an extended period in countries such as Brazil and Argentina is likely to
provoke substantial political repercussions. Another issue with regard to
the discontinuity of adjustment concerns needed reductions in regulations
and in the size of the corporate state. In Chile this process has gone the
furthest. How fast can it proceed elsewhere? For example, despite efforts
in Argentina to privatize the airlines and the telephone company, the
government has not been able to mobilize the political will to implement
these changes. Argentinian railways therefore continue to lose a billion
dollars a year while carrying only 6 percent of total traffic.
Although we need to talk about the economic issues of adjustment,
we should not lose sight of the political and social issues of adjustment.
These issues ultimately form the background of economic policies and



518 COUNTRY EXPERIENCE WITH ADJUSTMENT LENDING IN THE 1980s
can limit their effectiveness within a relatively short time. It is precisely
in this context that the approach of the latter half of the 1980s has become
untenable. We at the World Bank have assumed a total smoothness of
the adjustment process, a total synchronization of foreign assistance and
domestic adjustment. We have assumed that the frictions in the process
of adjustment were going to be very limited and that we could go on
increasing the debt of these countries as adjustment proceeded. That
approach clearly did not work. It is to be hoped that a meaningful process
of reduction of debt service burdens-and I purposely use the term "debt
service"-will provide governments with greater maneuverability in the
very difficult task of realigning the domestic benefits and costs of the
economic policy.
We also need to look more closely at the issue of poverty and the social
consequences of adjustment. We learned a good deal in the 1980s about
the issues of adjustment, but not enough about the process of adjustment
and the issues of transition. This book discusses some of these issues,
such as how, in the process of adjustment, governments can and should
institute measures and programs that cushion the less-privileged and vul-
nerable groups. A great deal more work needs to be done in this area if
the process of adjustment is going to work in the Latin American coun-
tries. This is particularly the case in countries that do not have author-
itarian regimes and where people therefore will feel more free to come
out onto the streets to protest. If the problem of the less privileged and
the poor is not addressed, such protests may contribute to the disruption
of the entire adjustment process.
That brings me to the role of the World Bank. Although we have
worked hard on adjustment, success has been greatest in cases in which
we have essentially supported and complemented the commitments of
the government and its own efforts. By and large, in countries where that
commitment has not been present, success has been marginal at best,
despite major efforts. And that simply means that more of the World
Bank's efforts need to go into countries where that fundamental com-
mitment and the proper conditions exist.
We in the World Bank and the international community will have to
show a great deal more patience with this process. The factors that led
to the present malaise developed over twenty or thirty years as a result
of the economic policies in these countries and, in some cases, the policies
of international lenders. Consider, for example, the case of Colombia
and Honduras, which have massive overinvestment in electric power.
This investment was externally financed and has contributed to high debt
service obligations. Most of this investment was financed by the World
Bank and the Inter-American Development Bank. So responsibility is not
one-sided; commercial lenders and to a certain extent official lenders



Mexico: Adjustment and Stabilization 519
such as the World Bank are also responsible for the economic problems
in many of these countries.
Essentially, this means that adjustment will take time. In many cases
there will be two steps forward and one step backward, and we will have
to look for the second- and third-best solutions.






PART IV
Assessment






23 Lessons of Experience and the
Future of Adjustment Lending
THIS CHAPrER uses the panel format to bring together a group of distin-
guished economists and practitioners from the World Bank, the Inter-
national Monetary Fund, universities, and the governments of adjustment
loan recipient countries. They were asked to draw lessons from the decade
of experience with adjustment lending that is described in this book. Their
views, comments, and criticisms and their forecasts for the future of
adjustment lending are presented here.
Joe Abbey
The real issue here is the search for a workable development strategy,
and adjustment lending must be viewed as part of that strategy. The
question of whether adjustment lending or structural adjustment should
be considered a medium- to long-term effort really is a nonissue for many
countries because their development effort is necessarily a long-term pro-
cess. Although this question becomes more pressing at times of crisis, it
must not detract developing countries from their major concern, which
is to get on with the process of economic development.
What is important, especially for sustainability, is how adjustment
programs are formulated and the role of each of the players-govern-
ments, the World Bank, and the International Monetary Fund (IMF)-in
that process. Governments are confronted by many serious tradeoffs.
First are the tradeoffs between the short term and the long term, and
quite often governments may be tempted to take the short-run benefits
at the expense of long-term gains. Second are the technical tradeoffs and
legitimate policy conflicts, such as fiscal prudence versus trade liberali-
zation. A third class of problems facing governments is the result of
conflicts in policymaking and advice between the IMF and the World
Bank. Those caught in the middle are not amused by such disagreements,
particularly when there are legitimate policy conflicts to be resolved.
A look at the totality of difficulties facing a country about to embark
on a program of adjustment-institutional rigidities, market weaknesses,
administrative and managerial weaknesses-brings certain issues into
523



524 ASSESSMENT
clearer focus. It is particularly clear that unless the government "owns"
the program, in the sense that the political tradeoffs have been taken care
of, the program cannot work. When that condition is lacking, the World
Bank might believe that a dialogue is taking place, but in fact what is
happening is that a country that needs money-very badly, too-is being
forced to take a few measures that it does not understand and that the
World Bank itself is not really confident about. As this book points out,
the countries that have done well are those that had their own programs
before imF or World Bank involvement or those that had the technical
capabilities to take these programs and adapt them to their needs. In this
regard, there seems to be some argument about whose program the Ghan-
aian Economic Recovery Program is. Some of us in Ghana, of course,
think the program is ours, but frequently it is referred to as the ESAF
Program or Ghana SAL I or SAL 11.
Nevertheless, the World Bank and the IMF do have a critical role in
the adjustment process: to force governments to think through what they
plan to do and to understand all the ramifications of their programs.
Frequently, the necessary markets or other institutions are not there, or
they do not work well or as expected. Failure to take into account in-
stitutional characteristics specific to the country can lead to poor advice.
Advice must be grounded in thorough and detailed analysis; the blanket
recommendation of generally accepted policies can be disastrous in par-
ticular circumstances. Financial deregulation, for example, if imple-
mented under unfavorable conditions, can actually lead to increased fi-
nancial instability. This was the case in Ghana in 1987, when poor advice
concerning financial deregulation was imposed on the country. The ex-
pectation was that as soon as financial deregulation took place, interest
rates on deposits would go up. It was clear to us in Ghana, however,
that the absence of nonbank financial intermediaries and of other debt
instruments made it unlikely that deregulating interests rates would lead
to the expected results. And as events unfolded, the banks in Ghana took
advantage of deregulation, widened their margins, and are now paying
very low rates of interest to depositors and even turning depositors away.
It is not surprising that interest rates are low. The institutional features
of Ghana's financial sector were not considered carefully by our inter-
national advisers, and now they tell us we must raise interest rates. This
is the outcome of inadequate knowledge and poor advice.
Other problems arise because of a narrow focus on aggregates or on
only part of the picture, without regard for the reality behind the data.
Thus, for example, after the 1987 conference of Ghana's external lenders,
there was concern that the country was walking into a death trap because
it was facing a 70 percent debt service burden. But everyone failed to
consider the net flows. The debt service ratio measures only gross out-
flow; it says nothing about net inflow or net outflow. Now, however,



Lessons of Experience and the Future of Adjustment Lending 525
the other side of the picture is finally being considered. Everybody is
talking about the large resources that are being given to the Ghanaian
program, and nobody is talking about the debt. Between 1983 and 1988
Ghana received from all sources US$3.6 billion and paid out (to clear
arrears in interest and principal) a little over US$2.6 billion. So the net
inflow to Ghana was about US$1 billion. This type of thing is missed
when people fail to look behind the aggregates.
Another example concerns cuts in social services or social expenditures
that accompany adjustment. These aggregates mean nothing unless one
takes account of the quality of the expenditures that are being cut and
those that remain. Thus, for example, total government expenditure in
Chile went down by 5 percent in real terms and yet the quality of life
did not deteriorate; in fact, there was an improvement.
When the social costs of adjustment are being analyzed, other inter-
vening factors must be considered as well as the adjustment program.
For example, in 1982-83 Ghana had the worst drought in its recorded
history, and about a million Ghanaians returned home from abroad. We
decided to launch the Economic Recovery Program at the height of the
drought because we believed it would be wrong to wait for the drought
to abate and then start a program that would require massive increases
in imported items. The program would then have been more difficult to
manage and would have provoked cynicism. Therefore, we had to move
in the middle of the drought. When malnutrition and high infant mor-
tality rates were reported in rural areas in 1983, the blame was put on
the program, which had not even started until 1983, instead of on the
drought. Of course, if the Ghanaian recovery program was responsible
for the malnutrition in 1983-84, it also caused one of the quickest turn-
abouts in history. As we had predicted, when the rains returned, when
the incentives began to work and food production picked up, infant mor-
tality rates improved and dietary problems were eased. But the wrong
story had gone out already, and many people believed that Ghana had
experienced massive social costs of adjustment.
The only way to go forward is for governments to clearly state their
objectives and how they intend to achieve them. Because these things are
often not clearly articulated, either by the government or by the inter-
national organizations involved, the structural benchmarks are rarely
achieved. I believe that the World Bank and IMF can best fulfill their roles
not by using their money as leverage, but by forcing the necessary anal-
ysis. By compelling governments to spell out how they expect to achieve,
if not their ultimate goals, at least their intermediate goals, the World
Bank and the IMF can help policymakers avoid the pitfalls experienced
by others who blindly tried to follow set paths and came to grief. This
imposition of discipline is the more important part of the dialogue pro-
cess, and many other practices that pass for dialogue are counterpro-



526 ASSESSMENT
ductive. Once we know the groundwork is firm, we can begin to worry
about sustainability.
Stanley Fischer
The evaluation of adjustment lending is not only extremely difficult, but
also essential. None of the methods of evaluation are entirely satisfactory.
First, there are statistical tests that look at countries with and without
adjustment lending. Although these tests provide some useful informa-
tion, they are vulnerable to criticism from those who look at each ob-
servation and say "yes, but Korea did this," "yes, but Mali did that"-
but that is true of every regression ever done. We work with statistical
methods because they provide useful data for evaluation. The second
method is case studies, and the information they provide is also useful.
Third are econometric models, which are themselves problematic in many
ways but do provide information. Fourth is a special group of case studies
prepared by the Operations Evaluation Department (OED) in the World
Bank. It is surprising how little these are used. That may be because they
tend to come in very late and are very long.
In evaluating adjustment, we have to pull the evidence together from
wherever we can get it. We are not going to find definitive evidence on
the success of adjustment lending or, for that matter, on anything in
economics. What caused the Great Depression? We do not know with
complete certainty. Why did the United States start to grow in 1982?
Was it supply-side economics, was it good old-fashioned Keynesian ex-
pansion? We do not have a definitive answer. We look at the evidence,
make as strong an argument as we can, and then we have to proceed on
that basis. As Arjun Sengupta of the IMF emphasizes, the fact that we do
not know does not excuse us from making decisions. Our obligation is
to evaluate the evidence as best we can, listen to all the arguments, and
then make a decision about the actions we want to take, taking into
account that our knowledge is imperfect and that we should remain flex-
ible and always on the lookout for new evidence. That is the best that
we can hope for.
Of course, there are people who prefer not to look at any evidence.
Frank Paish, the British economist and forecaster, said it was very dan-
gerous to look at our past forecasts because that made us lose our nerve.
But we have to look back in order to reach decisions that we are willing
to stand by on the basis of the best evidence we can find.
The problems of adjustment lending fall into three categories and are
related to adjustment in the countries themselves, relations between the
World Bank and the countries receiving adjustment loans, and internal
World Bank controls that affect adjustment lending.



Lessons of Experience and the Future of Adjustment Lending 527
In looking at problems of adjustment from the country's perspective,
it is important to emphasize both short-term macroeconomic adjustment
and long-term structural adjustment involving development strategy. An
adequate macroeconomic framework is absolutely essential. That is clear
from the many cases in which the macroeconomic framework has not
been appropriate. I would single out-in addition to the fiscal problem
that has already been emphasized-the problem of extremely high in-
flation rates. No program in which an inflation rate of 600 percent is
expected over the next nine months can conceivably be sound, and yet
such programs have occasionally been certified as acceptable.
These are long-term and extremely complicated problems. The eco-
nomic transformations now being attempted in adjustment programs
took decades and longer to achieve in the industrial economies. The ten-
dency is to introduce extremely sophisticated financial and governmental
structures into countries that probably cannot support them. We should
concentrate on fundamentals, however; the countries that require ad-
justment loans have major distortions and major gaps, and we had better
concentrate on correcting them. The refinements-the seventy-ninth con-
dition in an adjustment loan that at best would bring only an epsilon of
improvement-are way beyond what is known and what should be im-
posed on a country that may well lack the capacity to carry it out.
Adjustment is a continuing process. It is still taking place in the in-
dustrial economies, and it will take a very long time in the developing
economies. Our obligation is to differentiate between what is essential
and what is a luxury at a particular stage of development. Joe Abbey
emphasized the importance of thinking policies through and looking
ahead, and both the World Bank and the IMF can help with this kind of
rigorous analysis in their interactions with member countries.
Of the many complications that can arise in adjustment lending, those
involving the financial sector, foreign exchange controls, and so on
should cause particular concern. An enormous amount of information
is needed to deal with these delicate systems; there are matters of trust
and matters of prudential regulation. Even many highly developed coun-
tries have had great difficulty with their financial systems, and in many
developing countries the problems in this sector are clear: massive bank-
ruptcies, interest rates completely out of line, and so forth. These things
have to be fixed. When we begin to introduce ultrasophisticated financial
instruments, however, I worry whether that is what we ought to be doing.
With respect to the mutual problems between the countries receiving
adjustment loans and the World Bank, the crucial issue is "ownership"
of the adjustment program. For some countries, ownership is virtually
all that is needed. If a country "owns" a program and has the admin-
istrative capability necessary to design and implement it, we at the World
Bank do not have to do very much. We can discuss the program with



528 ASSESSMENT
the country and see it get under way, but then we should relax. If a
country has the commitment but does not have the administrative ca-
pacity, we can take a more active role. In those cases, slightly more
detailed adjustment lending and greater specification of conditions may
be useful.
Ernest Stern in the opening chapter touched on problems related to
internal World Bank controls when he said that adjustment lending is
addictive both to governments and to the World Bank. One could also
say that adjustment lending is seductive-its allure lies in being close to
the powerful, in being able to say "I spoke to the president" or "I was
assured by the finance minister." Because it is seductive, we need strong
procedures, rules within the World Bank delineating the conditions under
which adjustment lending will take place, such as assurances that the
macroeconomic framework is right. In particular, we need to ensure that
the patterns of disbursement will not cause trouble down the road. We
do have to control adjustment lending. Some people now believe that
World Bank staff cannot even talk to a country unless they give them a
great deal of money up front. That is wrong. It is up to the World Bank
to perform the economic and sector work and provide the technical advice
that will make project lending more attractive to its members and its
staff.
Internally, the problem is how to enforce our own controls on ad-
justment lending. There is pressure to lend from within the World Bank.
There is no denying that; indeed there should be pressure to lend. We
just asked our member governments for US$75 billion because we have
a task to do. The pressure to lend comes not from a belief that we should
shovel out money regardless of its use, but rather from an implicit con-
tract by World Bank management to deliver loans of a given quality in
given amounts. The pressure to lend is the pressure to go out, to find
the opportunities, and to set up the good loans.
The World Bank's internal procedures do not encourage simplicity in
lending, however. Consider, for instance, the system for reviewing loans
internally. Loan documents are distributed for comment, and it is in-
cumbent on everyone to add something, to show that he or she has
contributed. Similarly, in the Operations Committee, people pick up on
a point here and a point there. By an inevitable process of accretion, the
loans become more and more complicated as a result of the World Bank's
review process. I do not know how to stop that process, but we have to
think about improving it.
One change in internal procedures for which there are strong argu-
ments is from multi-tranche to single-tranche operations. When a country
has taken ownership of a program and we are confident, as we are, say,
in the case of Indonesia, that the government will deliver what it has
promised, then I cannot see the objection to single-tranche lending. These



Lessons of Experience and the Future of Adjustment Lending 529
are not one-shot operations; they involve a sequence of operations. Sin-
gle-tranche lending helps to clarify the ownership issue and it helps to
keep loans simple.
The balance of payments link is another important internal issue. We
need to maintain that link both for purposes of internal control and to
ensure that we do not unnecessarily increase a country's indebtedness.
But at the same time we should find a way of undertaking long-term
policy lending: providing money in small amounts over longer periods
for adjustment, including institutional development, is absolutely crucial
to the development effort. We played with that concept in the 1988 report
on adjustment lending but could not come up with a satisfactory instru-
ment. We should be using our ingenuity to create some type of loan that
disburses less rapidly than adjustment loans but that is nonetheless di-
rected at adjustment. Such a loan would give countries that have un-
dertaken a serious program of institutional and other development the
assurance that the World Bank would be there supporting a long-term
adjustment process.
With regard to the issue of World Bank-IMF coordination, we have
reached a very important new agreement with the IMF. We will no longer
try to sort out any macroeconomic disagreements with the IMF in com-
peting negotiations with the debtor countries, but will discuss these dis-
agreements with the IMF. The new agreement recognizes that countries
benefit from the different professional opinions offered through the two
institutions and requires the institutions to discuss these matters. When
there are genuine differences between the World Bank and the IMF con-
cerning macroeconomics from the viewpoint of development, it is ex-
tremely important to discuss those issues. We have to work out methods
for doing that.
In addition, the agreement gives the World Bank primary responsibility
for structural adjustment. From the perspective of development, struc-
tural adjustment policies are the crucial ones to deal with. The macro-
economics framework is important, and we have to be sure it is right and
that changes do not violate development priorities. But it is the primary
responsibility of the World Bank to tackle development issues through
the adjustment process. We are going to have to concentrate on this and
improve the quality of our analysis. That is the challenge-a crucial
one-that we must meet.
Manuel Guitian
A striking feature of the process of adjustment in the 1980s is the extent
to which it was dominated by the events of the 1970s. Even more striking
is that the factors that complicated the environment in the 1980s included
not only the problems but also the solutions from the 1970s. For brevity's



530 ASSESSMENT
sake I will focus on global developments without discussing the role of
specific policies pursued by individual countries. I cannot overstress the
importance of this caveat because, in many respects, the global trends
that I will outline are the joint results of all those policies.
Put briefly, the 1970s were characterized by events such as the sharp
changes in terms of trade and corresponding shifts in current account
imbalances, by problems such as high inflation and faltering growth, and
by solutions such as recycling and large capital flows, particularly from
commercial banks to sovereign borrowers. In many respects the 1970s
witnessed a revolt against the growing interdependence toward which
the international economy had been struggling during the 1950s and
1960s. This revolt was made evident not only by the surge of protectionist
pressures but also by the move to flexibility in exchange rates in search
of the widely heralded degree of freedom that would give autonomy to
national economic policies-a goal that, in the end, proved to be most
elusive. But another factor tended to counter these trends: growing capital
movements provided cohesion (at least for some time) to the international
financial system. The international lending-borrowing process was there-
fore perceived as part of the solution to the period's difficulties.
The start of the 1980s provided an excellent illustration of the inter-
dependence of economic phenomena, in particular the ease with which
the correction of one imbalance can thwart the solution adopted for some
other imbalance.
In the industrial world, inflation in the 1970s shifted the pendulum
toward policies to bringing it promptly under control. The resolution of
the inflation issue was accompanied by the emergence of high real interest
rates and a slowdown in economic activity. A by-product of this process
was that the decisions made in borrowing countries when real interest
rates were negative now became untenable, and the consequences of con-
tinuing to follow them soon surfaced: concerns about creditworthiness
arose among lenders, and the flow of international loans and capital dried
up abruptly.
In effect, the solution of the inflation problem was accompanied by
events that brought about the debt service difficulties. In other words,
what had been perceived as a solution in the 1970s (recycling) became
a problem of the 1980s (debt). Thus a first conclusion is that a most
urgent concern must be to ensure that the solutions devised for the dif-
ficulties of the 1980s do not become the problems for the 1990s.
Adjustment in the 1980s was aimed at resolving the debt problem,
possibly one of the most elusive economic objectives sought in this dec-
ade. By now the main stages of the solution are familiar: initially, the
pursuit of a debt strategy based on the assurance of external financial
support for domestic adjustment efforts; subsequently, reinforcement of
the strategy by the addition of the growth imperative-growth to be



Lessons of Experience and the Future of Adjustment Lending 531
achieved through structural reforms and continued financial assistance
in a variety of forms, including menus of financing modalities; and now,
the addition to the financing menus of schemes for reducing debt and
debt service, which are supported by international institutions and made
available to countries undertaking a strong adjustment effort.
Each stage provided evidence of the difficulties that led to successive
adaptations of the strategy. At the outset, improvements in the balance
of payments and economic recovery in debtor countries were slower than
expected; thus available resources were severely constrained and there
was a need for increased domestic efficiency and continued external sup-
port to ensure growth. The difficulty in restoring lending and capital
flows led, in turn, to menus of financing options and then to interna-
tionally supported debt-reduction proposals.
There was scope for dissent at each point, and dissent, of course, sur-
faced on each occasion. Questions arose concerning the mix and se-
quencing of adjustment and financing: Was the blend adequate? Which
should come first, adjustment or financing? Similar dilemmas surfaced
at later stages as perceptions differed (and will continue to differ) about
the appropriateness of an adjustment effort, a financing package, a proper
rate of growth, or-looking ahead-the proper scope of debt-reduction
packages, to name just a few. In the process, however, there is a risk that
attention may be diverted from what is possibly the most fundamental
question on this front: What are the conditions necessary for the res-
toration of normality to capital flows? This question brings up another
urgent concern for the 1990s: the need to ensure that both the level and
the structure of international capital flows freely conform to a market-
determined pattern of economic incentives.
The policy issues that will need to be addressed in the years ahead are
numerous. For the developing countries, the preceding chapters have
examined areas of policy reform and country experiences, and there is
little that I can add to these analyses. I will refer to the industrial countries
later in the context of the role of international institutions.
But certain general principles and areas are worth stressing with regard
to policy issues:
* The complementarity between macroeconomic and structural policies
and the importance of consistency among them as well as persistence
in their implementation
* The relevance of the varying time horizons in policy formulation and
the consequent need to ensure consistency between short- and long-
run aims and to take account of the different lags in the effects of
policies
* The search for operational criteria for the implementation of structural
reforms; for example, their scope (how widespread should they be?),



532  ASSESSMENT
their pace (how fast should they proceed?), and sequence (which one
should go first?).
As in the past, international institutions will continue to have an es-
sential role in the coming decade. They are uniquely placed to provide
a genuinely international perspective on issues confronting the world
economy. For them to do so appropriately and effectively will require a
concern with the system as a whole in order to safeguard common ob-
jectives and gauge the responsibility of the parts. A necessary means to
these various ends is a commonly agreed code of conduct, which of course
exists in the form of the Articles of Agreement of the IMF and of the
World Bank (and the General Agreement on Tariffs and Trade as well).
These institutions have a responsibility to demonstrate persistently that
the opportunities opened by economic interdependence far outweigh the
constraints-a responsibility that calls for a careful balance of interna-
tional and national considerations. Good economic analysis will help,
but also needed is the ability to promptly spot shifts in that balance over
time, because such shifts do occur-as happened, for example, in the
context of the debt strategy.
Given my affiliation with the IMF, I will focus on the very basic chal-
lenge lying ahead for this institution. The challenge calls for yet another
delicate balance, one between its jurisdictional or regulatory activities
and its financial assistance operations-that is, the balance between sur-
veillance and conditionality. The IMF has to ensure coordination between
surveillance and conditionality, so to speak.
Surveillance is needed to ensure a favorable and stable global economic
environment. The linkages between the whole and the parts have hardly
ever been clearer than in recent years. For example, the world economic
setting has been characterized by historically high real interest rates for
some time; it seems clear that their return to normal levels will be as
important as, if not more important than, some of the schemes under
consideration for the resolution of debt problems. To put it more gen-
erally, individual countries can be expected to pursue specific domestic
policies only if similar behavior is pursued elsewhere. Thus a challenge
for the IMF, in addition to the many others that it will have to meet both
by itself and cooperatively with the World Bank, will be to bring sur-
veillance back onto center stage along with the more visible condition-
ality.
The 1990s are likely to be an eventful decade, and we must seek to
ensure that its legacy to the next century is one of sustainable solutions
rather than recurrent problems.
These are some of the issues that lie ahead:
* Resolving industrial countries' payments imbalances. This will require
establishing the respective responsibilities of the deficit and surplus
countries, which are by no means symmetrical.



Lessons of Experience and the Future of Adjustment Lending 533
* Restoring balance to the composition of demand (and output) in sev-
eral dimensions: consumption and investment, public and private, ex-
ternal and internal.
* Determining the role of markets in the process.
* Ensuring continued effectiveness of the debt strategy.
* Returning capital markets to normal.
* Ensuring growth.
The experiences of the 1980s confirm the importance of consistency
in policies and persistence in their implementation. Perhaps the main
lesson is the limitations of econromic policy when it stretches far beyond
the setting of a predictable framework for economic decisions.
Gerald Helleiner
Perhaps an appropriate starting point is the observation that we do not
have total agreement concerning precisely what "adjustment lending" is.
Ernest Stern, in chapter 1, said that structural adjustment "ought to be
anchored in balance of payments problems"; others have argued that it
is a matter not of macroeconomic variables but of (microeconomic) re-
source allocation and the institutions and decisionmaking systems that
determine it. If we do not agree on the ultimate meaning and purpose
of adjustment loans it should not be surprising if we also cannot agree
on conclusions about either the key issues or the record.
There can be little doubt that whatever the original balance of pay-
ments-related origins of adjustment loans, they quickly acquired all man-
ner of other elements. Everyone who had a view about what was crucial
for growth in a particular country rapidly joined the effort to influence
local policies through program lending. A veritable Christmas tree of
structural policies began to be advocated, some with more obvious bal-
ance of payments implications than others. (Obviously, structural re-
forms frequently do have balance of payments implications, and mac-
roeconomic policies motivated by balance of payments difficulties can
and often do have structural implications.) Adjustment lending rapidly
became a euphemism for good old-fashioned policy-based program lend-
ing of the kind long debated in the development literature and aid agen-
cies.
How has adjustment lending actually worked? If there is one conclu-
sion that has dominated the discussion here it is that everything takes
longer than anticipated. Structural adjustment simply does not happen
through short bursts of intensive policy change. The message to those
undertaking structural policy reforms seem to be (1) be patient and (2)
have faith. They are being told that while there is as yet little firm evidence
of the developmental impact of these reforms, the reforms are nonetheless



534 ASSESSMENT
probably the correct ones. This advice to adjusters is strikingly similar
to that typically proffered by the International Monetary Fund. But the
IMF is probably on stronger ground than the World Bank because the IMF
promises less from its recommended policy changes: stabilization rather
than development.
What is the evidence that adjustment lending and the principal longer-
run reforms it has promoted have contributed to economic growth and
development? Granting all the methodological problems-arriving at a
suitable counterfactual case, possible sample-selectivity bias, and the
like-there is nonetheless evidence of modest improvement in overall
economic performance in countries undertaking World Bank-supported
adjustment programs. But the evidence is fairly thin. And there are some
disconcerting details in the record. Particularly deserving of further ex-
ploration is the finding that the intensity of World Bank and IMF lending
is negatively and statistically significantly related to the investment rate.
This relationship seems to place a very high premium on the increased
efficiency that structural reforms are supposed to achieve. Even more
provocative is the strong evidence, presented in more than one chapter,
that increased imports are strongly associated with improved economic
growth and the implication that increased utilization of capacity, rather
than any effects of investment, dominates medium-term performance.
There is also evidence elsewhere that the stability of import volume is a
powerful predictor of recent growth performance in developing countries.
A new mood of caution, humility, and uncertainty characterize this
book (as well as the 1988 summary report, Adjustment Lending: An
Evaluation of Ten Years of Experience). It was found in discussions of
all the major areas of adjustment policy.
In the sphere of trade, the record indicates greater success in encour-
aging exports than in liberalizing import policy. Much of the export
success was dependent, however, on earlier investments with a long ges-
tation period, as in Chile and Turkey, for example; only in Mexico (and
in the return of Ghanaian exports from smuggling) was there rapid
switching purely in response to current price incentives. Except for ex-
porters' inputs, import liberalization efforts have been fairly weak and
fairly modest in their effects. One is left yearning for more analysis of
such issues as the effects of different kinds of import liberalization, dif-
ferences in its timing, differing degrees of dispersion of trade incentives,
and differences in the stability of trade regimes. As is usual in recent
discussions of trade policy, there was also some handwringing over the
mysterious successes of the heterodox Republic of Korea.
Financial reforms have typically also been weak and, where initially
more forceful, have virtually always been subsequently reversed. It seems
that financial liberalization has frequently been counterproductive when
pushed too fast in the absence of macroeconomic stability, adequate pru-



Lessons of Experience and the Future of Adjustment Lending 535
dential regulatory mechanisms, or at least minimal competition in the
financial sector. It is striking that no systematic assessment appears to
have been made of the impact of interest rate reforms undertaken in
World Bank-IMF programs.
Despite the continuing frequency of advice on the desirability of pri-
vatizing public enterprises, the impact of privatization has yet to be sys-
tematically studied with the use of the traditional economists' criteria of
social efficiency. The summary record suggests that premature efforts to
privatize in the absence of market capacities or adequate official prepa-
ration have been costly and that privatization deadlines have often proven
counterproductive.
Fundamental tax reforms have been few and far between. None has
as yet been attributable to adjustment lending.
Chapter 7 on agricultural policy emphasizes the importance of long-
term perspectives and of nonprice considerations, particularly in the case
of public investment. Price reforms, it argues, do not accelerate growth,
whatever improvements in efficiency they may achieve, and they are easily
reversed. Price stabilization schemes are likely to remain in place, and
rather than waste effort in vainly seeking their elimination, we should
work to improve them.
There is some evidence that value added in industry grew a little faster
in countries receiving adjustment loans than in others, but the link with
conditionality is obscure. The author of the industrial policy study (chap-
ter 8) was reluctant to draw firm conclusions from such scanty data.
The social costs of adjustment and considerations of distributional
equity seem to have been universally neglected in World Bank-supported
adjustment programs. Where distributional outcomes were relatively be-
nign, they were accidental. Programs to defend the poor may be very
difficult to engineer, but the fact is that they have nowhere been seriously
tried. Such programs as Ghana's Program of Actions to Mitigate the
Social Costs of Adjustment were afterthoughts, added on after the pro-
gram was launched. The question at issue in the policy discussions was
not the desirability (which seems affirmed) so much as the efficacy of
targeting measures for the poor within adjustment programs.
In view of current environmental concerns, I note that nowhere in any
of the adjustment programs reviewed here was there any specific reference
to the environment or to environmental sustainability. There were only
glancing references to these issues.
What policy conclusions for the World Bank can one draw from this
very mixed and uncertain record of the impact of adjustment reforms
and the Bank's own adjustment lending? I would suggest at least the
following six.
1. Since structural change and development are long-run processes and
supply responses in developing countries are everywhere slow, one should



536 ASSESSMENT
not expect quick results-even from fast-disbursing loans to governments
that are doing everything "right." Trying to move too quickly can be
counterproductive. That some policies can be changed faster than others
does not mean that they necessarily should be. And overloading an ad-
justment program can similarly be unhelpful. What is required from out-
siders is long-run vision and commitment, steady financial support, and
ongoing supportive policy dialogue. The sustainability of change depends
upon a stable incentive structure and stable politics as well as stable
financing. But stable financing can go a long way.
2. It is perverse to rush to judgment on relative success or failure. It
is certainly premature, for instance, to proclaim, as does a recent World
Bank document, that in Africa "recovery has begun." Larry Westphal
of Swarthmore College has offered elsewhere the dictum-henceforth no
doubt to be known as Westphal's Law-that those who know the least
about a country are the most prone to seize upon its short-term record
of overall performance as an indicator of longer-run success.
3. It is therefore also perverse to impose too many conditions in the
short run, conditions for which failure brings cessation of external sup-
port. In the face of our uncertainty about the development consequences
of so many kinds of policy change, the only possible criterion for dis-
bursing short-term adjustment loan tranches is whether the country did
or did not undertake the promised policy changes. We are apparently
unable to tell in the short to medium run whether they worked. But is
this not a rather tenuous and inappropriate basis for a decision to drop
support? By imposing such short-run performance requirements, we risk
aborting the flow of external support so necessary for the development
process over matters of dubious developmental relevance. Conditions for
tranche release must be for a longer-term cycle and should relate to policy
outcomes rather than promises.
4. Policy conditionality should not impose unreasonably demanding
short-term structural shocks on fragile economies and politics. In view
of the considerable uncertainty concerning the effects of specific policy
reforms on the economy, political stability, and so on, it may be sensible
to prescribe programs of reform that, though less than optimal, are not
too disruptive and are both resilient to unforeseen shocks and amenable
to change with minimal costs. The skills of the economist are, after all,
best employed to reduce the prospect of egregious error, not to attain
policy perfection. But once the most grotesque distortions, particularly
those related to the real exchange rate and the fiscal deficit, are repaired,
further fine-tuning of prices and policies may not be very productive.
Straining toward further policy reforms risks disruption of the stable
investment flows on which development depends.
5. It is striking that among the case studies presented in this volume,
the countries that launched the strongest adjustment programs-Chile,



Lessons of Experience and the Future of Adjustment Lending 537
Korea, Mexico, and Turkey-undertook them out of their own convic-
tions. The World Bank does not seem to have persuaded them of much.
Sometimes their own convictions were more orthodox than those of the
World Bank! The World Bank and the IMF nevertheless were able to
provide assistance with technical specifics and contribute to ongoing pol-
icy debates. It would be no small achievement for the World Bank simply
to maintain everywhere an open dialogue on development policy, to pro-
vide specific technical assistance on matters of detail where requested,
and gradually to improve the quality of economic analysis both in the
borrowing countries and in the World Bank itself. If these functions were
all performed well, the World Bank would have quite enough to be proud
of. Its staff should not seek or expect to win all battles. Policy discussions
with adjusting countries should truly be dialogues, not a process akin to
negotiation in a bazaar.
6. Once gross macroeconomic distortions are overcome, the risk of a
program's failing because of underfunding seems at least as great as the
risk of failing because of not undertaking recommended policy reforms.
The contrast between underfunded Zambia, which, for a while, did all
it was asked and still failed, and Ghana, which this year received gross
commitments of US$800 million (its population is under 15 million),
could hardly be more striking. One cannot help returning to the strong
evidence cited earlier regarding the role of imports in growth. What, after
all, do we ultimately know about optimal policies for growth? We do
know-now better than ever-that more resources help. No effort
should be spared to raise them. Since total resources are limited, however,
difficult decisions will remain about which countries to support with
"adequate" funding.
The debate on adjustment lending has brought development econo-
mists back full circle to where they were about twenty-five years ago.
Once again they are debating the relative roles of program and project
assistance. They now emphasize macroeconomic management rather
more than most of them did in those earlier times. In other respects,
however, their knowledge has not progressed as much as one might have
hoped.
Perhaps the new humility and uncertainty about broad structural re-
forms will drive more of the World Bank's program lending into less
ambitious sectoral loans in the 1990s. And perhaps this move would
have the desirable side effect of reducing some of the World Bank's con-
flict with the IMF. To my mind, such developments would be all to the
good.
In any case, let me congratulate the World Bank on its new, more
nuanced, more careful, and more objective approaches to the analysis of
its policy-based lending activities. I am confident that I speak for the



538 ASSESSMENT
entire development community when I express strong support for this
more open intellectual environment.
Pedro S. Malan
After nearly a decade of adjustment lending-about 160 loans to about
60 developing countries-it is time to consider the lessons learned. Else-
where in this book the World Bank's experience in trade policy is de-
scribed as "a limited success." The Turkish experience is considered "un-
finished business" after five structural adjustment loans, three sectoral
adjustment loans, and three standby agreements. The Pakistani case is
called a "semisuccess." It is noted that the World Bank underestimated
the political difficulties involved in adjustment lending and the protracted
nature of the deep-rooted development problems facing too many de-
veloping countries in the 1980s. "Mild relative improvements" are seen
in countries undertaking adjustment programs with the support of the
World Bank and the International Monetary Fund (IMF).
Apparently, the first lesson to be drawn from experience is that the
record is a mixed one, which compounds the uncertainties about the
future of adjustment lending. There is no point yet in trying to reach a
verdict based on the evidence at hand. We need both more studies of
individual countries-their successes and failures-and more compar-
ative analysis that cuts across a wide range of experiences.
In evaluating the impacts of adjustment, I would like to organize my
remarks around four clusters of observations. The first one is related to
the sustainability of adjustment programs. The importance of political
commitment to structural reforms by the authorities of the adjusting
country has been strongly emphasized by creditor countries and the mul-
tilateral institutions they control. This emphasis seems appropriate, but
relatively less attention has been given to the importance of public support
for policy reform if adjustment efforts are to be sustained. In fact, the
political commitment of the authorities and the support of public opinion
are not unrelated, and both will be progressively harder to maintain the
longer adjustment continues without payoffs in terms of growth and the
longer human and social conditions are allowed to deteriorate.
The concern with the social costs of adjustment-or of delayed ad-
justment-is as legitimate as the concern with protecting the poor during
adjustment and reducing poverty over the long term. The argument is
often couched in terms of the need to ensure broadly based political
support for adjustment. What often derails adjustment programs, how-
ever, is not the outcry of the vulnerable poor but the opposition of the
vocal special-interest groups that are bound to be affected by reforms in
foreign exchange, subsidies, tax benefits, and other areas that affect rent-
seeking activities. The benefits of adjustment programs are often per-



Lessons of Experience and the Future of Adjustment Lending 539
ceived to be long term and diffuse, while the losses are shorter term and
concentrated. In this context, the vocal have an edge over the vulnerable.
Poverty alleviation is a legitimate concern, but to justify it as a means
of recruiting the political support of the poor to ensure the sustainability
of adjustment is to weaken the argument. This long-term objective should
stand on its own.
One last point on sustainability: a consensus seems to be emerging
that primary responsibility for the conception of structural adjustment
programs must lie with the national authorities that will implement and
sustain the program. Although the advice of the World Bank and the IMF
may be welcome, the national authorities must be satisfied that the policy
actions and reforms represent an appropriate, realistic, and feasible
course of action.
The second cluster of issues is related to design. Improving the design
of macroeconomic and structural adjustment policies could help to es-
tablish credibility for the policies at an early stage. But credibility cannot
be established unless the program design takes full account of the coun-
try's legacy of real or perceived structural weakness. These are frequently
deep-rooted structural problems, such as inadequate infrastructure,
poorly developed markets, institutional and managerial weaknesses in
the public and private sectors, lack of critical skills, and an undiversified
productive base with limited potential for a quick supply response. The
macroeconomic and structural adjustment policies will not be effective
unless these economic and social realities are addressed in the conditions
and performance criteria for the operation. These conditions should be
focused, capable of being monitored, few in number, properly sequenced,
and concentrated on variables that are under the effective control of the
authorities of the country in question.
The concern with social costs and poverty reduction in the design of
adjustment programs could well be summarized in a few propositions.
One, without growth and development there is no hope of reducing pov-
erty and social costs. Two, the trickle-down process is not enough to
make even a dent in the problem of poverty; specific measures directed
to the poor are needed, even though identifying the poor may be difficult.
And three, poverty-reduction objectives should be integrated in the design
of adjustment programs. But there are clearly tradeoffs in trying to put
rationality into the quest for equity and humanity into the quest for
efficiency. As noted by Arthur Okun, this inescapable conflict may be
the reason capitalism and democracy need each other.
A final issue related to design concerns the need for contingency mech-
anisms in adjustment programs because of the enormous uncertainties
they face. Such mechanisms are essential if the programs are to be sus-
tained and perceived to be sustainable over the relevant medium term.



540 ASSESSMENT
The third cluster of observations is related to the macroeconomics of
debt and its implications for the effectiveness of design and for the sus-
tainability of adjustment programs. As recognized explicitly by the man-
aging director of the IMF, "the debt overhang has interfered with the
sustained implementation of adjustment policies." Steps to accelerate the
reduction of debt and debt-service burdens are essential for the design
and implementation of effective and sustainable adjustment programs.
Both the World Bank and the IMF will have to pay increased attention
to two factors. First, realistic estimates are needed of sustainable current
account deficits or, more precisely, of the feasible balance in trade and
nonfactor services over the medium term. Second, adjustment programs
must be internally consistent-that is, they require a set of policies that
will bring about a financeable fiscal deficit that matches the private sav-
ings surplus over investment at levels of investment high enough to sustain
output growth. This situation clearly does not exist at the moment for
the indebted countries, given the magnitude of the net real resource trans-
fers abroad. For these countries, there can be no sustained implemen-
tation of a growth-oriented structural adjustment program unless the
design of the program squarely faces the problems posed by the debt
overhang.
The fourth cluster of observations concerns the existence of an external
environment that is supportive of structural adjustment. Such an envi-
ronment appears to be a necessary-but not sufficient-condition for
sustained domestic reform programs. Four elements are particularly cru-
cial: (1) medium-term external financial support that is adequate (in vol-
ume, terms, and conditions) for serious adjustment efforts, (2) growth-
oriented adjustment in industrial countries, (3) reduction of protectionist
barriers in industrial countries, and (4) no increase in real international
interest rates.
Moeen Qureshi
Both within and outside the World Bank there are two groups with op-
posing views on the issue of adjustment lending-the partisans and the
agnostics. My first point to both the apostles and the heretics is that
adjustment lending is a child born of sheer necessity. In the early 1980s,
following the second oil shock, some of the World Bank's largest bor-
rowers faced very severe balance of payments crises and were forced to
cut back drastically on their investment programs. Clearly, they wanted
help from the World Bank and the World Bank wanted to help them.
But its traditional instruments-especially investment lending-had lim-
ited relevance and effectiveness in that situation. These countries could
not afford and did not have the counterpart funds to undertake invest-
ments, and in any case that was not the highest priority. The highest



Lessons of Experience and the Future of Adjustment Lending 541
priorities in most countries were, first, effective demand management
and, second, in a more medium-term context, fundamental structural
change that would adapt their economies to a changed international eco-
nomic environment. So, necessity being the mother of invention, ad-
justment lending was invented by my predecessors in the World Bank.
Had that not been the case, the World Bank would have been out of
business, and it could not have provided large segments of the developing
world with both financial assistance and technical advice.
My second point is that it is important to recognize both the potential
and the risks of adjustment lending for the World Bank and the countries
that receive such loans. There are countries that received adjustment loans
but have nothing to show for them-it has largely been money down
the drain. In some of these countries that appear ready to make another
effort at adjustment we are prepared to plunge into battle again with
unabated faith. Other countries have achieved varying degrees of success
in implementing adjustment programs and are clearly better equipped
today to use whatever resources they have to support their development
process. The key arguments for and against adjustment lending would
appear very vividly to anyone taking a sight-seeing tour through the large
number of countries in which the World Bank has supported adjustment
efforts through both the International Bank for Reconstruction and De-
velopment (IBRD) and the International Development Association (IDA).
I will comment briefly on certain operational aspects of adjustment
lending. First, how do we formulate and prepare adjustment programs
these days? Quite frankly, this is a very imperfect craft, born of an inexact
science, and practiced on a patient whose diseases are very hard to di-
agnose. Therefore I agree with Gerald Helleiner that a measure of mod-
esty and humility is indeed needed. The tendency to overgeneralize should
be avoided. Above all, there should be a receptivity to learning the lessons
of experience-especially those of failure.
In the adjustment programs that I oversee, the macroeconomic frame-
work is conceptually the easier part but it is rarely done well in our
lending operations. Even by my own somewhat out-of-date standards,
it is clear that the analytical and conceptual basis for our macroeconomic
models needs strengthening. I am looking with expectation to Stanley
Fischer and other Policy, Research, and External Affairs staff to provide
the intellectual inputs that will refurbish that aspect of our work. The
sectoral or structural work in our adjustment loans is much better, al-
though it sometimes lacks thematic cohesion. In the sectoral area, the
World Bank's craft has been honed in the forge of experience. The more
experienced and skilled practitioners of the art in the World Bank-our
technical specialists-are beginning either to retire or to leave, however,
because they think adjustment lending is in vogue and their craft is less



542 ASSESSMENT
valued. Nevertheless, they are, and continue to be, the backbone of some
of our best work in the adjustment area.
Second, it should be recognized that adjustment lending, in most of
its aspects, is a pretty blunt instrument. Macroeconomic policy, for all
its conceptual and rather flamboyant architecture, is nonetheless an im-
precise operational tool. Therefore one must avoid the tendency, however
understandable, to turn this somewhat blunt instrument into a surgeon's
scalpel. This is attempted too frequently in our adjustment loans. As
Helleiner said, they are festooned with conditions in an attempt to achieve
myriad specific policy objectives and targets within the macroeconomic
framework.
Third, how much money should we provide to support an adjustment
program? The answer is, as little as possible-as little as might be needed
to grease the wheels of economic change and catalyze external support
for the adjustment programs. Without adequate external support, pro-
grams are not likely to succeed. But we cannot and should not take upon
ourselves the responsibility for providing the major part of that financing,
because that is not a sustainable role for the World Bank and IDA.
Fourth, how long should adjustment lending continue? Conventional
wisdom suggests five to seven years. My colleagues who sponsor ad-
justment lending proposals do not have a precise time period in mind
but generally argue for a longer period of adjustment the second time
around than the first. Clearly, there is no precise period within which
adjustment lending can be said to be completed. It varies a great deal
from country to country and depends on a host of factors. In addition,
adjustment can, and probably will in most cases, be an interrupted pro-
cess. In my experience, the period for most countries tends to be not less
than five years. But if the process of adjustment is not completed in ten
years, we had better ask ourselves whether the will to adjust is really
there. Indeed, if the political will is lacking, we might just be throwing
good money after bad.
Fifth, the sooner we can make the transition from quick-disbursing
assistance to investment lending the better the prospect of moving the
country from crisis management to more normal conditions. Planning
for the postadjustment period must become an integral part of the fi-
nancial and economic strategy. In the case of IBRD lending, in particular,
this process must be managed carefully so that the World Bank can main-
tain a role in a country long enough to see the adjustment process through
to the end. Otherwise, especially in countries receiving IBRD support, the
World Bank is likely to find itself constrained by the limits on its share
in total external lending to a given country and by its inability to provide
adequate and timely financial support to a country that needs it. In the



Lessons of Experience and the Future of Adjustment Lending 543
process of adjustment lending therefore we must not only define the
policy changes that are envisaged but also look ahead and plan for World
Bank support to the country both during the adjustment period and be-
yond it.



24 Summary and Conclusions
THIS BOOK has assessed the experience of countries with adjustment pro-
grams and the effectiveness of adjustment lending from the World Bank
in supporting these programs. Introduced by the World Bank in 1980,
adjustment lending has provided fast-disbursing, balance of payments
financing and support for policy and institutional reforms. During the
past decade, the use of adjustment lending has expanded dramatically.
This increase has brought mounting criticism of the appropriateness and
usefulness of this instrument as a means to provide financing and policy
support to countries in distress.
The book has taken an inventory of the decade of experience with
adjustment lending in fifty-one countries with loans totaling US$15 bil-
lion (from 1980 to 1987). The studies presented here have comprehen-
sively evaluated such aspects of adjustment lending as its design, policy
components, conditionality, social implications, macroeconomic inter-
actions, and country performances for a sample of nine countries from
all regions of the developing world. In addition, economists and prac-
titioners from both inside and outside the Bank have provided different
perspectives on the evolution and success of adjustment lending. Despite
the diversity of views and differences of assessments, a number of con-
clusions as well as open questions can be identified.
Three positive features of adjustment lending emerge from the inven-
tory presented here. First, among countries requiring adjustments, those
that have carried out major adjustments have performed better on av-
erage in terms of their aggregate economic activity than those that have
not. There is a general and positive connection between adjustment lend-
ing and this relative improvement, but the connection is not a tight one
and performance varies substantially among the loan recipients. A clearer
association emerges, however, when the focus is on the sample of early
recipients who received several loans rather than on all recipients. There
is a significant and positive correlation between relative improvements
in macroeconomic performance and the intensity of adjustment lending
operations. This result is noteworthy because the recipient countries were
subjected to more severe external shocks than were their comparators.
It therefore underscores the importance of timely adjustments and sus-
tained efforts in making a difference to performance.
S44



Summary and Conclusions 545
Second, the overall rate of implementation of loan conditionalities
seems to have been quite satisfactory. This finding is particularly clear
for the conditions in the monitorable components of action programs
that were considered especially important by the authors of the country
studies. In a sample of about fifty adjustment loans to fifteen countries,
60 percent of the loan conditions were subsequently fulfilled. Yet there
is a wide variation in implementation both among policy areas and within
a given country. Third, adjustment lending has become a necessary and
valuable instrument in the World Bank's kit of tools for supporting the
economic development of borrowing members. Furthermore, by virtue
of its focus on major economywide and sectoral reforms of policies and
institutions, adjustment lending has come to play a pivotal role in the
World Bank's dialogue with borrowing countries on the fundamental
aspects of macroeconomic and development Policy. It has thus placed
the World Bank in a central position in the economic adjustment process.
This emphasis should continue-provided the effectiveness of the in-
strument can be enhanced. Notwithstanding the generally positive review
of adjustment lending, the wide variation in its effectiveness raises many
questions about when and how to use it and how to improve it. Here
again three central issues emerge from the varied analyses in this book:
the ownership of the program, the adequacy of macroeconomic stability,
and the size of external funding. In each of these areas there is scope for
improvement and important lessons to be learned from past mistakes.
Program Ownership
By its very nature, adjustment lending is a high-profile activity. It touches
on sensitive policy variables such as the exchange rate, import tariffs,
taxes, and agricultural prices. Sustainable policy reforms in these areas
are predicated on a commitment by governments and policymakers to
the process of change and on the resolution of the tensions inevitably
associated with change. Consensus on the merits of the actions to be
undertaken is necessary for the success of adjustment operations. This
finding underscores the value of country ownership of an adjustment
program and implies that the role of the World Bank should be essentially
a supportive one, providing assistance in the design, implementation, and
supervision of various components of the program.
Judging the commitment of a government to an adjustment path is
inherently subjective. There is never uniform and widespread commit-
ment. Typically a core group of policymakers is in favor of a reform
program. This group needs and often seeks both financial and intellectual
support from outside agencies such as the World Bank. The World Bank,
on its part, must often make a commitment of funds and support based
on its judgment that this core group will succeed in pushing through a



546 ASSESSMENT
reform package. Mistakes clearly occur, as in Zambia, where the com-
mitted group in favor of adjustment was so small that the process had
very little hope of succeeding.
These judgments often fail because of the inherent uncertainty about
the underlying political processes that lead to or hinder reform. Outside
agencies such as the World Bank are not fully cognizant of the underlying
sociopolitical forces. One way for lenders to circumvent these problems
is to make loans only when the required policy changes have already
been executed. Indonesia has repeatedly demonstrated its commitment
through prior actions. A second-best alternative is tranching of loans.
This involves disbursing a part of the agreed loans only after policy ac-
tions have been undertaken on an agreed time schedule. The resort to
tranching reflects considerable uncertainty on the part of the lender that
agreed actions will be carried out. The record with tranched loans has
been mixed, and tranching remains an unsatisfactory and clearly inferior
solution to the lack of clear-cut commitment. It is usually very difficult
to stop the disbursement of second tranches, and often compromises are
made to justify their release. Smaller single-tranched loans seem to be a
better alternative, but they naturally involve higher administrative costs.
In a situation fraught with uncertainty the World Bank, as a public
international financial organization, must be willing to take greater risks.
There are two reasons for this: (1) to counteract what is likely to be the
excessive caution of international commercial banks and (2) to maintain
the World Bank's long-term relationship with client countries who are
also member shareholders of the Bank. The record indicates that the Bank
has taken risks in supporting a nascent reform program even when it
was not clear that the reforms would work. These actions have often
resulted in temporary setbacks or failures, but there have also been no-
table successes. One case of success is Ghana. In 1982 Ghana's record
on reform would have given little hope and much concern to an inter-
national lending agency. Yet the World Bank pursued the course with
subsequent high payoffs.
Macroeconomic Justification
Taking appropriate risks does not imply being foolhardy. One situation
in which reforms are not going to work, however genuine the intentions
and commitment, is when there is macroeconomic instability. Macro-
economic stability should be an integral part of the adjustment effort
and is critical to its success. This requirement raises many complex ques-
tions about the scope, mix, and sequencing of policy measures that must
be determined country by country. In particular, the effectiveness of struc-
tural adjustment involving changes in relative prices depends critically



Summary and Conclusions 547
on a country's ability to restore fiscal balance through more judicious
management of aggregate demand.
How much macroeconomic stabilization is necessary in a particular
situation also requires judgment. Clearly insufficient macroeconomic sta-
bility threatened Turkey's reforms over the past decade, and the World
Bank could have been more strict in ensuring greater stability. But both
the size and the speed of stabilization matter, and if they are too great
they can also threaten reforms. Eliminating in one year a large fiscal
deficit of, say, more than 10 percent of GDP can threaten reforms in one
of two ways. Either the sharp reduction in the deficit can slide the econ-
omy into a sharp recession, or the sharp cuts in income from very rapid
and large stabilization can hurt particular income groups (such as the
poor or organized urban workers) who may threaten social stability. The
size and pace of the required stabilization effort depends on the amount
of external funding available to support the reforms.
External Financing
External financing can assist and promote adjustment or it can substitute
for it (that is, delay adjustment). There are examples of both. The risk
of delayed adjustment is a real one, and it is not easy to judge beforehand
which risk is greater. Yet such judgments are made all the time. Once a
program is initiated on the assumption that certain external funds will
be available, the failure to provide them will kill credibility and can defeat
the program.
The adequacy of the funding of adjustment programs, especially in
low-income countries with characteristically low supply elasticities, re-
ceives considerable attention in the various country studies in this book.
Adjustment programs failed in some African countries primarily because
they were underfunded. In other countries, particularly the highly in-
debted ones, the legacy of the debt crisis continues to diminish the do-
mestic availability of the countries' financial resources and to delay the
recovery of investment. Low investments in productive assets have in
turn handicapped the process of adjustment.
Under these circumstances, taking appropriate measures for debt relief
will have a high payoff in terms of macroeconomic stability. At the same
time, the use of funds available from debt relief should be monitored so
as to improve the position of both creditors and debtors. Debt relief
should come with enhanced conditionality to provide the country with
the incentive to adjust and, perhaps more critically, to avoid the re-
sumption of unsustainable macroeconomic policies.
With the continuation of debt problems and the likelihood of higher
oil prices and higher real interest rates in the early 1990s, the World
Bank will be required to assist a new round of structural reforms. Ad-



548 ASSESSMENT
justment lending will remain a key instrument for the World Bank in
providing financial and policy support to its member countries. The les-
sons from the experience of the past decade must therefore be taken
seriously so that future adjustment operations can be made more effec-
tive.



Index
Accounting rules, 81, 105, 113, 121        transition to project-based lending
Adjustment: and analysis of structural     and, 5-7
adjustment, 11, 12-14; attempts to     Adjustment programs: in Asia, 412-13;
minimize industrial, 159-60; defining,   in Chile, 473-86, 490-93;
405-6; new poor created by, 247-49;      conditionality and differences in,
performance of manufacturing during,      199-201; conditionality and nature
160-76, 187-88; poverty with, in         of, 198-99; in C6te d'Ivoire,
C6te d'Ivoire, 264-65; poverty           288-304, 358, 359; design and, 539;.
without, in Peru, 266-67; shocks of      design and agricultural, 136-48;
1980s and, 11; stabilization and, 11,    economic analysis and, 524-25;
12; types of policy responses and, 11    extent of trade reform and, 47-54;
Adjustment with a Human Face (Cornia,      external financing and, 547-48;
Jolly, and Stewart), 247                  functioning of, 533-34; funding of,
Adjustment Lending: An Evaluation of       542, 547; in the future, 532-33;
Ten Years of Experience (1988            general principles of, 531-32; in
summary report), 534                     Ghana, 311-26, 359; global
Adjustment loans: agricultural reform      development and, 530; growth-
and, 132-36; conditionality and,         oriented fiscal, 38-40; impact of
198-201, 545; defining, 533; defining    trade, 67-69; import liberalization
adjustment and, 11; effectiveness of,;   and, 534; in Indonesia, 338-42,
in financial sector, 76, 91-92;          362-66, 378-80; internal World
evaluation reports and, 197-98;          Bank procedures and, 528-29;
expenditure and, 32; experience with,    investment and, 542; in Korea,
544; export performance and, 59-65;      391-402, 403-4, 406-11; length of,
fiscal deficits in loan-intensive        542; loan review and, 528;
countries and, 23-26; fiscal reform      macroeconomic aspects of fiscal,
and, 20; forms of financial sector,      20-31, 44; macroeconomic
92-94; and implementation of             justification of, 546-47; methodology
conditionality, 202-13; import           of fiscal, 44-45; in Mexico, 497-509,
protection and, 57; internal reviews     511-14, 515-16; new poor created
and, 197; introduction of, 1-2, 193,     by, 247-49, 263; operational aspects
544; lessons from experience with,       of, 541-42; ownership question and,
2-4; manufacturing and, 19, 159-76,      279-82, 527-28, 545-46; in
187-88; policy and financial sector,     Pakistan, 415-28, 432-33; policy
81-85, 95; political factors and, 4-5;   and, 535-37; policy change and trade
positive features of, 544; project loans  and, 54-67; positive features of, 544;
in agriculture and, 131; public          public sector and fiscal, 31-34;
enterprise reform and, 109-23,           resolving problems of debt and,
124-25; risks involved in, 541-42;       530-31; social costs and, 247-49,
trade policy proposals and, 48-49;       256-57, 525, 538-39; surveillance
549



550 INDEX
and, 532; sustainability and, 523,        18, 133, 147; input subsidies and,
536, 537-38; tax reform and, 34-38;       140, 144, 154; institutional reform
technical and political tradeoffs and,   and, 133-36, 148; investment and,
523-24; trade and, 534-35, 536; in        139, 140-41, 155; irrigation and,
Turkey, 208, 441-56, 463-65; views        139; macroeconomic policies and,
on, 540-41; in Zambia, 334-48,            135-36; Malawi and, 212; nonprice
342-50, 355-58, 359                       factors and, 139-41, 152, 153; policy
Adjustment programs, evaluation of:        reforms and, 132-36; price
conditionality and, 199-213, 215-19;     stabilization and, 137, 142-44,
difficulties of, 217-20, 222; external    155-56; pricing and, 132-33,
environment and, 226-28, 244;             137-38, 139, 145, 146-47, 152-53,
Ghana and, 317-26; how adjustment         154; privatization and, 135, 147, 148,
has worked and, 533-34; indicators        158; program design and, 136-48;
of effective performance and, 223-25,    research and, 139, 154; trade
229-33; limitations of, 222-23; loan     liberalization and, 133, 144-46, 147,
objectives and, 193; macroeconomic        157
performance and, 225-26; measuring      Agriculture, 209; and analysis of
impact of, 538; measuring success of,     adjustment, 13; bias against, 200;
526; performance of loan recipients       Chile and, 137, 138, 139; C6te
and, 233-35; poverty and, 257-67;        d'lvoire and, 264, 265, 300; Ghana
social welfare and, 249-56, 257-63,       and, 309, 318; Indonesia and, 376,
271-75; and tradeoffs between             378; Mexico and, 132, 133, 136, 140,
growth and external adjustment,           147, 149, 262, 503, 510, 511, 513;
235-39; Zambia and, 342-50                Pakistan and, 415, 417-18, 420, 421,
Adjustment programs, reform under. See     423, 430, 432, 434, 436, 437; Peru
Agricultural reform; Exchange rate;       and, 266; policy reform and, 17-18,
Financial sector reform; Fiscal reform;   535; poverty and subsistence crops
Industrial reform; Policy reform;         and, 258; Turkey and, 441, 447,
Public sector reform; Trade reform        452-53, 455-56; Zambia and, 335,
Africa: adjustment programs in,             344, 353, 354, 355
199-200, 203, 211, 216-17, 353,        Algeria, 262
358-60; agricultural policies and, 17;  Alliance for Progress (Latin America),
export promotion and, 146; growth         216
in, 322; industrial policies and, 18;   Argentina, 16, 472, 517; export taxes
industrial reform and, 161, 162, 164,     and, 142; financial sector reform and,
173, 174, 175, 176, 179, 180; LSMS       76, 87; fiscal deficits and, 22; land in,
data and, 255; performance                138; social expenditures in, 254, 263;
evaluation and, 231; price reforms        tax reform and, 36; trade reform and,
and, 132; public sector reform and,      52
108, 109, 111, 112, 113, 114, 121;     Arrears, 112, 123, 316, 334, 338, 342
reorganization of public enterprises in,  Asia: adjustment and, 412-13; import
135; social welfare analysis and, 248,   substitution and, 174; industry and,
250, 272, 273, 274; trade reform and,     162, 164, 175; LSMS data and, 255;
49, 53. See also names of individual     policy-based lending and, 216; social
countries                                welfare analysis and, 248. See also
Agricultural procurement schemes, 145      names of individual countries
Agricultural reform: agricultural growth  Auditing rules, 81, 105, 121
and loans and, 131; agricultural       Automobile industry (Korea), 399-400,
protectionism and, 145-46, 157; and      402
analysis of adjustment, 17-18;
deregulation and, 135, 147; extension  Baker Plan, 358
and, 139; India and, 216;              Balance of payments: adjustment and,
infrastructure and, 139; inputs and,      11, 529; Chile and, 469; financial



Index 551
sector reform and, 93; Ghana and,    Bureaucracies, 3, 217; high
311; Indonesia and, 361, 363,          administrative costs of, 254; Mexico
366-69, 370, 379, 383; industrial      and, 503; social service, 263; Turkey's
reform and, 161; Korea and, 403;       exports and, 447
Mexico and, 512, 516; Pakistan and,  Burundi, public sector reform and, 114
437; policy reform and, 6; structural  Busia administration (Ghana), 311
adjustment and, 4, 5; trade reform
and, 48, 52; Turkey and, 440;        Cameroon, 153
Zambia and, 348                      Capacity utilization, 534; adjustment
Balassa, Bela, 72, 73, 146, 233         lending and, 235; manufacturing in
Bangladesh: credit program in, 260; tax  Zambia and, 338-39; Turkey's export
reform and, 36; trade reform and, 53;  promotion and, 448
utility companies in, 34             Capital markets, 79, 80; development
Bank associations, 104                   of, 85, 105; Indonesia and, 365;
Bankruptcy, 389                          Korea and, 389
Banks, 16, 228; adjustment and, 524;   Caribbean: industrial reform and, 162;
causes of failure of, 76-78, 101; Chile  price reforms in, 132. See also Haiti;
and, 471, 473, 476-77, 478, 480,       Jamaica
481, 484, 486, 493; commercial,      Central African Republic, 114
80-81, 101; and credit schemes for   Chad, 258-59
the poor, 260; functions of, 79;     Chemical industries, 18
Ghana and, 319, 325, 331; Indonesia  Chile, 16, 128, 206, 208, 211, 279, 283,
and, 365, 368, 373; inflation and,     284, 463, 534, 536; adjustment
530; Korea and, 389, 397, 398;         program in, 473-77, 517; agriculture
Mexico and, 495, 497, 502, 512;        and, 137, 138, 139; automatic
ownership links and, 94; Pakistan      adjustment process and, 471;
and, 416, 425; performance             conditionality and, 486; copper
evaluation and, 228, 243; reduced      stabilization fund and, 143; economic
lending and, 200; restructuring of,    crisis in, 468-73; economic
81-84; Turkey and, 444, 445-46,        performance and adjustment and,
451, 454, 460; Zambia and, 336, 347    477-86; export promotion and, 146;
Benin, public sector reform and, 109,    external shocks and, 472; financial
111, 114                               sector and, 76, 87, 92, 104, 471,
Black, Eugene, 353                       472-73, 474, 476-77, 481-84;
Black market, 333, 337, 357, 463         government expenditure and, 481,
Blejer, M. I., 104                       485-86, 525; health expenditures
Bolivia, 66, 209; exchange rate and      and, 485; industrial reform and, 161,
deficits and, 30; LsMs data and, 255;  166, 175, 176; public investment and,
social welfare analysis and, 248, 250,  33; social welfare and, 248, 262, 263,
258, 259, 261, 262, 263                478, 484-86; trade and, 47, 49, 53,
Bonds, 16                                57, 67, 473, 475, 480-81, 487; VAT
Borrowing: Chile and, 471; fiscal deficit  and, 481, 485; World Bank and, 473,
and, 22; Indonesia and, 368; Mexico    474, 475, 492
and, 495-96, 514; during 1970s, 530;  China, 260
Pakistan and, 416, 424, 437; Turkey  Choksi, Armeane, 50
and, 440, 441, 450, 463-64           Cocoa: Cote d'lvoire and, 287, 288,
Bourguignon, Francois, 225               289, 295, 296, 298-99, 300-301,
Branson, William H., 225                 305, 306; Ghana and, 309, 315, 316,
Brazil, 16, 517; price controls in, 31;  317, 330
social expenditures in, 254, 262; trade  Coffee, C6te d'Ivoire and, 287, 288,
reform and, 48                         289, 295, 296, 298-99, 300-301,
Buchi, Herman, 491                       305, 306



552 INDEX
Colombia, 18, 209; agricultural reform   303-4, 358, 359; commodity boom
and, 136; industrial reform and, 166,  in, 287-88, 300, 305; deficit and, 24,
175, 176; trade reform and, 53, 54,    287, 288, 294, 306; exchange rate
58; utility companies in, 34           and resource allocation in, 302-3;
Compensatory programs, 258-59, 263,      fiscal policy and, 299-302; industrial
265                                    reform and, 166, 173, 175;
Conditionality, 194; adjustment and,     investment and, 33, 287, 288; labor
198-99, 536, 545; Chile and, 486;      market and wage policy in, 302; LSMS
differences in programs and,           data and, 255; public sector reform
199-201; difficulty of monitoring,     and, 118, 120, 121, 122; savings and,
218; economic environment and,         303; trade and, 52, 53, 73, 289, 298
212-13; factors affecting            Credit, 201, 208, 209; bank failure and,
implementation and, 206-13, 213-14,    77; Chile and, 469, 471, 472, 475,
215; Ghana and, 322-26; government     491; control on, 84, 92, 104; Cote
commitment and, 210-11;                d'lvoire and, 295; direct, 103;
implementation and, 202-6, 215-19;     excessive regulation and, 78; Ghana
industrial reform and, 162, 176-79,    and, 316, 319, 323, 325; Indonesia
180; Korea and, 402; lending           and, 363, 365, 368, 372; industrial
selectivity and, 214; macroeconomic    reform and, 161; intermediation loans
policies and implementation and,       and, 81, 103; Korea and, 397-98,
207-8, 214; Pakistan and, 428-32;      403, 406, 409; Mexico and, 494, 502;
tranche-release conditions and, 199,   Pakistan and, 416, 425; performance
202, 208-10, 326; Turkey and,          evaluation and, 226, 227, 228, 229,
456-57, 465; World Bank advice and,    231, 241, 243, 245; poverty programs
211-12, 214; Zambia and, 343-45.       and, 260; selective, 102; subsidized,
See also Key conditions                88, 90, 104, 140, 146, 266; Zambia
Congo (Republic of), public sector       and, 344
reform and, 109, 111, 112            Credit rationing, 86, 245
Consumption: and analysis of           Creditworthiness: Mexico and, 506;
adjustment, 12; Chile and, 478;        Turkey and, 463; Zambia and, 349
Ghana and, 320; household surveys    Cruzado Plan (Brazil), 31
and, 254; Indonesia and, 361, 376,   Current account: deficit in Chile, 469;
385; Korea and energy, 395-97,         deficit in Indonesia, 366-67; deficit in
400-401; Pakistan and, 419, 424,       Korea, 390; deficit in Pakistan, 427,
425; price controls and, 142, 143;     435; deficit in Turkey, 464; deficit in
taxation and, 35; Zambia and, 333,     Zambia, 333, 340, 344, 347, 348;
334, 335, 339                          deficits and World Bank and IMF, 540;
Contract plans, 114, 116, 122, 123       evaluation and, 224, 232; surplus-in
Convertible Turkish Lira Deposits        Mexico, 516; trade and imbalances in,
(CITDs), 440, 445                      530
Conwvay, P., 451, 452                  Current account payments, 88
Copper, Zambia and, 332, 333, 335,     Customs duties, 30, 41; tax reform and,
338, 340, 342, 343, 347, 348, 349,     36; trade reform and, 54
354, 357
Corden, W. Max, 235                    Debt: arrears and, 112, 123, 316, 320,
Corina, G. A., 249                       334, 338, 342; Chile and, 474-75,
Costa Rica, 209; social welfare analysis  476, 478-80; Cote d'lvoire and, 289;
and, 250; trade reforms and, 52, 53    cross-debts and, 112; Ghana and,
Cost recovery measures (public sector),  309, 312, 316, 325; Indonesia and,
257                                    362, 368, 379, 382; Korea and, 386;
C6te d'lvoire, 205, 208, 211, 279;       macroeconomic performance
adjustment program in, 288-99,         evaluation and, 224, 235, 241;



Index 553
Mexico and, 494, 508, 515; Pakistan    fiscal policy and exchange rate and,
and, 424, 427-28, 431, 435, 437;       29-30, 45; Ghana and, 311, 312,
public enterprise, 122; repayment of   315, 316, 317, 324, 327; India and,
foreign, 14; resolving problems of,    216; Indonesia and, 363, 383, 384;
530-31, 540, 547; Turkey and,          industrial reform and, 167, 173, 174,
444-46, 450, 451, 456, 464-65;         178; Korea and, 388, 393; Mexico
Zambia and, 338, 346, 347, 356         and, 501, 516; Pakistan and, 426;
Debt service, 14, 161; adjustment        social welfare and, 256, 267; trade
process and, 524-25, 540; Chile and,   reform and real exchange rate and,
480; C6te d'lvoire and, 301; deficit   54, 73; Turkey and, 440, 441,
and exchange rate and, 40;             447-48; Zambia and, 355; Zimbabwe
devaluation and, 30; economic          and, 40, 41
indicators and, 224, 235; Ghana and,  Devarajan, Shantayanan, 302, 303, 359
309; Indonesia and, 362, 368; Mexico  Developing countries: bank failures in,
and, 494, 508; Pakistan and, 427-28,   76-78; deficits in, 20-21;
431; Turkey and, 444, 464; Zambia      disprotection of agriculture in, 144;
and, 338, 340, 347, 356                economic indicators for, 279; financial
Deficit, 14; adjustment and fiscal, 12,  sector reform and, 76, 78, 101, 102,
540; adjustment loan-intensive         103, 105; growth in, 322; industrial
countries and, 23-26; causes of,       sector and development in, 179; price
20-21; Chile and, 474, 478; C6te       policy and, 153, 155; public sector
d'lvoire and, 24, 287, 288, 294, 306;  reform and, 108; tariffs and, 47-48;
economic indicators and, 224;          tax reform and, 36; trade reform and,
financial sector reform and, 16;       47, 59, 73
Ghana and, 309, 319, 320; growth     Development finance institutions (DFIS),
and fiscal, 26-29, 39; Indonesia and,  80, 102, 103
364, 366-67, 369-70, 383; Korea      Development strategies: adjustment and,
and, 386, 390; Mexico and, 22, 24,     11, 523, 527, 536; distortions in,
494, 495, 502, 515; Pakistan and,      1-2; financial systems and, 88;
416, 417, 424, 427, 434, 435, 437;     industrial, 179, 188; misconceptions
reform and causes of fiscal, 21-22,    concerning, 3; role of government
44-45; tax reform and, 36; Turkey      and, 13
and, 441, 449, 450, 461, 464;        The Diplomacy of Economic
Zambia and, 24, 30, 333, 334, 335,     Development (Black), 353
336, 340, 341, 344, 347, 348, 354,   Divestiture (public sector), 113-14,
356; Zimbabwe and, 40, 41              115-16, 116-20, 123-24, 128-29
Deindustrialization, 164, 179          Dutch disease, 142, 283
de la Madrid, Miguel, 495, 501, 503
de Melo, Jaime, 225, 302, 303          East Asian countries, 13. See also Asia
Deposit insurance, 81, 82-83, 471      Eastern Europe, 13. See also names of
Depreciation, 14; Chile and, 469;       individual countries
Indonesia and, 363, 367, 368, 373;   Economic growth, 213; adjustment and,
Indonesia and U.S. dollar, 363, 368;   537; fiscal adjustment programs and,
Pakistan and, 427; trade reform and    38-40, 45; fiscal deficits and, 26-29;
real exchange rate and, 54, 66, 67,    Ghana and, 329, 330; Indonesia and,
72-73; Turkey and, 448, 452;           361, 368, 375-76, 384-85; Korea
Zambia and, 336, 337, 338, 343, 347    and, 388, 403; macroeconomic
Devaluation, 14, 19, 136, 142, 159;      performance evaluation and, 224;
agricultural reform and, 133; bank     Pakistan and, 415, 421-23, 433-34;
failure and, 77; Chile and, 471, 474,  the poor and, 261; tradeoffs between
475, 491; debt service and, 30;        external adjustment and, 235-39,
economic indicators and, 224, 231;     244, 246; trade reforms and, 59, 69;



554 INDEX
Turkey and, 440, 460; Zambia and,      146, 475; Mexico and, 501; social
332                                    welfare and, 256; Turkey and, 446,
Ecuador: agricultural reform and, 133,   447-49
135; financial sector reform and, 76  Exports, 29, 77; adjustment and, 12;
Egypt, 262                               agricultural, 133, 144, 145-46, 153,
Employment, 260, 264; Chile and, 472,    200, 264, 265, 273, 298, 378, 447,
481, 485; Indonesia and, 377, 379;     452, 453, 480; agricultural price
job creation schemes and, 257;         stabilization and, 156; Chile and, 480,
Pakistan and, 423; public sector and,  491; conditionality and, 204, 208,
200, 333, 335; 338; Turkey and, 455    209, 212, 213, 217; Cote d'lvoire
Energy: Pakistan and, 417, 419, 423,     and, 289, 296, 298, 304, 307;
430, 433; Turkey and, 447, 453-54.     exchange rate and, 14-16; fiscal
See also Oil                           deficits and, 21; Ghana and, 309,
Energy conservation (Korea), 395-97,     315, 316, 322; Indonesia and, 362,
400-401                                363, 366, 367, 368-69, 375, 376,
Environmental issues, 136, 535           377, 378, 382; industrial reform and,
Evaluation. See Adjustment programs      161, 162, 167, 173, 174-75, 176,
(evaluation of); Industrial performance  177, 178-79, 189; inward-looking
evaluation; Macroeconomic              strategy and, 13; Korea and, 388,
performance evaluation                 389, 390-91, 407-8, 410-11;
Exchange rate: agricultural reform and,  macroeconomic performance
133, 135-36, 145, 148, 152, 153;       evaluation and, 231, 241; Mexico
agriculture and overvaluation of, 144;  and, 494, 496, 497, 500, 501, 507,
and analysis of adjustment, 12;        508, 510, 512, 513; Pakistan and,
auctions and trade reform, 52; Chile   415, 416, 418, 421, 423, 426, 427,
and, 469, 473, 474, 477, 480-81,       431, 432, 433, 435; Peru and, 266;
486, 491; conditionality and, 208,     trade reform and, 48, 49, 50, 52, 53,
213, 217; Cote d'lvoire and, 298,      54, 57-58, 59-65, 67, 69, 73-74;
302-3, 304, 307, 359; fiscal reform    Turkey and, 440, 446, 447-49, 452,
and, 29-30; Ghana and, 311, 312,       453, 460, 463; Zambia and, 332,
315, 320, 324; Indonesia and, 363,     333, 334, 335, 336, 337, 338, 340,
367, 371, 373, 382, 383; industrial    342, 344, 346, 347, 348, 349;
reform and, 160, 161, 162, 173, 177;   Zimbabwe and, 41
inflation and, 40; inward-looking    Export taxes, 30, 37, 50, 52, 53, 74,
strategy and real, 13; Korea and, 386,  142, 309, 316, 335, 340, 497. See
388, 393; macroeconomic                also Taxes
performance evaluation and, 224,     Extension, agricultural reform and, 139
231; Mexico and, 500, 510, 514, 515;  External shocks, 159; Chile and, 472,
Pakistan and, 416, 418, 423, 427,      490; and countries with industrial
431, 436; policy involving, 14-16;     adjustment loans, 161, 179; fiscal
price stability and, 142; trade reform  deficits and trade-related, 24-25;
and, 48, 54-5-5, 59; trade reform and  Indonesia and, 361-62; Korea and,
depreciation of, 50-52; trade reform   388, 389-91; of 1980s, 11; Pakistan
and real, 54, 66, 67, 72-73; Turkey    and, 414, 436; performance measures
and, 441j 447, 452, 463, 464;          and, 227, 228, 241, 244, 245
Zambia andi 333, 334, 335, 336-37,
338, 343, 346, 347; Zimbabwe and,    Faini, Ricardo, 232, 243, 384
40                                   Fertilizer, 147, 344, 346, 453, 503, 510,
Expenditure. See Government              511; subsidy for, 34, 154-55, 417,
eipenditure                            428
Export licensing, 452, 500             Financial audits, 113
Export prothotion, 162, 177; Chile and,  Financial sector: adjustment process and,



Index 555
524, 527, 531; bank failures and,     Foreign exchange: adjustment and, 527;
76-78; Chile and, 76, 87, 92, 104,      Ghana and, 311, 312, 315, 317, 329;
471, 472-73, 474, 476-77, 481-84;       Indonesia and, 363, 367, 368, 384;
excessive regulation and, 77, 78-79;    industry and, 160, 161, 167, 173,
external financing and, 547-48;         174, 177, 188; Mexico and, 495, 501;
highly indebted countries and, 201;     Pakistan and, 415, 425, 426, 434;
Korea and, 386, 402, 403; Mexico        Turkey and, 440, 446, 451, 463, 464;
and, 502-3; policy and, 16-17; taxes    Zambia and, 206, 333, 336, 340, 343,
and, 78, 79, 85, 88; Turkey and, 452,   344, 346, 347, 348, 349, 355
454-55, 461
Financial sector reform: adjustment and,  General Agreement on Tariffs and Trade
16-17; bank failures and, 76-78,        (GA1T), 47, 146, 148, 149, 157, 449,
101; bank restructuring and, 81-84;     500, 513, 532
conflicting policy arguments and,     Germany, Federal Republic of, 445
87-89; forms of loans and, 92-94;     Ghana, 17, 18, 205, 206, 209, 211, 213,
Indonesia and, 365, 373;                279, 284; adjustment program in,
intermediation loans and, 80-81, 93,    311-17, 359, 524-26, 535;
103; liberalization and, 77, 84,        agricultural reform and, 136; and
86-88, 90, 92, 95, 104; need for,       assessment of conditionality, 322-26;
76-81; nonbank sector and, 79-80;       cause of economic crisis in, 309-11;
other adjustment programs and,          compared with C6te d'lvoire, 308;
86-91; policy package for, 81-85,       economic performance and adjustment
95; policy successes and, 91-92;        in, 317-22; financial sector reform
regulation and, 77, 78-79, 81, 84, 86,  and, 76; industrial reform and, 166,
90, 101                                 175; LSMS data and, 255; public
Finger, Joseph M., 48                     investment and, 33; public sector
Fiscal crises: causes of deficits and,    reform and, 111, 112, 114, 122, 123,
20-21; fiscal reform and causes of,     125; social implications of adjustment
21-22; Ghana and, 309-11; Zambia        in, 326; social welfare analysis and,
and, 332-34                             248, 250, 259, 261, 262, 263;
Fiscal reform: adjustment and, 14, 20;    stabilization and, 208, 312; trade
adjustment loan-intensive countries     reform and, 48, 50, 59; workers
and, 23-26; causes of deficits and,     returning to, 309
20-21; causes of fiscal crises and,   Glewwe, Paul, 264, 266, 267
21-22; C6te d'lvoire and, 288,        Goldstein, Morris, 224
289-95, 296, 299-302; exchange rate   Government: commitment to programs
and, 29-30; fiscal deficits and growth  and, 198, 210-11, 217, 279, 329,
and, 26-29; growth-oriented fiscal      487, 545-46; lack of commitment to
programs and, 38-40; Indonesia and,     trade reform on part of, 53; role of, in
363-64, 383; macroeconomic aspects      development, 13; technical and
of, 20-31; Mexico and, 502; Pakistan    political tradeoffs and adjustment and,
and, 437, 439; price policy and,        523-24
30-31; public sector resource         Goverment expenditure: and analysis
allocation and, 31-34; study            overnment         Chile and 481i
methodology and, 45-46; tax reform      of adjustment, 12; Chile and, 481,
and, 34-38, 44; trade policy and, 30;   485-86, 525; Cote d'lvoire and,
Turkey and, 449-51; Zambia and,         301-2; deficits and, 21, 39; effects of
335-36; Zimbabwe and, 40-41             reduced, 26; Indonesia and, 369-70,
Fischer, Stanley, 541                     376, 378; Korea and, 406; Mexico
Food sector (Mexico), 503                 and, 494, 495, 502; Pakistan and,
Food stamp program, 144, 258, 506         415, 417, 424-25, 437-38; recurrent,
Food subsidies, 143-44, 258, 262, 505,    in public sector, 31-32, 44; Zambia
506                                     and, 340-42, 347, 348, 354, 356



556 INDEX
Gross domestic product (GDP): and        inward- looking strategy and, 13;
analysis of adjustment, 12; Chile and,  Korea and, 389, 390, 391, 397,
472, 473, 477-78, 490; C6te d'lvoire   398-99, 403, 407, 410; liberalizing,
and, 296, 297, 301, 302, 304, 305,     54, 55-58, 66, 74, 201, 216; Mexico
306; economic indicators and, 224,     and, 500-501, 503, 507, 508,
229, 231, 232, 237; financial sector   512-13, 514; Pakistan and, 415, 416,
restructuring and, 92; Ghana and,      431, 434; performance evaluation
324, 330; Indonesia and, 364, 369,     and, 241, 245; tax reform and, 36;
376, 384; industry and, 164, 166;      trade reform and, 54, 55-58, 64,
Korea and, 388; Mexico and, 494,       65-66, 69, 74; trade reform and
495, 508; Pakistan and, 415, 416,      restrictiveness and, 47-48, SO; Turkey
424, 425, 433, 434; trade policy       and, 446, 449, 463; Zambia and, 333,
analysis and, 66, 67; Zambia and,      335, 337, 339, 340, 344, 346, 347,
332, 335, 338, 339, 340, 347           349; Zimbabwe and, 40. See also
Gross national product (GNP): and        Trade
analysis of adjustment, 12; Indonesia  Import substitution, 12; C6te d'lvoire
and, 362, 367; industrial development  and, 289; Indonesia and, 367, 376;
and, 160; Korea and, 395; Pakistan     industrial reform and, 167, 173, 174,
and, 435; Turkey and, 440              175-76; Korea and, 406, 410;
Growth. See Economic growth              Mexico and, 508; Pakistan and, 416,
Guinea: public sector reform and, 110,   418; trade reform and, 48-49, 54;
114, 119; social welfare analysis and,  Turkey and, 440; Zambia and, 333
250, 259                             Incentives, 189, 206, 525, 536;
Guinea-Bissau, 250, 257, 259, 262, 329   agricultural, 17, 420, 452; and
Guyana, 209, 259; trade reform in, 48    analysis of adjustment, 12; Chile and,
492; for energy conservation, 400;
Haiti, 66, 259, 260                      export, 14, 204, 213, 336, 431, 446,
Halevi, Nadav, 47                        448, 475, 480; fiscal, 389, 502;
Hanson, James A., 102                    industrial, 407; industrialization and,
Helleiner, Gerald, 541, 542              18; price, 213, 307; private sector
Household surveys, 254-55; LSMS and,     and, 327; for privatization, 119;
255, 266                               production, 317; tax, 205, 415, 502;
Housing, 265, 295, 298, 379, 486         trade reform and, 49, 54-58, 59, 64,
Hungary, financial sector reform and, 76  506, 513, 515
Income: and analysis of adjustment, 12;
Import licensing, 206, 344, 365, 367,    Chile and, 472, 485; C6te d'lvoire
407, 500, 514                          and, 264; Ghana and, 322, 327;
Imports: agricultural, 145, 146, 147; of  Indonesia and, 361, 372, 376, 377,
agricultural inputs, 133; agricultural  378, 384; Korea and, 388; Mexico
price stabilization and, 156;          and, 494; Zambia and, 339, 348, 349,
agriculture and taxes on, 144; and     354
analysis of adjustment, 12, 534, 537;  Income distribution, 27, 265, 267; Chile
Chile and, 473, 475, 480-81;           and, 485, 493; Turkey and, 441,
conditionality and, 209, 212; C6te     455-56
d'lvoire and, 298; exchange rate     India, 18, 138; reform and, 216
policies and, 16; fiscal reform and  Indonesia, 14, 16, 22, 282; adjustment
trade policies and, 30; Ghana and,     program of, 362-66, 378-80, 528;
312, 315, 316, 318, 319, 320, 322;     balance of payments and, 366-69;
Indonesia and, 361, 363, 365-66,       economic efficiency and, 374-75;
367, 370; industrial reform and, 161,  economic growth and, 361, 375-76;
162, 167, 173, 174, 176, 178, 179,     economic performance and adjustment
180; industrial sector and, 160;       in, 366-78; external shocks and,



Index SS7
361-62; financial sector policy and,    and, 502, 506, 507, 511-12, 516;
365; fiscal deficit and, 369-70; fiscal  during 1970s, 530; Pakistan and,
policy and, 363-64; inflation and,      433-34, 435, 436; price liberalization
370-72; interest rates and, 372-73;     and, 31; reform and, 207, 208;
monetary policy and, 365; social        Turkey and, 441, 449, 452, 460, 461,
welfare and, 377-78; trade and,         463, 464; Zambia and, 346, 347,
365-66; utility companies in, 34; VAT   348; Zimbabwe and, 40-41
and, 37; World Bank and, 378-80       Inflation tax, 16. See also Taxes
Industrial performance evaluation,      Infrastructure, 154; agricultural reform
160-62, 166, 187-88                     and roads and, 139; Ghana and, 313;
Industrial reform: and analysis of        Indonesia and, 378, 379; Pakistan
adjustment, 18-19; Chile and, 161,      and, 415
166, 175, 176; C6te d'Ivoire and,     Institutional change: agricultural reform
166, 173, 175; general patterns of      and, 133-36, 148; conditionality and,
manufacturing growth and, 173-74;       217; fiscal costs of, 93; poverty
Ghana and, 166, 175; Korea and,         alleviation programs and, 261; public
161, 166, 175, 176, 388-89, 395-96,     investment and, 33; public sector
397, 398, 399, 400, 406, 407, 408,      reform and, 112-13, 122-23, 124,
409; loan conditionality and, 162,      125; trade policy and, 67
180; loan conditionality and          Institutions: elements of adjustment and,
performance and, 176-79;                3; structural adjustment and, 11, 12,
manufacturing growth and, 162-66,       13
180-84; manufacturing performance     Interest rates: adjustment and, 13, 524;
indicators and, 160-62, 187-88;         Chile and, 468, 469, 472, 477, 478,
Mexico and, 166, 175, 176, 513;         481, 486, 491; conditionality and,
Pakistan and, 421; performance          217; C6te d'lvoire and, 289, 294,
evaluation and, 166; regional patterns  301, 303; decontrolling, 201; financial
of growth and, 164, 174-75; sectoral    sector policy and, 16, 87, 92, 104,
performance analysis and, 159-76;       105; fiscal deficits and, 21, 22, 27,
sources of growth and, 166-76;          44; Indonesia and, 362, 365, 369,
Turkey and, 161, 166, 173, 174, 176;    370, 372-73, 376, 382, 384;
Zambia and, 166, 173, 339               industrial reform and, 162; Korea
Industry: agriculture and, 131; and       and, 389, 397, 406; liberalization of
analysis of adjustment, 13; Ghana       financial sector and, 87; Mexico and,
and, 318; intermediation loans and,     494, 495, 502, 507, 508; Pakistan
80; loans and reforms in, 205; policy   and, 424, 425; performance
reform and, 18-19; Zambia and, 339,     evaluation and, 226, 227, 235, 241,
347. See also Manufacturing             244; reform and, 207, 208, 535;
Infant industries, protection of Korean,  regulation of, 78, 84, 86, 101;
407, 409, 410                           subsidized, 90, 92, 486, 505; Turkey
Infant mortality, 250, 254, 525           and, 444, 450, 451, 452, 454;
Inflation, 14, 148; adjustment and, 12;   Zambia and, 333, 335, 336, 340, 341,
bank failure and deflationary policy    343, 344, 346, 347, 356
and, 77; in Brazil, 31; Chile and, 474;  Internal reviews (World Bank), 197
economic indicators and, 224, 232,    International Bank for Reconstruction
239; exchange rate and, 40; financial   and Development (IBRD), 541, 542
sector reform and, 86, 87, 88; fiscal  International Development Association,
deficit and, 22, 27, 29, 45; Ghana      114, 359, 541, 542
and, 309, 312, 316, 319; income       International Food Policy Research
taxes and, 38; Indonesia and, 363,      Institute, 157
370-72, 383, 384, 388; Korea and,     International Monetary Fund (IMF), 2,
386, 391, 393-95, 403, 404; Mexico      193, 194, 213, 384; and analysis of



558 INDEX
adjustment, 523, 524, 525, 527, 529,  Jamaica, 17, 205, 208; agricultural
538; Articles of Agreement and, 532;   reform and, 132; financial sector
and assessment of adjistment, 222,     reform and, 76; industrial reform and,
223, 225, 226, 227, 228, 232, 233,     166, 175, 176; LSMs data and, 255;
239, 243, 244; Chile and, 473, 474;    public sector reform and, 109, 110,
conditionality and, 177, 207-8, 210,   114, 115, 117, 120, 121, 125, 128;
215; C6te d'lvoire and, 288, 289,      social welfare analysis and, 250, 254,
299, 306; economic factors and, 540;   260, 263; trade reform and, 48, 50,
fiscal deficits and loans from, 24, 29,  52, 53
44, 45; fiscal issues and, 22; Ghana  Japan, 128
and, 311, 312, 316; Ghana's public   Job creation schemes, 257
investment and, 33; interest rate    Jolly, Richard, 249
reform and, 535; Korea and, 397,
400, 403; Mexico and, 494, 495, 496,  Kanbur, S. M. Ravi, 255, 264, 265
511, 512; Pakistan and, 414, 415,    Kefauver, Estes, 130
416, 417, 418, 424, 425, 428, 429,   Kenya, 205, 206, 209, 212; agricultural
431, 435, 436, 437, 438; policy and,   reform and, 133, 138, 153; financial
537; stabilization and, 216, 228, 247,  sector reform and, 76; industrial
534; trade reform and, 47; Turkey      reform and, 161, 166, 173, 175; tax
and, 440, 441-44, 449, 465; Zambia     reform and, 36; trade reform and, 52,
and, 334, 336, 338, 342, 347, 354      53, 54, 58
Investment, 14, 17, 206, 208, 212, 540;  Key conditions, 206, 214
adjustment and, 534, 535, 542;       Khan, Mohsin S., 225, 232, 233
agricultural, 139, 140-41, 155; bank  Knight, Malcolm, 225
failure and, 77; capital markets and,  Korea (Republic of), 16, 22, 128, 205,
79; Chile and, 469, 473, 475, 476,     211, 213, 279, 282, 283, 463, 534;
480, 481, 487, 490, 492, 493; C6te     adjustment program in, 391-401,
d'lvoire and, 33, 287, 288;            403-4, 406-11, 537; economic crisis
expenditure cuts and public, 26, 28;   in, 386-91; industrial reform and,
fiscal adjustment and public, 32-33,   161, 166, 175, 176, 388-89, 395-96,
44-45; Ghana and, 318, 320, 323,       397, 398, 399-400, 406, 407, 408,
330, 331; Indonesia and, 361, 366,     409; subsidized health and, 259; trade
367, 368, 370, 372, 374, 375, 376,     and, 47, 49, 53, 54, 57, 58, 59, 67,
379, 384, 385; Korea and, 388, 389,    73, 389, 390, 391, 397, 398-99, 403,
393, 395, 398, 404, 408, 409; Latin    407, 410-11; VAT and, 37, 38; World
America and, 518-19; Mexico and,       Bank and, 397, 399, 402, 403, 411
502, 505, 512, 514, 516; Pakistan
and, 415, 417-18, 419, 421, 424,     Labor: C6te d'lvoire and, 302;
428, 430, 432, 434, 437; performance   Indonesia and, 377, 378; Korea and,
evaluation and, 231, 244, 246; project  388, 408; Mexico and, 506; public
lending and, 6; public, in Peru, 266;  sector reform and, 111; Turkey and,
standards of efficiency for, 3; tax    455-56
changes and, 34; tax reform and, S9;  Labor productivity: Korea and, 393;
Turkey and, 29, 440, 447, 451-52,      trade reform and, 59, 64
460, 465; unproductive government,   Laird, Samuel, 48
200; Zambia and, 343, 349, 353, 354  Land, 138
Inward-looking strategies, 13, 58, 407,  Land reform, 266
452, 455, 459                        Latin America, 283; adjustment process
Iran-Iraq war, 448, 464                  in, 516-19; conditionality and, 216;
Irrigation, 139, 266, 379, 417-18, 421,  credit subsidies in, 140; financial
424                                    sector reform and, 87, 101, 104;



Index 559
industrial reform and, 162, 164, 174,  Madagascar, 259; agricultural reform
175, 176, 180; price reforms in, 132;   and, 133, 149; public sector reform
public sector reform and, 113, 129;     and, 109, 111; trade reform and, 50
social welfare analysis and, 248, 260,  Malawi, 205; agricultural prices and,
272, 273, 274. See also names of        153; export crop taxation and, 212;
individual countries                    fiscal deficits and, 22, 24; industrial
Liberalization, 19; Chile and, 469, 471,  reform and, 161, 175; infant mortality
487, 490; constraints on trade, 52,     in, 250; tax reform and, 36; trade
53; in financial sector, 77, 84, 86-88,  reform and, 49, 53
90, 92, 95, 104; imports and, 54,     Malaysia, deficits and, 28
55-58, 66, 74, 201, 216, 446, 449,    Mali, public sector reform and, 109, 111
534; Korea and, 397, 398, 400, 403,   Management, bank, 81, 89, 95
404, 407, 410; Mexico and, 501, 506,  Maniagers, company, 190
509, 512-13, 514; Pakistan and, 416,  Managers, public enterprise, 12, 123,
418, 430, 431, 434, 439; of prices,     124, 125; policy reform and, 17, 111,
30, 31, 45, 337, 346; taxes and trade,  205; reform programs and, 113, 129;
36; trade, 50, 77, 133, 144-46, 147,    steel mill example and, 127, 128;
162, 177, 178, 208, 216, 315, 398,      training programs for, 112
403, 407, 410, 416, 418, 501, 506,    Manufacturing, 19; duty drawback and,
509, 512-13, 514; Turkey and, 440,      204; Ghana and, 318; growth in,
446, 449, 461, 465                      162-66; Indonesia and, 376, 377;
Life expectancy, 248, 249, 250, 254       Korea and, 386, 397; loan
Little, 1. M. D., 72                      conditionality and, 162, 180; loan
Living Standards Measurement Study        conditionality and performance and,
(LSMS), 255, 266                        176-79; and methods of analyzing
Loans: adjustment and review of, 528;     growth, 180-84; Mexico and, 501,
capital markets and long-term, 79;      507; Pakistan and, 418, 419, 420,
hybrid, 140; intermediation, 80-81,     421, 423; patterns of growth in,
93, 103; Korean bank, 398; new poor     173-74; performance evaluation and,
and, 248; performance evaluation and    166; performance indicators and,
bank, 228, 243; project-based, 5-7,     160-62; regional patterns and, 164,
131; reduction in number of bank,       174-75; sectoral performance and,
200; sectoral, 537; Turkey and bank,    159-76; sources of growth in,
454-55. See also Adjustment loans       166-76; trade reform and exports
Long, Millard, 101                        and, 59; Zambia and, 333, 335,
L6pez Portillo administration (Mexico),   338-39
494, 502                              Marcos administration (Philippines), 24
Marris, Richard, 127
Macroeconomic performance evaluation:   Mauritania, 257; LSMS data and, 255;
country classification data and,        public sector reform and, 109
239-41; difficulty of assessment and,  Mauritius, trade reform and, 49, 59, 67
217-20, 222; indicators of effective  McKibbin, Warren, 383
performance and, 229-33; limitations  McNamara, Robert, 1
of, 222-23; and measure of            Mexico, 14, 128, 200, 205, 279, 282,
deterioration in external environment,  283, 284, 534; adjustment program
226-28, 244; methods of, 225-26;        in, 497-506, 511-14, 515-16, 517,
performance of loan recipients and,     537; agricultural reform and, 132,
233-35; selection of performance        133, 136, 140, 147, 149, 262, 503,
indicators and measuring, 223-25;       510, 511, 513; conditionality and,
and tradeoffs between growth and        510-11; deficits and, 22, 24, 494,
external adjustment, 235-39. See also   495, 502; economic crisis in, 494,
Adjustment programs (evaluation of)     495-97, 515, 516; economic



560 INDEX
performance during adjustment in,    Oil shocks, 1, 11, 388, 414, 436, 440.
506-9; export development loans and,   See also External shocks
208; food stamp program in, 144;     Okun, Arthur, 539
food subsidies and, 258; foreign     Operations Evaluation Department
investment and, 505; industrial reform  (World Bank), 197, 223, 353, 354,
and, 166, 175, 176, 513; trade and,    526, 528
48, 50, 52, 54, 57, 58, 59, 67, 494,  Organisation for Economic Co-operation
495, 496, 497-501, 507-8, 510,         and Development (OECD), 47, 101-2,
512-14; World Bank and, 496, 497,      389; Turkey and, 444-45, 448, 460,
500, 503, 506, 509-11, 512, 513        464
Michaely, Michael, 50, 513             Outward-looking strategy, 58, 72, 188,
Middle East, 162, 216; Pakistan and,     459-60
423, 427                             Overstaffing (public enterprises), 18,
Monetary policy: Chile and, 474, 491;    121, 124
Indonesia and, 365, 373, 383-84;     Ozal, Turgut, 441, 462-63, 464, 465
Pakistan and, 425-26; Zambia and,
336, 344-45. See also Devaluation    Paish, Frank, 526
Money markets, 80                      Pakistan, 22, 205, 211, 213, 279;
Monopolies, 110, 119; agricultural       adjustment program in, 415-20,
reform and, 147, 158; in commodity     432-33, 466-67; conditionality and,
trade, 18; Ghana and state, 317        428-32; economic performance and
Monopsonies, 110                         adjustment in, 421-28; economic
Montiel, Peter J., 224                   review (1960s-80s) and, 414-15;
Morocco, 18, 205; agricultural reform    export development loans and, 208;
and, 132, 135, 136, 140, 149; food     during fiscal 1988 and 1989, 433-34;
subsidies and, 258; industrial reform  import liberalization and, 216;
and, 161, 166, 174, 176; public sector  industrial reform and, 161, 166, 176;
reform and, 109, 110, 111, 112,        phasing out of fertilizer subsidy and,
117-18, 120, 122; riots in, 262; tax   34; public investment and, 32; social
reform and, 36, 38, 54, 58, 59, 67;    welfare analysis and, 250, 262; trade
trade liberalization and, 208; trade   reform and, 48, 53, 54, 58
reform and, 30, 52, 54               Panama, 17; agricultural reform and,
135; public sector reform and, 110,
National Investment Fund (Korea), 388,   111, 114, 115, 117, 119, 121, 128;
398                                    trade reform and, 49, 53
Nationalization, 103, 495              Papageorgiou, Demetrios, 50
Nepal, financial sector reform and, 76  Papua New Guinea, 143
New Zealand, 17, 128                   Paris Club, 358
Niger, public sector reform and, 109,  Performance evaluation. See Adjustment
110, 112, 114, 115, 117, 120, 121,     programs (evaluation of); Industrial
122, 123, 125                          performahce evaluation;
Nigeria, 200, 329; financial sector      Macroeconomic performance
reform and, 76; fiscal deficits and, 22;  evaluation
Ghanaian workers and, 309            Peru: LSMS data and, 255; public sector
Nutrition, 250, 258, 525                 reform and, 109; social welfare
analysis and, 250, 266-67
Oil: C6te d'lvoire and, 288, 289, 295;  Philippines, 205, 206, 208; financial
Ghana and, 309; Indonesia and, 361,    sector reform and, 76, 89, 92; fiscal
363, 368, 369, 373, 374, 375, 376,     deficits and, 22, 24; industrial reform
379, 382; Korea and, 388, 399, 400;    and, 161, 166, 176; public sector
Mexico and, 494, 495, 504, 505;        reform and, 109; social expenditures
Pakistan and, 423, 433; Turkey and,    in, 254; tax reform and, 36; trade
453                                    reform and, 50, 52, 54, 57, 58, 67



Index 561
Please effect, 27                        12, 13, 235, 265, 267, 535, 546-47;
Policy: adjustment and, 11-14;           agricultural, 17-18, 258, 333;
adjustment and general principles and  agricultural reform and, 132-33,
areas of, 531-32, 535-37;              137-38, 139, 146-47, 152-53, 154;
agricultural reform, 132-36, 535;      agricultural reform and stabilization
conditionality and macroeconomic,      of, 142-44, 155-56; Chile and, 468;
207-8, 214; development, 1-2, 3, 11,   conditionality and, 205, 206, 213,
13; exchange rate, 29-30; financial    217; Cote d'lvoire and, 287, 288-89,
sector reform and, 81-85, 91-92;       295, 296, 298, 299, 300-301, 305,
food, 143-44; forward-looking          307; fiscal deficits and declining, 21;
industrial, 65; implementation and     fiscal reform and, 30-31; food, 248;
changes in, 203-5, 212-13; inward-     Ghana and, 317, 319, 327; increase in
looking industrial, 13, 58, 407, 452,  utility, 34; Indonesia and, 332, 333,
455, 459; Ivoirian fiscal, 299-302;    334, 337-38, 342, 343, 346, 347,
macroeconomic performance              348, 349, 350, 353, 354, 355, 357;
evaluation and choice of, 224;        industrial reform and, 162, 173, 189;
outward-looking, 58, 72, 188,          Korea and, 391, 393, 400;
459-60; price, 30-31; price            liberalization of, 30, 31, 45, 256;
stabilization, 142-43; public sector   Mexico and, 495, 500, 504-5, 506,
reform and, 110-12; role of            507, 510, 512; Pakistan and, 414,
government in, 13; trade, 30, 86       415, 417, 419-20, 421, 430, 432,
Policy reform: conditionality and, 197,  437; producer, 200; public enterprise
203-5, 207-8, 211-14, 215, 217-18,     reform and, 111, 115, 120-21, 124;
219-20; possibilities of, 6; in sectors  trade reform and, 48, 52, 67; Turkey
not tied to balance of payments, 6;    and, 447, 452, 453; Zambia and, 332,
structural adjustment and, 4. See also  333, 334, 337-38, 342, 343, 346,
Agricultural reform; Exchange rate;    347, 348, 349, 350, 353, 354, 355,
Financial sector reform; Fiscal reform;  357
Industrial reform; Public sector     Primary commodities. See Agriculture;
reform; Trade reform                   Cocoa; Coffee; Copper; Dutch
Political unrest, 142, 262, 346, 356     disease; Oil
Politics: acceptance of programs and,  Private sector, 14, 244; financial sector
210; conditionality and, 217;          bailouts and, 104; financial sector
expenditure cuts and, 26; Ghana and,   policy and, 16; fiscal deficits and, 22,
327, 328; Indonesia and, 385; Mexico   27, 28, 44; Ghana and, 327, 330;
and, 508-9; social welfare analysis    Indonesia and, 372, 376, 379, 384,
and, 261-62; structural adjustment     385; Korea and, 389, 404; Mexico
and, 4-5; Zambia and, 348, 349-50,     and, 494-95, 504, 508; Pakistan and,
356                                    414, 415, 418, 421, 423, 424, 425,
Population growth, 200, 297              426, 428; public enterprise reform
Population program (Indonesia). 379      and, 110; public investment in Turkey
Poverty, 136, 361; and analysis of       and, 32; trade reform and, 53; Turkey
adjustment, 539; Ghana and working     and, 440-41, 447, 449, 451, 465;
poor and, 326; Mexico and, 506; and    Zambia and, 339
new poor created by adjustment,      Privatization, 17; and advice on
247-49; with adjustment (C6te          adjustment, 535; agricultural reform
d'lvoire), 264-65; without adjustment  and, 135, 147, 148; Chile and,
(Peru), 266-67                         469-71, 476, 480, 484, 490; of
Price controls, 31, 142, 143, 147, 161,  hospital services in Jamaica, 260;
177, 205, 213, 311, 317, 334, 414,     Pakistan and, 420; public sector
419, 487, 507                          reform and, 114, 115, 116-20, 125,
Prices, 77; adjustment lending and, 11,  205



S62 INDEX
Procurement schemes, direct, 145        Public sector reform: adjustment loans
Production, 206, 302; adjustment          and, 109-14; assessment of, 114-23,
lending and, 235; agricultural, 136,    124-25; attempted, 109-10;
138; Chile and, 476, 485; copper,       divestiture mechanisms and, 113-14,
332; Ghana and, 313, 317, 318, 319,     115-16, 116-20, 123-24, 128-29;
320; industrial, 164, 176; Mexico       institutional framework for, 112-13,
and, 504; monetary restraint and, 88;   122-23; overstaffing and, 121, 124;
Pakistan and, 421; price instability    performance levels and, 108-9, 113,
and, 155; price stability and, 142,     122; policy framework for, 110-12;
143; reform and, 207; taxation and,     pricing issues and, 111, 115, 120-21,
34, 35; trade reform and competitive,   124; privatization and, 114, 115,
59; trade reform and domestic, 48,      116-20, 125; and rehabilitation of
57; Zambia and, 332, 337, 343, 346      enterprises, 113, 116, 121-22, 124;
Productivity: Indonesia and, 374-75;      results of, 116-20; types of, 115-16
industry and, 160, 188, 189, 190
Project-based lending, 2; agriculture and,  Rawlings administration (Ghana), 328
131; transition from structural       Recession, 161, 248, 261; Cote d'lvoire
adjustment lending to, 5-7              and, 296, 297
Protection: agricultural, 145-46, 157;  Reform in adjustment programs. See
agriculture and industrial, 144; Chile  Agricultural reform; Exchange rate;
and, 206; C6te d'Ivoire and, 330;       Financial sector reform; Fiscal reform;
industrial reform and, 161, 174, 189;   Industrial reform; Policy reform;
of infant industries in Korea, 407,     Public sector reform; Trade reform
409, 410; Mexico and, 501, 508, 512,  Regulation, 13, 16, 524; agricultural
513; trade reform and, 48-49, 54,       reform and, 135, 147; bank failure
55-58, 64; Turkey and, 440, 449;        and, 77; bank restructuring and, 81,
Zambia and, 337                         84, 86, 90, 101; excessive financial,
Public sector, 200, 535; assessment and   78-79; Ghana and, 317; Indonesia'
investments of, 231; Chile and, 468,    and, 365, 366, 367, 371, 374, 379;
473, 476, 478, 481, 486, 490;           Pakistan and, 418-19
conditionality and, 205, 206; Cote    Remittances, workers', 414, 426, 427,
d'lvoire and, 295, 302; cutbacks in,    429, 432, 434, 440
247, 248, 316, 317, 504; employment   Research, agricultural reform and, 139,
and, 200; expenditure cuts and          154
investment of, 26, 28; fiscal reform  Resource allocation: and analysis of
and resource allocation and, 31-34;     adjustment, 12; Cote d'lvoire and,
Ghana and, 316, 317, 327, 329, 330;     301, 302-3; fiscal reform and public
ghost workers in, 317, 327; Indonesia   sector, 31-34; inward-looking
and, 363, 364, 370, 371, 376;           strategy and, 13; Korea and, 388;
industry and, 161; intervention in      productivity and, 189; Zambia and,
agriculture and, 18; Korea and, 397,    355
407; management reform and, 17;       Revenue, 14; Chile and, 481; C6te
Mexico and, 494, 495, 496, 501, 502,    d'lvoire and, 299-300, 303, 306;
503-4, 505, 516; Pakistan and, 414,     deficits and cuts in, 26-27; fiscal and
415, 417, 420, 421, 423, 424, 426,      exchange rate policies and, 30; Ghana
432; public enterprise pricing and, 12;  and, 312, 316, 324, 327; Indonesia
social welfare and, 254; staff          and, 363; Pakistan and, 417, 424,
reduction and, 256-57, 316; Turkey      425, 437; Zambia and, 340-42, 349,
and, 32-33, 110, 116-17, 120, 121,      354. See also Taxes
122, 123, 125, 441, 444, 447, 449,    Riots, 262, 346, 356
450, 451, 452, 456, 464; Zambia       Rocha, Roberto de Rezende, 102
and, 333, 334, 335, 338, 339, 344,    Rural sector: Indonesian social welfare
345, 347, 348                           analysis and, 378; Mexico and, 506;



Index 563
Pakistan and, 417; poor in, 248;      Stabilization, 53; as adjustment policy
Turkey and, 455                         response, 11, 546, 547; Chile and,
468, 477-78; Cote d'lvoire and price,
Sachs, Jeffrey, 383                       298; defining, 11-12; Ghana and,
Salaries: cuts in public sector, 256-57;  312; imF-supported, 216, 228, 247;
Ghanaian civil service, 316;            inflation and, 31; inward-looking
Indonesian civil service, 363; Ivoirian  strategies and, 13; Korea and,
civil service, 287; severance pay and,  391-95, 402, 404, 406, 409, 411;
248, 258; Turkey and, 455, 461. See     Latin America and, 216; Mexico and,
also Wages                              494, 507, 509, 511-12, 516; Pakistan
Salinas administration (Mexico), 509      and, 416, 431; reform and, 207-8;
SALS (structural adjustment loans). See   Turkey and, 444, 449, 461, 465;
Adjustment loans                        Zambia and, 334, 342, 345, 348
Salter, W., 382, 383                    Standard of living: and analysis of
Savings, 206; Chile and, 476, 481, 490,   adjustment, 12; Ghana and, 320-22;
493; CGte d'lvoire and, 303; deficits   LSMS and, 255; reduced, 200; Zambia
and fiscal crises and, 20, 22, 27, 28,  and, 350
44; Ghana and, 318, 330, 331;         Statistical evaluation. See
Pakistan and, 415, 424, 425, 437;       Macroeconomic performance
trade reform and, 65; Zambia and,       evaluation
340                                   Steel, William, 164
Scitovsky, Tibor, 72                    Stewart, Frances, 249
Scott, M. F., 72                        Structural adjustment. See Adjustment
SECALS (sectoral adjustment loans). See   entries
Adjustment loans                      Structural adjustment lending. See
Senegal, 17, 263; public sector reform    Adjustment loans
and, 109, 111, 112, 114, 115, 118,    Structural adjustment programs, reform
121, 122, 123, 125; trade reform and,   under. See Agricultural reform;
50, 73                                  Exchange rate; Financial sector
Sengupta, Arjun, 526                      reform; Fiscal reform; Industrial
Service sector: expenditure cuts and      reform; Policy reform; Public sector
social and economic, 26; Ghana andt,    reform; Trade reform
318                                   Subsidies, 26, 30, 155, 206; adjustment
Severance pay, 248,258                   Fprograms and, 32; for agricultural
Severance pay, 248, 258                   inputs, 140, 144, 154, 453; Chile and,
Sierra Leone, 328; exchange rates and     486, 487, 492; credit, 88, 90, 104,
deficits and, 30; trade reforms and, s2  140, 146, 266, 397-98, 4S2; export,
Social services, 263, 267, 525; Indonesia  307, 418, 448, 449, 463; fertilizer,
and, 378; reduced, 247, 257, 259-60     34, 154-55, 417, 428; food, 143-44,
Social welfare: adjustment measures       258, 262, 505; in form of price
affecting, 256-57, 525, 538-39;         controls to encourage consumption,
Chile's adjustment and, 48, 484-86;     142, 143; health program, 259;
costs of external disturbances, 227,    housing, 265, 295, 298; Indonesia
244; Ghana's adjustment and, 326;       and, 365; interest rate, 90, 92, 486,
measures to protect the poor and,       505; Korea and, 389, 397-98;
257-63; methods of measuring,           Mexico and, 502, s04-5, 506, 509;
249-56; new poor created by             Pakistan and, 414, 415, 417, 418,
adjustment programs and, 247-49;        419, 420, 428, 434; phasing out of
and poverty in Cote d'lvoire with       fertilizer, 34; the poor and, 157;
adjustment, 264-65; and poverty in      social welfare and, 256; Turkey and,
Peru without adjustment, 266-67;        447, 448, 449, 4s2, 453, 463;
Zambia's adjustment and, 350, 357       Zambia and, 333, 334, 335, 344, 346,
Somalia, trade reforms and, 52            350
Sri Lanka, 262                          Swan, T. W., 382, 383



564 INDEX
Tanzania, 18, 355; agricultural reform    fiscal reform and, 30; Ghana and,
and, 135                                315, 320; highly indebted countries
Tariffs, 30, 35, 39, 87, 206; agricultural  and, 201; Indonesia and, 361,
reform and, 145; Chile and, 473, 481;   365-66; Korea and, 47, 49, 53, 54,
C6te d'lvoire and, 296, 298, 307;       57, 58, 59, 67, 73, 389, 390, 391,
Ghana and, 327; industrial reform       397, 398-99, 403, 407, 410-11;
and, 161, 162; Korea and, 398, 399,     liberalization, 50, 77, 133, 144-46,
403; Mexico and, 500, 501, 513;          147, 162, 177, 178, 208, 216; Mexico
Pakistan and, 212, 418, 426, 432,       and, 48, 50, 52, 54, 57, 58, 59, 67,
434; reduction in, 201; trade           494, 495, 496, 497-501, 507-8, 510,
liberalization and, 36; trade reform    512-14, 516; during 1970s, 530;
and, 47-48, 50, 53, 54, 57, 67, 74;     Pakistan and, 414, 415, 418, 437;
Turkey and, 449; Zambia and, 334,       promotion, 159; taxes and, 36-37,
335, 337, 343, 344                      40; Turkey and, 48, 50, 54, 57-58,
Taxes, 12, 14; adjustment programs        59, 440, 446, 447-49, 452, 453, 460,
and, 535; Chile and, 481, 485, 490;     463; World Bank and policy for, 538,
C6te d'lvoire and, 299-300, 303,        540; Zambia and, 332-33, 334, 337,
306, 307; export, 30, 37, 50, 52, 53,   343, 344
74, 142, 309, 316, 335, 340, 497;     Trade reform: constraints on
farmers and, 18, 200; financial sector  implementing, 52-54; degree of
and, 78, 79, 85, 88, 105; fiscal reform  restrictiveness and, 47-48; effects of
and, 34-38, 44; Ghana and, 312,         change in policy and, 65-67;
316, 324; income, 35, 38; inflation,    exchange rates and, 54-55; exports
16; Indonesia and, 363-64, 369;         and, 59-65; implementation record
industrial incentives and, 205;         of, 49-52; imports and, 55-58;
industrial reform and, 162; Korea       incentives and, 54-58; performance
and, 406; Mexico and, 495, 502;         before and after loans and, 58-59;
Pakistan and, 415, 417, 420, 425,       positive effect of, 67-69; proposals
434, 436, 437, 438; poverty programs    for, 48-49
and, 258, 260, 264; price stabilization  Tranche-release conditions, 199, 202,
and, 142, 153; public investment and,   208-10, 326, 336, 343, 344, 345,
27; trade liberalization and, 36-37,    429-30, 456, 486, 510, 511, 536,
S4; value added, 37-38, 339, 457,       546. See also Conditionality
481, 485, 535; Zambia and, 334,       Transportation, 139, 316, 505
335, 337, 340                         Tunisia, 262; public sector reform and,
Taylor, Lance, 465                        109
Technology, 13, 397, 407, 409, 410;     Turkey, 14, 16, 17, 205, 206, 211, 213,
industrial, 160, 187, 190; price        279, 282, 283, 284, 534; adjustment
changes and, 153-54, 155                program in, 441-47, 463-65,
Thailand, 205, 206, 211, 212, 213;        466-67, 537, 538; conditionality and,
deficits and, 28; industrial reform and,  456-57, 465; economic performance
166, 175, 176; tax reform and, 36;      and adjustment and, 447, 456;
trade reform and, 52, 59                fertilizer subsidy in, 34; financial
Togo, 17; public sector reform and, 110,  sector reform and, 76, 87, 452,
112, 115, 117, 119, 120, 121, 122,      454-55, 461; fiscal deficits in, 24, 29,
123, 125, 128                           45, 441, 449, 450, 461; IMF and, 440,
Trade: adjustment policies and, 14-16,    441-44, 449, 457-59, 460, 465, 467;
534; bank failure and liberalization    industrial reform and, 161, 166, 173,
of, 77; Chile and, 47, 49, 53, 54, 57,  174, 176; public sector in, 32-33,
67, 473, 475, 480-81, 487, 491;         110, 116-17, 120, 121, 122, 123,
financial sector reform and, 86; fiscal  125, 441, 444, 447, 449, 450, 451,
deficits and trade shocks and, 24-25;   452, 456, 464; successful adjustment



Index 565
and, 208; trade and, 48, 50, 54,       526; external financing and, 547-48;
57-58, 59, 440, 446, 447-49, 452,      financial sector reform and, 76, 79,
453, 460, 463; VAT and, 37, 38, 457;   80, 84, 89, 90, 91, 93, 100, 102, 103,
World Bank and, 441, 446-47, 449,      104, 105, 106; fiscal deficits and,
451, 453, 454, 455, 456, 457-59,       23-24, 39-40; fiscal issues and, 22;
460, 461, 464-65, 466-67               Ghana and, 311, 314, 315, 316, 318,
322, 323, 324, 325, 326, 330, 331;
Uganda, trade reforms and, 52            governmental tradeoffs and, 523, 524;
Unemployment, 247; Chile and, 472,       government commitment and, 198;
490; industry and urban, 160; Mexico   growth and development and, 534;
and, 494, 509; Pakistan and, 423;      how adjustment has worked and,
Zambia and, 347, 350, 357              533-34; Indonesia and, 378-80, 384;
Unions, Mexican food stamp program      industrial reform and, 159, 160, 164,
and, 144                               174, 175, 177, 179; interest rate
United States, 216, 445                 reform and, 535; internal procedures
Urban sector: Indonesia and, 377, 379;   of, 528-29; internal review and, 197;
Mexico and, 509; Pakistan and, 414;   introduction of adjustment lending
poor in, 248, 254, 326                 and, 1-2, 193, 544; Korea and, 397,
Uruguay, 138, 142                        399, 402, 403, 411; Latin America
Utilities, 17, 267; price or rate increases  and, 518-19; level of funding and,
and efficiency and, 34                 542; loan review and, 528;
macroeconomic performance and,
Veneroso, Frank, 101                     223, 225, 226, 227, 228, 232, 233,
Venezuela, 263                           237, 239, 243, 244; Mexico and, 496,
497, 500, 503, 506, 509-11, 512,
Wages, 247; Chile and, 469, 485, 490,    513; operations and maintenance
492; Cote d'lvoire and, 302, 359;      expenditures and, 32; ownership and,
Ghana and, 312; Indonesia and, 371,    545-46; Pakistan and, 415, 416, 417,
377, 378; Korea and, 393; Mexico       418, 420, 425, 428, 429, 430, 431,
and, 501, 506; Pakistan and, 423;      432, 433, 434, 435, 436, 437, 438,
Turkey and, 441, 455, 456, 461;        439; policy and, 537; project-based
Zambia and, 334, 335, 338, 340. See   lending and adjustment and, 5-7;
also Salaries                          public enterprise reform and, 109,
Western Europe, 128, 162                 110, 112, 114, 117, 118, 119, 120,
Westphal, Larry, 536                     122, 123, 125, 128; risks involved in
Women: household surveys and,            adjustment loans and, 541-42, 546;
254-55; new poor and, 248              social consequences and, 247, 248,
Workers' remittances, 414, 426, 427,     254, 261, 272; subsidies and, 32;
429, 432, 434, 440                     trade policy and, 538; trade reform
World Bank: adjustment problems and,     and, 30, 47, 72, 73, 74, 75; Turkey
526-28; and advice on adjustment,      and, 441, 446-47, 449, 451, 453,
211-12, 214; agricultural reform and,  454, 455, 456, 457-59, 466, 461,
131, 132, 136, 137, 141, 147, 148,     464-65, 466-67; VATS and, 38;
149, 152; Articles of Agreement and,   Zambia and, 334-35, 336, 342, 343,
532; and assessment of adjustment,     345, 347, 348, 349, 352, 353, 354,
2-4, 44-45, 222; Chile and, 473,       355, 358
474, 475; conditionality and, 199,   World Bank Living Standards
200-201, 202, 203, 206, 207-8, 209,    Measurement Study (LsMs), 255, 266
210, 211, 213, 215, 218, 219; control
and, 528; C6te d'Ivoire and, 288,    Yeats, Alexander, 48
289, 299, 306; economic analysis and,  Yugoslavia, 209; trade reforms and, 52,
525, 540; evaluation and, 193, 194,    53, 58



566 INDEX
Zaire, 250                                341; food subsidies in, 262; historical
Zambia, 206, 209, 211, 213, 279, 282,     overview of, 352-55; industrial
284, 546; adjustment program in,        reform and, 166, 173, 339; political
334-38, 355-58, 537; adjustment         support for adjustment program and,
program design and, 348-49;             349-50, 356; social impact of
conditionality and, 343-45; economic    adjustment and, 350; trade reform
crisis in, 332-34; economic             and, 52,53,58,59,332-33,334,
performance and adjustment in,          337,343,344
338-42; and evaluation of             Zimbabwe: fiscal reform and, 30,
adjustment, 342-47; fiscal deficits     40-41; social welfare analysis and,
and, 24, 30, 333, 334, 335, 336, 340,   250; trade reform and, 53






E  -The World Bank.
E  xperience has shown that even the best-designed development
project-whether a dam, a highway, or an irrigation scheme-
might not yield the expected results if it had to be undertaken in a
poor economic environment. In 1980 the World Bank began structural
adjustment lending to its member governments to support broad eco-
nomic, institutional, and policy reforms. It was hoped that these adjust-          
ment loans-by restoring the balance of payments, strengthening institu-  t    J
tions for managing the economy, and bringing pricing and other policies
in line with the world market-could generate sustainable economic
growth and alleviate poverty in developing countries.
To examine how well structural adjustment lending has worked,
this volume brings together leading practitioners and scholars from
the World Bank, from other international financial institutions, from       0
government, and from academia. Their assessments, based on theoret-      *
ical and practical experience with these programs, contribute to the  h I
ongoing debate about how best to bring about macroeconomic stabil-
ity and sustainable growth.                                        C.        "
Adjustment loans are made on the understanding that specific reforms
be carried out in such areas as public finance, public enterprises, and
trade policy. The authors review these conditions and how they have
been designed and implemented. They also consider the social and politi-
cal consequences of adjustment, particularly for the poorest people,         9
and describe programs designed to protect the vulnerable. Case studies
of nine countries under adjustment explore how well they complied
with loan conditions and how their economies performed before,
during, and after adjustment.
Vinod Thomas is staff director of the World Bank's World Development
Report 1991. Ajay Chhibber is with the Bank's West Africa Depart-  t.
ment. Mansoor Dailami is with the Bank's India Department. Jaime de
Melo, affiliated with the University of Geneva and the Centre for Eco-
nomic Policy Research, is in the Bank's Country Economics Department.        J .-YrL    
:                                                              'Oxford Umerit Prs  4iLi; r_;
-'0ve9 desVgn        WT
-~ .~-- -~~~'-ISBN 0-19-520870O 6